I don’t need to explain the attraction of low power operation; if you’re reading this, the chances are that you are already a convert. I’ve been operating with low power ever since first being licensed in the UK in the late 70’s as G8RYQ, and then G4IFA. One of my first rigs was a homebrew VFO-controlled FM rig for 2M. I don’t remember how much power it put out, but it was at most only a few watts. Then there was an 80M DSB rig, built from a kit, that put out a watt or two. I had a series of 2M FM rigs, including an Icom IC-22A and a Trio TR2200 (1 watt out). I had a hand-rotated 5 element 2 meter beam. To change the beam heading, I leaned out of my bedroom window and twisted the aluminum pole that was supporting it. One of the PA transistors in my Icom IC-22A was blown, so the rig only put out 4 or 5W. I remember using that rig and the beam to talk regularly on simplex with a fellow young ham who was in Wales, 90-100 miles away. I think I used the TR2200 to talk with him as well, which was even more impressive, as the Trio only put out 1 watt of RF power.

My checkered amateur radio life did include one 100W rig. It was a TS520 that I owned for a couple of years in the early 1990’s. Other than that though, every rig I have ever built or owned has been 5 watts or less. After a while doing QRP, 5 watts becomes the norm. 5W is known as “the full QRP gallon”, and it does feel like it! I still run 5W as my default most of the time, on both CW and SSB. Recently though, I’ve been turning the power down, to see how lower power levels get out. A fun moment recently, was when I clearly heard the backwave from my little two-transistor transmitter on the KPH SDR, which is 41 miles away as the crow flies. That backwave was about 1mW, and being able to hear it on a remote receiver was something of a revelation. It was that moment that kickstarted my interest in even lower power levels than the mighty force of the full QRP 5 watts.

Once a day, I check into The Noontime Net on 7284KHz. It is virtually the only time I use SSB. They welcome check-ins from QRP stations. Once a year, they have a QRP day, when operators are encouraged (though not required) to check-in using QRP power levels. There is an honorable mention on their website for the station who checks in using the least power and this year, I took the prize for checking in with 10mW of SSB. The check-in was with Don KY7X in Wellington, NV. The distance between us is 168.5 miles as the crow flies. At an equivalent distance of 16,850 miles per watt, I think that would easily qualify for the QRPARCI 1,000 miles per watt award. Given that most members who apply for that award have achieved it with CW, I think that a QSO of 168.5 miles with 10mW of SSB is even more inspiring.

The lowest power I can turn my Elecraft K2 down to on SSB is 1 watt. I achieved the power of 10mW out by connecting an inline attenuator with a fixed attenuation level of 20dB in the antenna lead. It is one of those “barrel” attenuators, with a BNC connector on each end. With the success of a check-in with 10mW under my belt, I resolved to try for even lower power in next year’s QRP Day. This led me to a nifty little kit offered by QRP Guys. It is an inline attenuator, with switchable levels of attenuation of 10, 20, and 30dB. Also included is a bypass switch that allows the operator to easily switch the attenuator out of circuit when on receive. For $25 + shipping, it was the obvious solution to my QRPp needs. I was very close to pulling the trigger, when my “QRPp extreme sports” gene kicked in, and I thought it could be useful to be able to attenuate a signal by an extra 10dB, for a total of 40dB attenuation. This would reduce the 1 watt of SSB from my K2 down to the truly flea power level of 100µW, and the 100mW CW output to the mind-bogglingly low level of just 10µW! Granted, this extra 10dB of attenuation may never be needed, but if and when I succeed in making a QSO with 30dB of attenuation in the antenna line, I may always wonder if it could also have been made with an extra 10dB. It’s the desire to constantly push our achievements just that little bit further.

With that in mind, I decided to build my own attenuator box. There is no circuit design involved really, as it is simply a series of 50 ohm pi-attenuator pads and a few DPDT switches. I took the circuit from the QRP Guys’ attenuator and added an extra pi-section and switch –

All resistors are metal film 3W types. The 100 ohms ones are 1%, while the other values were 5%. If you can get 1% tolerance for all values, then all the better. If you want to calculate resistor values for any other degree of attenuation, you can use this online calculator.

There’s not much to the build. The diecast enclosure, switches, and BNC connectors all came from Tayda. I am not thrilled with the quality of the connectors and switches from Tayda. The terminals were hard to solder to, presumably due to the use of a cheaper alloy than the quality brands such as Kobiconn and Switchcraft use. Nevetheless, I persevered, and managed to obtain a reasonably satisfactory result.

The bypass/attenuate switch is useful when going from transmit back to receive, for ensuring that your reception is not also attenuated.

Daytime band conditions weren’t too good on first finishing this attenuator, though I did manage to just be heard by Don KY7X, 168.5 miles away, with an output power of 10mW SSB, using 20dB of attenuation from an original 1W signal. I’m not sure if it would have been enough for a positive ID of my signal if he didn’t already know who I was. Nevertheless, band conditions were poor that day, so this was a good sign. I decided to hook it up to my VK3HN WSPR beacon (thanks Paul), which puts out 200mW. I applied 30dB of attenuation, for a 200µW WSPR signal – that’s just 0.2mW! Incidentally, to figure out the various attenuation levels, and what output power they give you, there are several online calculators. I found this one to be useful.

Unfortunately, the lowest power level that can be encoded into a WSPR signal is 0dBM, equivalent to 1mW. I’m not keen on misrepresenting the power level, but as I was so eager to see what a mighty 200µW of WSPR could snag me, and as I considered a power level of 20mW (10dB of attenuation of the 200mW signal) to be too high, I decided to WSPR for the night on 40M, and encode the signal at 0dBm. Check out the following results from a night of WSPR’ing. There were 486 spots in total. This is not a lot by normal standards but a good result, I think, for such a low power signal. In this screen grab, they are sorted in order of distance, so these are the most remote spots. AI6VN/KH6 in Maui tops the list, for a distance of 3778km = 2347 miles. That’s 11.735 million miles per watt! In a normal night of WSPRing on 40M with the relatively high power of 200mW, I would expect multiple spots from Hawaii, VK and ZL land, as well as a spot or two from DP0GVN in Antarctica. However, 0.2 mW is a whole new ballgame, and I was very happy to get just one spot from HI –

Here’s the attenuator sitting on my desk, on top of the VK3HN WSPR beacon. It would be nice to have a tidy desk and a nice, clean operating position but every time I tidy it, it slowly gets like this again (the 3rd law of thermodynamics in action!) At the bottom of the stack is the Sproutie SPT Part 15 Beacon, which is currently not QRV. DK, if you’re reading this, you may notice something familiar at the very bottom of this picture –

I want to be able to WSPR on a more regular basis, and have the encoded power information on my transmissions actually be fairly accurate, so the next step was to build an attenuator with a fixed attenuation level of 23dB, to reduce the output of the WSPR transmitter to 1mW. This way, when a spot from me shows up with a power level of 0 dBm, it actually is 0 dBm. This online pi-attenuator calculator was pressed into service, and yielded the following values. If you don’t have a 12 ohm resistor, then a single 330 ohm part will be close enough –

As this attenuator will only be used to attenuate the 200mW output of the WSPR beacon, I used 1/4 W resistors. The two 56 ohm resistors were left over from a cheap resistor kit that I bought years ago. They had short, thin leads, and were ostensibly 1/4W parts. The left-hand one (the one closest to the transmitter) dissipates the most amount of power, and was getting very warm during 2 minute transmissions from the 200mW transmitter. I do think it was sustainable, but would have preferred it to run cooler. I didn’t have any 1/2W or bigger resistors in appropriate values, so decided to parallel 2 x 1/4W parts. There are plenty of fresh sheets of white paper here, but drawing schematics on envelopes is more fun. A 100 ohm and 150 ohm resistor in parallel makes 60 ohms. Using 352 ohms as the “top” resistor (achieved with a 330 and a 22 ohm resistor in series) makes for an attenuation level of 22.7dB, which is pretty dang close –

The two resistors on the left-hand side run much cooler than the single 56 ohm did. The 56 ohm one was a cheapie resistor, and I suspect it’s stated power dissipation of 1/4W was being a bit optimistic. All the resistors in the final attenuator are 1/4W metal film types. The project box came in a pack of 5 from Amazon, for $7.50. I have used the same ones recently to build QRP baluns and ununs. There are all sorts of fun and novelty projects they would be useful for. The lid snaps on. I can supply the link to anyone who is interested –

The increase in power from 200µW was noticeable after the first night of WSPR’ing on 40M with 1mW. Here is a screen grab of the most distant spots received. In just under 10 hours, I had 1114 spots. It’s a lot fewer spots than what I would receive with 200mW, but that many spots with just 1mW sounds very encouraging. I love this little WSPR beacon (thanks to Paul VK3HN once again). Out of 1114 spots, all of the drift figures were a big honking zero, with the exception of a single 1 and a single -1. Instead of the single spot from AI6VN’s remote listening station in Maui, I now had 7. Sure, propagation on different nights could be some of it, but I’m pretty sure the 7dB power increase from 200µW to a gigantic 1 milliwatt was a significant factor.

This QRPp experiment has been a huge success so far, and I haven’t even begun to work on the goal that I had in mind when beginning this. That was to use the switched step attenuator to see how far I can go with very low power on CW, my mode of choice. WSPR is very instructive and interesting, but an actual QSO, even a brief one, carries the extra appeal of contact with a distant person, with all the unpredictable intangibles that come along with that. I do wish there was a way of encoding lower powers than 0 dBm (1mW) into a WSPR transmission, as I would then be WSPR’ing with successively lower powers. I’ve already received a significant number of spots with 200µW of transmitted power. It would be great to see what could be done with, say, just 10µW (0.01mW), if anything. In the meantime though, there will be a lot of 1mW WSPR’ing emanating from the AA7EE radio ranch, as well as some extreme QRPp CW too, with the help of the switchable step attenuator.

As well as the small stash of finished projects that grace my living space, I also have two small boxes containing various boards. Some of them are boards from part-finished projects that didn’t work. For whatever reason, I ran out of steam and, instead of troubleshooting them, put them carefully into a small box along with their cellmates, and conveniently put them out of my mind. A few of the boards actually worked, but I decided not to case them up. One of these is the two-transistor transmitter I built a few years ago, that was basically the TX side of the Pixie 2 design. I like to pull boards like this out of the box from time to time and power them up. Then they go back in the box, waiting for the next time I feel partial to some extra-curricular fun.

Extricated unceremoniously from it’s cardboard box on the shelf, this little transmitter (pictured above) was putting out about 200mW when connected to a 12V supply. I got a real kick from hearing the signals from this simple circuit on the KPH online SDR at the Point Reyes coastal station, 41 miles away as the crow flies. This design has the oscillator running continuously. In the Pixie, from which it was taken, the oscillator is needed on receive as well. As the circuit consists of just an oscillator and a PA stage, there is a small amount of oscillator leakage into the PA even when the PA stage is not being keyed. Imagine my surprise on discovering that I could hear this backwave on the KPH online SDR! Putting the transmitter on my OHR QRP Wattmeter revealed that although accurate measurement was not possible at such low power levels, it looked as if something of the order of a milliwatt of RF energy was making it’s way to the antenna when the transmitter was not being keyed. Now, the fact that 1mW could be heard 41 miles away, though impressive, is by no means unprecedented. Nevertheless, it fired up my imagination, and made me want to build another little transmitter.

The train of logic that I followed in order to finally settle on building GM3OXX’s little OXO Transmitter was, as it turns out, not very logical and quite convoluted. I will not attempt to explain it here, as it will only serve to confuse! However, I did want a design in which the oscillator wasn’t running on key-up, so that I could monitor the signal in the receiver to serve as sidetone.

The OXO transmitter was first featured in the Autumn 1981 issue of SPRAT, the journal of the G-QRP club. The original circuit didn’t include an LPF; the builder was expected to provide their own. Here is the original circuit, with the addition of an LPF –

The BCY39 PNP transistor keys the +12V supply to the 2N3866 PA. It is not necessary to key the oscillator, as the oscillator won’t run unless the PA transistor is switched on. In the LPF, the 1.38µH inductors can be made with 21 turns on a T37-6 toroid, and the 1.7µH with 24 turns on a T37-6 toroid. Values were taken from the assembly instructions for the QRP Labs low pass filter kits. Numbers of turns for different toroids can be figured out from the very useful calculators on the Kits and Parts website.

I mocked it up on a breadboard and it worked well. I noticed that the VXO, although not oscillating on key-up, was emitting some very low level spurii. In retrospect, this could well have been due to stray capacitances in the breadboard, and the fact that the circuit wasn’t built over a ground plane. Nevertheless, I decided to have the keying transistor key both the oscillator and PA. This is what I came up with, drawn on the back of an envelope –

George GM3OXX added a 0.1µF cap across the key contacts to help with shaping. It wasn’t on the original schematic as published in SPRAT, but I added it here. The RFC in the collector of the PA transistor can be a molded choke. I wound 17 turns on an FT37-43 toroid to serve the same purpose. I also added a spotting switch, with the 1N5817 diode to prevent the PA from being switched on when only the oscillator signal is wanted, for “netting” the transmitter frequency on a receiver. Unfortunately, the spotting switch didn’t work out quite as planned. More on that later.

I had an old LMB Heeger 143 enclosure lying around from my build of N6KR’s SST. I had drilled the front panel holes in the wrong places, so ditched it and used a fresh enclosure. I had considered this enclosure to be unusable for another project, until realizing that I could simply place a piece of PCB material over the front panel to cover up the unused holes, and keep it attached with the nuts and screws that were holding the controls in place. It worked well and looks pretty good. The slide switch is for transmit/receive switching, the red button is for spotting, and the big knob is the VXO tuning –

The board was scrubbed with a steel wool soapy pot scrubber and given a couple of thin coats of clear spray-on lacquer. A fresh board holds so much potential, and never looks as good as before construction begins. It’s almost a shame to glue any pads onto it!

The slide switch was part of an order that I received from Dan’s Small Parts & Kits back in 2011. The polyvaricon was from the VRX-1 direct conversion receiver, designed by NT7S and kitted by 4SQRP. It was one of my first ever Manhattan construction attempts and, although the receiver worked well, I wasn’t happy with my layout and the way my build looked.

All the back panel connectors (key jack, DC power connectors, and BNC’s) were from Tayda. I’m not blown away by the quality of the connectors from Tayda but considering the very reasonable prices, I am making an exception.

I built the VXO first, and tested the frequency coverage. Most any small signal NPN transistor will work in this position. I used a 2N3904. A 2N3866 was used in the original circuit for the PA. I didn’t have one of those, but I did have some 2N3866 equivalents in the form of the Motorola 4-247 CG9949. A while back, the G-QRP club were giving away small quantities of these transistors for free to their members. I took advantage of the offer, knowing that I could use a 2N3866 equivalent or three. This is the exact same transistor that Kanga UK are using for their kit version of the OXO transmitter.

At this point, after having built the VXO and PA circuits, the OXO will function as a transmitter, by keying the +12V line. If you only ever intend to use a manual key with this transmitter, it is not necessary to build the keying switch, and you’ve got yourself a nice and simple two-transistor transmitter. However, I like using a paddle, and the keying circuit is very simple. Here’s the board with the basic transmitter built. You can see the crystal, in a holder made from an SIP strip, along with the VXO transistor, right next to the polyvaricon. The PA transistor is the one with the big honking heatsink on it, just behind the toroid. The keying transistor, a 2N3906, is to the right of the toroid –

For transmit/receive switching, a primitive solution was found, in the form of a DPDT slide switch that came from Dan’s Small Parts and Kits years ago, and has been sitting in one of my parts drawers, just waiting for an opportunity to be used. It was wired up like this (another diagram drawn on the back of an envelope!) –

On receive, the antenna is connected to the antenna input of the receiver. The output of the transmitter is connected to a 50 ohm dummy load. In case the transmitter is accidentally keyed while in receive mode, it’s output will be protected by the 50 ohm load. On transmit, the output of the transmitter is routed to the antenna, while the receiver antenna input is connected to the 50 ohm load. The receiver is being used for sidetone, so the idea behind this is to do whatever can be done to prevent receiver overloading while in transmit mode.

The OXO transmitter, all wired up and ready to go. I used an LPF from QRP-Labs. Band changing can be accomplished by plugging in a different crystal and plugging in a different LPF –

Regarding the PA emitter resistor that is marked as 39Ω. You can fine tune the value of that resistor, depending on the output power you want. Be careful not to go too low, or the transistor could overheat and be destroyed. In the Kanga UK kit version that uses the exact same PA transistor, the emitter resistor is 2 x 33Ω resistors in parallel, for a total effective resistance of 16.5Ω. In the build instructions, Paul reports that with a 13.8V power supply, he gets 1.5W out on 80M and 1W on 40M. I was cautious, and began with 2 x 100Ω resistors (=50Ω). I added resistors in parallel, until I got to 4 x 100Ω (=25Ω). My OHR QRP Wattmeter indicated an RF power out of 400mW. My NM0S QRPometer indicated a power out of 580mW. I wasn’t sure which one was more accurate, so split the difference and called it 500mW. I could get more power by adding another couple of 100Ω resistors, but I rather like the idea of having a 500mW transmitter. By comparison, 1W seems so pedestrian! With an emitter resistance of 25Ω, the voltage across it was 3.2V. According to ohm’s law, the current through it was 128mA, for a dissipated power of 0.41W. The resistors are 1/4W parts, so their total power dissipation capability is 1W. Sounds well within the margin of their capability.

In the following picture, you can see the two 100 ohm 1 watt resistors that form the 50 ohm dummy load in the background –

The heatsink on the PA transistor is probably overkill for a power output of just 500mW, but it gives a good safety margin. I held the key down for 2 minutes (into a dummy load, of course), and it only became mildly warm to the touch. I think it would safely survive even if my cat fell asleep on the key

The two 12V DC connectors on the back panel are wired in parallel, to allow one 12V lead to power both the transmitter and a companion receiver. It is not shown in the schematic, and cannot be easily seen in any of these pictures, but a 1N5817 diode is wired between the +ve side of the 12V DC connectors and the board, for reverse polarity protection.

I paired it up with my Rugster direct conversion receiver and QSO’ed with K6KWV in Diamond Bar, CA on 40M – a distance of 363 miles as the crow flies. Not mega-DX, but it was a very enjoyable contact. He said that I was the first contact he’d had with a station using a homebrew rig, and I was also running the lowest power of any station he’d QSO’ed with. That was nice to hear! He has only been a ham for 4 months, and has a good fist. His code is pleasant to listen to, and easy copy.

When using the Rugster, as well as throwing the TX-RX switch when going from receive to transmit, I also have to turn the RF gain control down to zero to prevent the receiver from overloading. If on headphones, I also have to turn the AF gain down somewhat. Although it is an easy process to get used to, I find my Belka-DX even easier. Due to the AGC in the Belka, I don’t have to change a thing when going from receive to transmit, and vice-versa, other than flipping the TX-RX slide switch on the OXO transmitter.

The CW note sounds good on 40, and chirp-free. There is a little chirp with my 14060 crystal, and a lot with the 21060 and 28060 crystals. I credit this to the PA loading down the oscillator, and figure that a buffer stage between the oscillator and buffer would cure it. Thanks to John KC9ON, at 3rd Planet Solar, I have a pack of 40M crystals that are in HC49/S cases – the short cases. My 14060, 21060, and 28060 crystals are all in the tall HC49/U cases. Perhaps the higher band crystals are not fundamentals?

On a related note, comparing my HC49/U 7030 crystal with the HC49/S one, the tall crystal pulls over a wider frequency range. I didn’t think to measure the capacitance of the polyvaricon before installing it, but I think it is 270pF per gang. Putting both gangs in parallel made little difference to the pulling range. With the tall HC49/U 7030 crystal, the range was 7028.77 – 7032.83KHz, representing a swing of 3.61KHz. By contrast, the short HC49/S crystal pulled from 7029.57 to 7031.1KHz – a swing of just 1.53KHz. Pulling range became increasingly greater with the higher frequency crystals. The 28060KHz crystal pulled over a range of 13.42Khz, but chirped so much it was comic.

Oh, about that spotting button. I thought I’d be able to use it to net the transmitter precisely on a received station’s frequency. Unfortunately, the frequency of the VXO is significantly lower in spot mode than on full key-down. On 40M with a short crystal, the difference is 250Hz. I assume this issue wouldn’t exist with a buffer stage between the VXO and the PA.

For the time being, I am going to give this project a rest and concentrate on other things. We all need a break sometimes. However, if and when I revisit this little transmitter, I’d like to rewire it so that just the PA is keyed, as was intended with the original circuit. Given that the oscillator doesn’t run unless the PA transistor is switched on, I see no reason not to do this. It is an easy change to make. I’d also like to acquire and try different crystals for the higher bands, in the hope that will eliminate the chirpiness.

Another future possibility would be to add an SMA connector on the back panel for an Si5351 VFO. At that point though, the project is becoming more complex, perhaps negating the point of such a simple transmitter to begin with.

Some years ago, I purchased and assembled an Oak Hills Research WM-2 QRP Wattmeter from Milestone Technologies. As far as QRP wattmeter kits go, it was something of a classic at the time, and as such, I wanted one. I’m glad I made this purchase, as they are no longer available – at least, in this form. Another company is offering a very similar kit, but without the decals on the case. I was told that they acquired the rights to it from Milestone Technologies, so this would be appear to be a direct clone of the WM-2. The WM-2 is a great little wattmeter, with 3 ranges representing 100mW, 1W, and 10W full-scale – an ideal selection of ranges for the QRPer. You can read both forward and reflected power. While direct readout of SWR is not offered, it can be calculated from the forward and reverse power readings. The WM-2 has an analog meter and can be left inline while operating. Apart from the satisfaction of being able to see the needle bob up and down when transmitting, this type of indicator is very useful when peaking circuits for maximum output. It can certainly be done with a digital readout, but an extra stage of “translation” needs to happen in the brain, converting the number on the readout to a “level”. This process when looking at an analog meter is more immediately intuitive.

That was so many photos of my WM-2, that you might be thinking, “Hang on – isn’t this a post about the QRPoMeter? Well, it is – and we’ll get to that very soon. I don’t think I ever blogged about the WM-2 when I built mine years ago, so felt it was time to give it some air time on my blog.

For my purposes, the WM-2 meets my needs. However, I don’t have any other instruments with which to check the accuracy of it’s readings. A Bird wattmeter would be nice, but the expenditure is hard to justify. Another option is to use an oscilloscope to measure the peak to peak voltage a transmitter develops across a 50 ohm dummy load, and use that to calculate power. This is a definite possibility in the future, as I do intend to add a digital storage oscilloscope to the shack at some point. In the meantime, it would be good just to have another wattmeter of similar accuracy, simply to increase my confidence in the readings I am getting from either one. For the kind of operating us QRPers do, absolute accuracy is not essential. 5% of full scale is good enough which, on a 10W scale, means ±0.5W. If I claim to be transmitting with 5W, then the difference between 4.5 and 5.5 W is unlikely to even be noticed at the receiving end.

The QRPoMeter, designed by Dave Cripe NM0S, has been on my radar for a very long time. Originally offered by the 4SQRP group, it is a very affordable instrument for measuring power and SWR. It has a built-in dummy load, to make measuring the power into a 50 ohm load an easy task. Also, when measuring SWR, it uses a resistive bridge, so that the maximum SWR your transmitter will see is 2:1. I’ve long wanted to assemble this kit. A few times, I’ve waited too long to purchase a kit, only to find that it was no longer available (the SST, once offered by Wilderness Radio, was one example). With that in mind, and also because the QRPoMeter is so reasonably priced, I went ahead and placed an order with NM0S Electronics.

A few days later it arrived, in a small flat rate Priority Mail box. I love getting radio parts and kits in the mail. It’s exciting! The PCB pieces that, as well as forming the circuit board, also comprise the case, were slipped in between pages of the assembly manual to protect them. There was also a little bag of goodies. I love little bags of radio parts!

The bag of parts, emptied out into a styrofoam tray –

Also included was a piece of thin 2″ x 3.5″ PCB material, etched and silkscreened on both sides. On one side was the business card of NM0S Electronics. On the other side were these handy little band plans –

The pieces that form the sides of the case have to be broken off from the larger pieces of PCB material, and given a very light filing to remove the rough edges. It was immediately obvious how smart the final product was going to look. It was raining very, very lightly when I took this photo. If you look carefully, you might see some very small raindrops on the panels –

For anyone who has assembled a few kits, construction is uncomplicated. Dave’s instructions are clear and straightforward, consisting mainly of a checklist for populating the board, and instructions for constructing the included PCB case. The switches and input/output BNC connectors are all mounted directly on the board. The only wiring required is a 4-conductor ribbon cable that is used for the connections between the board and the LCD panel meter. Other than wiring up this meter, and soldering the case pieces together, construction of the QRPoMeter consists of populating the single PCB with the parts. This picture is of the finished board before the two switches were installed –

I only discovered two very slight issues during assembly. Neither will present a problem for anyone with a little experience, but they might slightly confuse a beginner. These were due to a change in sourcing parts, and Dave said he will take care of them in future versions of the construction manual. These were –

1. U4, the TLC2272, had no dot or u-shaped indentation to denote the correct orientation. I used the printing of the device number on the top of the IC as a guide instead, and this turned out to be right. See picture below, with U4 circled in red –

2. When using the 4-conductor ribbon cable to connect the LCD panel meter to the board, because the instructions refer to an earlier version of the meter, a beginner might experience some uncertainty as to which holes on the meter board to connect to which holes on the main board. On the main board, there are two pads next to each other marked +Vin and -Vin. These are connected to the pads on the meter board that are marked “IN” and “COM” respectively. The other two connections are more obvious. The pads on the main board, next to the schematic symbol for a 9V battery, marked + and -, are connected to the pads on the back of the meter board, next to the “DC 9V” text, that are marked + and respectively.

A portion of the board in greater detail, showing the 8 large surface mount resistors that form the dummy load/resistive bridge (or, to be more accurate, 7 of them, and a small part of the remaining one) –

Calibration is straightforward, and requires a fairly accurate DMM. I used my Brymen BM235 (the EEV Blog version). The only other piece of equipment needed is any HF QRP transmitter with an output of between 2 and 10 watts. The output power doesn’t need to be known, as long as it falls within that range. When the unit is calibrated, you have a very handsome and useful piece of QRP kit!

The QRPoMeter seems to be accurate enough for my purposes. Power measurements are in line with the ones reported by my WM-2, taking into account the accuracies of both instruments. SWR measurements are similar at the lower readings. They differ by fairly large amounts at higher SWR’s. This doesn’t concern me though, as once the SWR goes much above 2 or 3:1, it’s exact value is of little interest to me. I just know that I want to get it back down below 2:1! A useful feature of the resistive bridge in the QRPoMeter that is used to measure the values of forward and reflected power, is that when SWR is being measured, the transmitter never sees an SWR higher than 2:1. This was verified with the SWR indicator in my Elecraft K2.

Thanks Dave. A good-looking and worthy little piece of QRP test gear! The QRPoMeter is available from NM0S Electronics.

I first tried WSPR out in 2009, with a Signalink USB interface attached to my FT-817 and PC. For anyone interested in QRP and QRPp, the process of being able to decode a signal that is up to about 34dB below the noise level is quite fascinating. Morse code, sent by way of CW, engages and tickles my brain in ways that other modes don’t. WSPR though (and other weak signal modes), has it handily beat in terms of it’s sheer ability to extract data from a signal that the human ear cannot even detect. A few years later, in 2018, I assembled an Ultimate 3S QRSS/WSPR beacon transmitter from QRP Labs for a ham friend. This project opened me to the appeal of a standalone WSPR beacon that, unlike my earlier foray into WSPR, didn’t require tying up my main station gear. The addition of a GPS unit, as well as setting the timing of the transmissions, could also automatically insert the Maidenhead grid locator – no need to manually program that, making it ideal for travel.

Fast forward to the current day. I’ve recently become a bit more active on the bands, and decided that I wanted to “stop the rot” of my CW skills, which were slightly degrading due to lack of use. I signed up for an online CW course with the CW Academy, offered by CW Ops. I just completed their intermediate course, and enjoyed it immensely. The Intermediate course is designed to take ops from 10-20 wpm. I was already comfortably having conversational QSO’s at about 16-18 wpm. At CW Academy, the emphasis is on head-copying, so that you can converse without needing to write anything down other than the occasional piece of essential info (name, rig, etc.) This, they explain, is an important skill, if you are to increase your speed. I, along with most of the other students, found it surprisingly challenging to listen to short stories in code, and extract meaning from them without writing anything down. It helped that we had a fantastic advisor, in the form of Randy N1SP. Practice sessions in between our online Zoom sessions could be challenging, but the prospect of classes led by Randy were a great incentive. He made learning fun.

Along with my renewed interest in CW came interest in weak signal modes generally, as well as a slight stirring in the desire to build radio things again. Over the last 3 years, I’ve been putting time and effort into working on my camper van, which took energy and money away from amateur radio. Well, I’m gradually angling towards selling the campervan, which will free up some mojo for other pursuits. Anyone want to buy a 1993 Airstream B190, with 67K miles, 200w of solar on the roof, and a 2″ lift?

Back to radio. The Autumn 2022 issue of SPRAT contained an article by Paul VK3HN, detailing the WSPR beacon he had built using modified open source code from Harry at ZachTek and, of course, the JTEncode and Si5351 libraries from Jason NT7S (Jason’s libraries pop up everywhere). If you don’t have access to SPRAT, and even if you do, Paul describes his beacon on his blog here.

As long as you know how to upload a program to an Arduino, or flash firmware to a microprocessor (same thing), the barrier to entry to building a WSPR beacon is now quite low – even lower if you don’t build a PA stage, and take the ~10mW output from the Si5351 clock output directly to the LPF and the antenna. Here’s what I built –

The output is taken from the CLK 0 output of the Si5351 and feeds directly into the PA stage that Hans Summers uses in the QRP Labs Ultimate 3S QRSS/WSPR transmitter. I’ve built both the Ultimate 3S and QCX rigs, and liked the class E PA’s he used in both designs. Simple in design – and I also like the fact that, because the BS170 is a MOSFET that doesn’t suffer from thermal runaway, you can simply parallel them up for greater power, without the need for balancing. Details of how to wind the bifilar transformer can be found in the assembly manual for the Ultimate 3S on the QRP Labs website.

In his beacon, Paul runs the Si5351 at it’s default of 2mW output, and follows it with a W7ZOI-designed 2 stage PA from the pages of EMRFD . Due, I suppose, to sheer laziness, I wanted to keep the PA stage as simple as possible, so opted for higher output from the Si5351, and a single MOSFET, with very few supporting components, for the PA. Paul mentioned that in the earlier days of the Si5351 being available to experimenters, he heard some talk of higher phase noise and jitter from the Si5351 at higher output levels. Perhaps running it at a lower output level, and making up for that later, is a worthy strategy? To run the Si5351 at it’s maximum power of about 10mW out into 50 ohms, I found the following line in Paul’s modified code –

si5351.init(SI5351_CRYSTAL_LOAD_8PF, 0, 0);

and inserted the following line after it –

si5351.drive_strength(SI5351_CLK0, SI5351_DRIVE_8MA);

This sets the chip to produce the maximum power at the CLK0 output.

The very first iteration of this project used a passive patch antenna, as I didn’t realize that the GPS module supported active antennas. The patch antenna, with it’s very short piece of coax, was quite difficult to implement in the diecast enclosure I had chosen for the project. I mounted it on top of the lid, with the main board mounted on the inside of the lid, and the coax passing through a hole in the top. When I took the lid off to work on the circuit, the antenna was shielded from satellites by the lid, which was inconvenient. Once I discovered that the GPS module supported active antennas, I installed one. I have no photos of the implementation with the passive antenna.

Here’s a view of the next version of the board, with the clock generator and Nano boards unplugged, to allow viewing of the wiring underneath. As usual, I have used Rex’s wonderful MePADS and MeSQUARES for the Manhattan pads, and strips of header to plug the Si5351 board, Nano, and LPF boards into. Operating on a different band just requires changing the output filter, and reprogramming the Nano via it’s ICSP header –

The first version of this build used a single 7805 voltage regulator, bolted straight onto the board for heatsinking. I had forgotten how very hot these 1 amp regulators get. The IC itself got very hot, as did much of the ground plane on the board to which it was bolted. Although not my best idea, it turned out to be dwarfed by a particularly poorly thought-out aspect of the layout –

It’s perhaps not immediately obvious from the above photo, but might become more apparent from this image –

That is the BS170 PA transistor mounted directly underneath the frequency synthesizer board. The problem, is that the PA transistor gets very warm. Warm air rises – and what is directly above? Yes indeed – the most frequency sensitive part of the whole circuit. What a fool, an oaf, a bumpkin, a buffoon, and a rube! When laying out the build, I was mainly concerned with fitting everything in, and not having a long wire between the output of the Si5351 and the PA. I’m not sure why, as a short length of RG-174 would have worked just fine. Nevertheless, slightly disheartened at my mistake, I forged on, and proceeded to attempt to calibrate the unit using Jason NT7S’ calibration script. I’ll spare you the long, dull version, and just say that I couldn’t get Jason’s script to work. My suspicions lay with either the cheap Nano board, or the cheap Si5351 board that I had bought from Amazon. Not pictured here, the first Si5351 board I tried was a direct clone of the Adafruit board, with a purple board instead of the Adafruit blue color, and without the Adafruit branding on it. I ditched Jason’s script, and went for a rough calibration by beating the output of the board against WWV, and making adjustments to the correction factor, until I was within a Hz or two of zero-beat.

I then uploaded VK3HN’s script to the Nano. The unit was indeed WSPR’ing but, despite the fact that I had calibrated it fairly accurately, quite a few of the WSPR transmissions were out of band by anything up to 100Hz. This didn’t seem right, so I tried calibrating the board again, only to find that each time I calibrated the board, I came up with a significantly different correction factor. Replacing it with a genuine Adafruit board solved the problems. Suddenly, Jason’s calibration routine worked beautifully, and the board began producing consistent, repeatable frequencies. All subsequent WSPR transmissions were in-band. The Si5351 board that I had purchased from HiLetgo was only about $3 less than the genuine article from Adafruit. In retrospect, it was not worth the trouble just to save a few bucks. Lesson learned. In contrast, their Nano boards are significantly cheaper than the “real” thing, and seem to work just fine.

My first foray into WSPR with this mini concoction was on 10M. Drift figures were nearly all -4’s, and I wasn’t getting anywhere near as many spots as I would have expected to get. Because nearly all the drift figures were -4’s, that indicated to me that many spots were very possibly being missed, due to a drift figure of higher than -4. Placing the board in a diecast enclosure with the top on helped. I was then getting more spots, but still all with drift figures of -3 and -4, with more -4’s than -3’s. I went down to 20M, where the drift figures were a little better, but still not good enough. From cold, the first few transmissions produced no spots. After an initial warm-up period of about 30 minutes, I was getting more -3’s, and even a few -2’s. Still not good enough.

One obvious change would be to relocate the PA to the opposite side of the board, away from the clock generator board. If I did that, it would be in another build completely so, for the time being, I concentrated on other ways to bring the drift down. Here’s what I did –

1. Mounted the 7805 regulator on the side of the diecast enclosure, to which it was bolted. I also added a 7808 regulator, thinking that it wouldn’t hurt to spread the heat generation between two devices, even though these parts are designed to run very warm.
   
2. Added a 1N4001 diode in series with the 12V DC input. As well as providing reverse polarity protection, the forward voltage drop of about 0.7V should help to spread the heat dissipation from the regulators out just a little more.
   
3. Secured the clock generator board with 2 nylon screws and a threaded nylon spacer. I had been waiting for the parts to arrive from Adafruit, so hadn’t done this earlier. (This kit of nylon screws and spacers should last a while!)
   
4. Although not a modification, one thing I did differently this time before testing out the beacon, was to screw down the top of the enclosure tightly, instead of just placing the top on.
   
   

After these changes, the difference was dramatic. No spots were picked up on the first transmission. On the second transmission (on a 50% transmission duty cycle), several spots were received, all with drifts of -4. Things improved with every cycle until, after about 45 minutes, all spots were -1’s and 0’s, with the very occasional -2. Much better, and very encouraging.

Some more shots of the board with the regulator removed, and replaced with 2 regulators in series (a 7808 and 7805), both bolted to the side of the enclosure. The enclosure is a bigger mass of metal that provides more effective dissipation of heat from the devices –

Here’s a view of the board with the Si5351 breakout board and Nano board unplugged, to show the wiring underneath –

Looking dead sexy in it’s diecast enclosure from Tayda –

In attempt to further improve the drift figures, I made a heatsink from a piece of brass strip, and epoxied the BS170 PA transistor to it with JB Weld. A clamp held the mighty little MOSFET in place while the epoxy set –

A pair of round-nose pliers were used to bend the leads. The leads on some of these parts are quite delicate, so I prefer to coax them round the bend, rather than foisting an abrupt 90 degree angle on them –

I am unsure of the dielectric properties of JB Weld so, to avoid any problems, made sure to keep the area around the leads free of epoxy –

I think this heatsink looks mighty spiffy. Brass is such an attractive alloy –

A close-up of the heatsink –

Unfortunately, with the heatsink fitted, the drift figures were worse. After about a 90 minute warm-up period, I was getting drift figures of mainly -2’s and -1’s. Removing the heatsink got me back to drift figures of mainly -1’s and 0’s, with the occasional 2, after warm-up. After 2 hours, the drift figures are equally split between -1’s and 0’s. All of the figures I have quoted are from 20M operation, by the way. A quite satisfactory result, I think, from a frequency generator board that is not temperature compensated.

I was already fairly satisfied with this result, but then things became better. The heat from the PA transistor was rising, and heating up the Si5351 board, forming a sort of crystal oven. Because of this, it occurred to me that if I were to adjust the bias on that BS170, it would affect the amount of heat the transistor gave off, and might also affect the drift figures reported by wsprnet.org. The transistor was currently providing about 200mW to the antenna. Although, by adjusting the bias, I could have coaxed some more power out of it – perhaps as much as 250mW, I didn’t want the transistor to run much hotter than it was already running. Likewise, I didn’t really want to run much less than 200mW. Fiddling around with the bias trimpot, I ended up with it in almost the same place as it was before. The transistor was probably putting out a mW or two more, but not much more. However, the difference in the drift was dramatic. Check out this wonderful result (still on 20M), achieved after a warm-up period of around 40 minutes or so –

These fantastic drift figures almost made me giddy! The only other thing I had changed, was to swap out the 4 oxide black panhead 4-40 machine screws on the sides of the diecast enclosure, for regular stainless steel machine screws. Perhaps they have slightly different thermal properties, but I think the main factor responsible for the improvement in drift, was the very slight change in the bias setting. I had haphazardly settled on a near-perfect bias setting, and created a very effective crystal oven! I did have a couple of other ideas I was going to try, namely placing foam over the Si5351 board, to insulate the Si5351 and crystal from air currents, and looking for a TCXO to replace the crystal in the Adafruit board. However, at this point, I don’t think it’s necessary. Running the beacon for another 6 hours, the results were much the same, though a single -2 and +4 drift figure did pop up. I think the +4 was an anomaly, probably caused by drift in the other station. This is a better result than I had hoped for. I’m ecstatic!

On 10M, it takes 3 hours to fully settle down, after which, drift figures are mainly -2’s and -1’s, with a few -3’s, the occasional 0, and the very occasional -4. However, I do notice that after running it all night, drift figures in the morning are a little worse, with a lot of -3’s and a few more 4’s. This suggests to me that the ambient temperature of the room might be playing a part.

Incidentally, changing bands only involves changing the band in the code, which requires simple changes to two lines in the sketch, plugging in a different LPF, and uploading the new sketch via the ICSP header on the Nano board in the WSPR beacon. As far as initial setup goes, before you upload Paul’s modified code, you will need to insert your callsign, and the power level in dBM. Mine puts out about 200mW, which is 23dBm. You can input your grid locator if you want, but the unit will calculate that from the GPS, once it has gotten a fix. Although I haven’t tested it yet, I assume that if the unit moves into a different grid square, it will report the new locator. (EDIT – Paul informs me that, although it would be easily possible to insert code that calculates the grid locator, his modified code doesn’t do that. I assumed it did, based on the fact that although I input my locator as CM87, wsprnet reports it as CM87ut. However, they are probably doing that based on their knowledge of my location. Looks as if I have something else to work on!)

By the way, when you’re changing bands, remember to also change the LPF. When assembling the LPF boards from QRP Labs, I always check the response curve on my NanoVNA. As an added testament to the fact that they do indeed work, I recently flashed the unit with firmware to change the operation from 80M to 20M and left it to run overnight. In the morning, there had been absolutely no spots. I was flummoxed, and even thought I might have fried the Nano board, until it dawned on me that I had not changed the LPF. The beacon was running on 20M, with an 80M LPF still plugged in. No wonder!

In the future, I may experiment with an Si5351 board that has a TCXO, in order to improve the drift figures on the higher HF bands. In the meantime though, I am deliriously happy with the performance on 20M (and presumably below). This project was inspired by VK3HN’s SPRAT article, and the realization that “throwing together” a few boards, and constructing a simple PA and LPF should be easy, and wouldn’t constitute a full-blown project. I have become somewhat shy of such lengthy endeavors these days. I wasn’t expecting it to turn into a cased-up and very serviceable WSPR beacon though. I tend to let it run in the evenings and overnight, when I’m not operating. That way, in the morning, I can check wsprnet to get an idea of what propagation is like. As many others have said, it’s a handy propagation tool. If you don’t want to build one, you can buy a ready-made WSPR beacon from Harry at Zachtek.

At the risk of posting too many pictures, here are a few more –

A cigar box from the local tobacconist, and some packing foam, makes a good storage box for my growing collection of QRP-Labs LPF’s and BPF’s. Only the LPF’s are used in this project. The BPF’s, in the front row, are for receivers (though they could be used in the early stages of transmitters, where only very low signal levels are involved) –

Definitely a successful project. Thank you Harry Zachrisson of ZachTek, Paul VK3HN, and Jason NT7S –

One last gloat. Look at these great drift figures. Pretty good for an Si5351 board without a TCXO! To date, this 200mW powerhouse has been spotted all over Europe, North, South, and Central America, Hawaii, Taiwan, Hong Kong, Australia, New Zealand, and several island clusters and nations in the middle of vast oceans. Exciting stuff!

Oh, and one last thing. Paul included an LCD display in his transmitter, which shows some extra useful information. The code will support it, and his blog shows how to connect the display. I think there is just enough space to fit a display into my unit. I didn’t chance it however, as I seem to have the thermal balance inside the case just right (for 20M and below, at least) and I don’t want to upset anything. My desire for a display isn’t strong enough to want to make any more changes. I’m fine with this, as I think of it as a set and forget kind of beacon. In the evenings, I plug it in, and forget about it until morning.

I haven’t been building much at all, for quite a long time now. However, the urge occasionally returns. When it does, it’s wonderful to have a small stash of parts on hand, so I can pull the soldering iron out and start building before the desire dissipates. I’ve been interested in beacons for a while, and this interest has followed a logical progression. I first noticed that my interest in talking with other hams over the air using phone (i.e. SSB) was waning. During this time, I would still check in daily with the Noontime Net on 40M (now on 7284 KHz). A quick check-in to a net was fine, and it was good to hear the other stations, many of whom were regulars, also check-in, and have the occasional quick conversation. In addition I noticed that my inclination towards CW QSO’s was also diminishing. I’m not exactly sure why. I enjoy talking with close friends, acquaintances, and neighbors in person, but my enthusiasm for chatting with relative strangers who I can’t see, and don’t know that well, all but disappeared. Too many QSO’s seem very cookie cuttter. Either that, or the conversation is about subjects that don’t interest me.

My main interest in amateur radio was fast becoming the medium of radio communication and the science behind it, rather than the content. I enjoyed quick contest exchanges, as it was a way of seeing where my signal was getting to. Some folk dislike contests, preferring to ragchew, but I find the average ragchew on the bands rather dull. Mike Rainey AA1TJ referred to brief QSO’s as being akin to high-fiving someone when you’re walking down the street. It’s an acknowledgement – a quick, “Hi! It’s good to see you. Talk later!” If you just built something, it’s good to know that it works, and a way to marvel in the mechanism of propagation that made it all possible.

This is why I like beacons. The ones I listen to send CW at regular speeds. I tend not to look for QRSS beacons, or data signals. I like CW that I can listen to and decode in my head, even if it’s just a few letters that are constantly repeated. I can put in the work to decode a very weak signal from a QRP (and often QRPp) beacon, figure out roughly where it is in some cases, and feel the satisfaction of having received a very weak signal, without listening to some chap talking about the model number of his transceiver, his latest medication, his political/social opinions, or whether he mowed the lawn and watered his plants today.

I’ve probably explained this before, but there are two main types of beacon I like to listen out for. The first are the so-called “unlicensed HF beacons” (for which read pirate). They tend to operate in the lower half of the HF spectrum. The lowest frequency one I know of is on 2097.3 KHz, is located somewhere in the southwest (many of them are in the SW deserts), and is relatively high power – probably in the range of 5-15W. It sends the letter A once every 10 seconds (approximately). There are a number clustered around 4096 KHz (a popular crystal frequency), quite a few around 8000 KHz, and others up to about 8500 KHz. A good one to listen out for, is the fairly new Desert Whooper, on 4095.65 KHz. It must be relatively high power, as I have heard it regularly in 6 Western states, on a recent campervan trip, with a portable receiver and set-top whip. It sends a whooping sound for a few minutes then, in CW, it sends the battery voltage, the outside temperature, the inside temperature, and a number that is related to the solar panel voltage. Then, if I remember correctly, it sends it’s ID a few times (DW, for Desert Whooper), and goes back to whooping for the next few minutes. The current list of known active unlicensed HF beacons is here, on the very excellent HF Underground forums –

https://www.hfunderground.com/board/index.php/topic,9478.0.html

I’m a bit too chicken to deploy one of these types of beacons, partly because I’m a licensed amateur. Also, how do you really know that your HF beacon is not operating on, or close to, some frequency that is used for critical communications? Certainly, there are published bandplans, but I’m not sure how much detail they really go into. On top of that, I just cannot bring myself to leave something that emits RF in a remote location such that I would have trouble getting back there quickly, were the beacon to malfunction. It’s a control thing, I suppose. I’m responsible for my stuff, and I like to be able to switch it on and off, and service it, at will. I know that the likelihood of a 100mW HF beacon interfering with critical communications is pretty unlikely but even so. Besides, as an amateur, there is a wide swath of spectrum I am allowed to use legally (though not for unmanned beacons below 28MHz).

The other type of beacons that I find quite fascinating to listen for, are the HiFERs. These are also unlicensed, though if you keep your power low enough, they are legal (in the US, at least). The FCC, for the 13553 – 13567 KHz ISM band, regulate the power level allowed in terms of the field strength a certain distance from the antenna. For those of us without a professional calibrated field strength meter (most of us, because they are expensive!) a couple of helpful gentlemen have expressed this in terms of power to various antennas. According to W1TAG’s calculations, you should be within, or close to, the allowed field strength with 4.6mW into a half-wave dipole, or 2.3mW into a quarter-wave groundplane. K6STI uses several different antenna installations over varying degrees of ground conductivity. Under certain conditions, he calculates that up to 473mW would be allowed. His calculations are at –

http://ham-radio.com/k6sti/hifer.htm

I’ll leave it up to you to decide what power level to use, but it sounds to me as if you’re pretty safe if you keep it under about 5mW. What? you think. Is it possible to even hear beacons at distance at those sorts of power levels? It certainly is. If you, like me, either have a poor antenna installation, or none at all, you can still hear some of these HiFERs by utilizing online SDR’s. One of my favorites is the KFS SDR, that uses several large antennas on a 150 foot cliff overlooking the Pacific Ocean, 6 miles south of Half Moon Bay, CA. This SDR has good ears! In the last few days, on this SDR, I have heard TON in Tonopah, Arizona, PCO in Pine, Colorado, and TSN in Tucson, Arizona.

Which brings me to this project. I wanted to build a HiFER beacon that would operate from a LiPoly battery charged by a small solar panel, for a compact installation that could be located outside on a hilltop, or in some desert area. These batteries are 4.2V at full charge, so I built up my beacon with a LP2950 5V regulator on board, figuring that if it worked at 5V, it would probably work at 4.2V. The regulator and it’s bypass capacitors are not shown in the following schematic, but there is a 100uF and a 0.047uF bypass cap on the input of the regulator. The bypassing on the output is shown in the schematic, and consists of a 0.047uF on pin 8 of the ATTiny85, and another 0.047uF on the +ve supply line side of the 8.2uH choke that feeds the collector of the PA transistor. A molded choke was used in this position.

13.56MHz crystals are easily available from many suppliers of electronic parts. I bought a bag of 20 on eBay. The RF chain is very straightforward, and is the same one seen in the transmit section of the Pixie transceiver. The output filtering is more than is necessary for this application. 2 poles of filtering would have sufficed, but it doesn’t hurt to add an extra pole. The 47pF capacitor across the center toroid forms a parallel tuned circuit with that inductor to increase rejection at the second harmonic. It’s a design by W3NQN. I took the values from the 20M LPF –

https://www.gqrp.com/technical2.htm

The ATTiny85 keys the oscillator. There is no chirp or instability caused by doing this. I like this approach, rather than keeping the oscillator running and keying the PA. I build most of these beacons for use around the house, and am usually in close proximity to them. If the oscillator is keyed, I don’t hear the constant backwave from the oscillator transistor. The code for keying, as in my Boris Beacon, was courtesy of Nick SV1DJG. I use pin 2 of the chip as the keying line, and had to change that in Nick’s sketch. The line –

#define ledPort 1

was changed to

#define ledPort 3

The feedback capacitors C1 and C2, as well as forming part of the feedback loop to maintain oscillation, help to pull the crystal, to determine the frequency of operation. Initially, C2 was 100pF, and there was no capacitor in the C1 position. The oscillator ran reliably, but it was on around 13561, which is too close to the center of the band where all the RF from the various ISM and RFID devices is. Take a look at 13560 ± 1 or 2 KHz on a waterfall display – on your own receiver if you have a good antenna, or a good online SDR. You’ll see why you wouldn’t want to run a QRPp beacon there. Too much RFI! However, above about 13562, and below around 13559.5, the band is quite clear. It’s a gift for very low power beacon enthusiasts. Where else in the HF spectrum can we legally operate an unmanned beacon that stands a chance of being heard? (Well apart from 10 meters, that is.)

I wanted to avoid frequencies that established and known beacons are operating on, as well as the 3 other devices in this band that I use at my house. I ended up with a 100pF cap for C2, and a 150pF one for C1, which gave me an operating frequency of 13557.49 KHz ± a few tens of Hz, depending on the ambient temperature.

With the above setup powered by the 5V regulator, I measured 0.913V across the 51 ohm resistor with the N5ESE RF probe and an EEVBlog Brymen BM235 DMM. The voltage drop across the diode in the probe was 0.234V, so this translates to a power of about 25mW. I removed the 51 ohm resistor and measured the power with a freshly calibrated WM-2 QRP Wattmeter. The reading was about 6 or 7mW. I am not sure why the discrepancy between the two readings, but if I’m able to get my hands on a DSO in the near future, I’ll be interested to see which of the two readings is the more accurate. I suspect it’s the wattmeter reading. If the power is closer to 25mW, it can easily be dropped with an attenuating pad, or a lower supply voltage from either a LiPoly battery, or a 3.3V regulator.

A successful project, I think. I may even press it into service one day!

In this post from May of last year, I detailed the construction of a 1mW solar-powered HiFER beacon. I named it the Boris Beacon, in tribute to my neighbor’s cat. The beacon was never mounted permanently outside. I kept it indoors, powered from a small solar panel in the window, and feeding an “antenna” of sorts, consisting of the original dipole wires folded up into two small bundles. Obviously, I had no serious intention of it being heard by anyone; I just liked having it come on every day when the sun came up, and transmitting until later in the day, when the light was too low to sustain operation.

Recently, another location became available in my house that seemed like a good place to install a beacon outside. The Boris Beacon was still in operation from inside my apartment. Moving it outside onto this first floor balcony and spreading the dipole legs would be a straightforward task. You’ll notice from the original post on this beacon that, in attempting to seal the holes where the leads entered the enclosure, I used Plastidip. It’s a rubbery solution that sprays on. It’s great for some applications, but not for this one, as I ended up getting the rubbery liquid all over the enclosure. I do like my projects not to look too messy, so for this new iteration of the Boris Beacon, I moved the circuit board into a new enclosure –

Here it is, close to it’s final installed position, on a first floor balcony (Edit – I just noticed, after a year, that I should have called it a second floor balcony. In the UK, where I haven’t lived since I was in my early 20’s, we call the second floor the first floor, and the first floor the ground floor. I guess old habits die hard!) –

In it’s final installed position. The solar panel is fixed to the top of the wooden railing with 2 wood screws, as is the beacon enclosure. The dipole is stretched out behind the wooden  fence at the top, and then trails down onto the balcony floor in one direction. In the other direction, it is attached at the other end to the wall of the house, so is partially elevated –

A close-up view, showing the silicone caulk around the entrance/exit holes. The underside of the lid has a foam weather sealing strip embedded in it, which can be seen in the original post, linked to at the beginning of this post –

I was unsure how impervious the little solar panel would be to the elements, so I caulked around the edges. If it fails, these kinds of low wattage panels are cheap and easily available anyway –

The panel I’m using is a small 1.8W one, intended for use as a 12V battery maintainer –

It is probably overkill, but I popped a silica gel packet in the enclosure, to mop up any excess humidity that might find it’s way inside. The dessicant turns pink when saturated, and is blue when dehydrated and ready for action –

Another view, with the gel packet flipped –

The beacon sends the letters “BRS” at 10wpm, with a break of 3 or 4 seconds between the end of one transmission and the beginning of the next, with a mighty power to the dipole of about 1mW. The frequency is a nominal 13556.9KHz (13.5569MHz), which varies either way by a few tens of Hz, depending on the outside ambient temperature. I will be overjoyed if anyone, anywhere hears it! There is no battery, so it transmits during daylight hours only. It comes on about half an hour after local sunrise, and goes off about half an hour before local sunset. I’ll update this with more accurate information, as I observe the on and off times over the next few days.

The Boris Beacon is definitely a successful project. I just need someone to hear it. Even one person will do! If we were allowed to run 100mW on this band then getting spots would be much easier. In fact, if the dipole were situated more up and in the clear, that would help too. As it is, 1mW into a compromise dipole will make this little beacon a super DX catch. I don’t know how long it will remain in operation, as the long-term future of my current living situation is in doubt. I suspect that it will be up and running for much of 2019 though. I will update this page if and when it goes off the air.

Reception reports greatly appreciated!

EDIT –

Almost a week later, and it seems to be faring well in the rain, although it’s early days –

Rain was pooling on top of the panel and although it’s supposed to be weatherproof, I’m not too sure how waterproof this panel really is –

I raised one end slightly, to help a bit –

Still no reports!

EDIT – As of Aug 2019, the BRS beacon is off the air, probably permanently. The space from which I was operating it from is no longer available. It was put to sleep, having received not one report. I put it down to two things. Firstly, it was active during a period of particularly poor HF propagation. Secondly, the power was around 1mW. Even so, I was hoping for at least one report. I think it would have been worthwhile to have reprogrammed the chip to send QRSS.

Back in May of this year, Sheldon N6JJA began sending me information and details of his version of N1BYT’s WBR regenerative receiver. It went through several iterations, before ending up at the final version as shown here. Even this version is still a work in progress – as all good products of experimentation are. Sheldon took the original WBR circuit, as described by N1BYT, and made a few changes. Firstly, he added a preselector. Regens are well-known for having poor strong signal performance. A pre-selector can’t help with very strong signals close to the received frequency, but it may well help to protect the fragile front end of a regen from very strong signals away from the frequency the regen is tuned to. Secondly, a tunable preselector is a very handy tool for anyone who is experimenting with simple receivers, whether direct conversion, regenerative, or simple superhets. It could be well worth building this as a separate unit, for future use and experimentation. Of great interest also, are the changes that Sheldon made to the classic WBR circuit. He describes them in some detail in this article. It makes for a good read, and may encourage you to try building it for yourself. If you do, please let me know how it goes, as I haven’t built this version of the circuit yet.

Sheldon, amazingly, found enough time away from his very busy silicon valley job, and responsible position as a first-rate cat dad, to write up this project as a complete article. Rather than attempt to interpret his words and re-write them in my own style, I’d rather that you get all this great information straight from the cat-dad’s mouth so, without further ado –

(Oh, one more thing – check out the way he draws his schematics. Accurate and beautiful!)

 

How it started (by Sheldon N6JJA)

In 1957 I was in second grade in a small town in rural Illinois. Our “library” was a bookmobile that came through once every two months. But even that long ago I was completely in love with all things radio and electronic, so when Alfred P. Morgan’s book The Boy’s First Book of Radio and Electronics appeared I checked it out and devoured it over the next two months. For me, the centerpiece of the book was a design for a simple, one-tube regenerative receiver. My desire to build such a thing knew no bounds, but a lack of money and parts made it a non-starter at that time. I became a ham in 1965 and my attention turned to more modern equipment and kits to build things, but I never got over that old regen. Fast forward eight more years and I’d gotten my hands on my first “boat anchor,” an SP-600-JX7. Alas, I was once again in love, but the radio wasn’t mine to keep, and I again resolved, “Someday…”

Oddly enough, that’s all part of how and why the WBR-Oscar came to be. Over the past few years I’ve been buying and restoring a variety of “boat anchors,” and now have a lot more than I can keep. And that radio-crazy second-grader wound up with a Ph.D. in electrical engineering and a huge junk box and some decent test equipment. So, somewhat naturally, my first major project last year was to build something to help those old radios perform with a little more pizazz. It started as a wide-range antenna coupler, then added a preamplifier that became also a preselector, then added an audio amplifier, then a DSP filter…well, you get the idea. One of our cats, Oscar, helped me with all of this, making sure that now parts got included that hadn’t been checked for obedience to the laws of physics, particularly gravity. As you can see in Figure 1 below, Oscar became the name of that “helper” unit. Oscar sits proudly atop his namesake.

 

I considered, for my next project, that one-tube regenerative receiver. But then, along the way, I came across an article from QST from August 2001 for a regen that I’d started to build but never finished. (Life gets in the way sometimes.) The article was by Dan Wissell, N1BYT, and titled “The WBR Receiver.” (Citation at the end of the article.) Rummaging through my collection of parts I decided that this was the right project at the right time. Mr. Morgan’s radio would wait a bit longer, but my desire to build a regen was finally going to be fulfilled.

Over the years a lot of these receivers have been built, and the results have been mostly good, it seems, even spawning a lot of designs based on the WBR detectors but without the WBR tank circuit. Lately, thanks to the ongoing blog of Dave Richards, AA7EE, interest in this design has been renewed and some of the design’s deficiencies noted and, to a certain extent, addressed by experimenters. Now, however, buoyed up by the information in Dave Richards’ blog and some other QEX articles, I decided that it was time to put my own spin on things and see how far I could push the design. The result is the WBR-O receiver, and it now covers, fairly easily, 6 to 15.6 MHz with a single tank circuit, making it now a true “40-30-20” meter amateur and SWL radio. As an additional feature, the “O” (for Oscar) part refers to a preselector/preamplifier that both isolates the input of the WBR circuit and adds front-end selectivity and some amplification. The preselector/preamplifier actually tunes from 80 to 10 meters (in two bands), includes widely adjustable gain, and can easily be built as a stand-alone for folks who only want that part. The photo in Figure 2 shows the pair as built.

The design you see in the photo is actually the culmination of months of work and numerous revisions and tweaks. Some things helped, some didn’t, but I learned a great deal along the way. I also, on purpose, used several “lousy” construction techniques to convince myself that, simply, “If I can build this, anyone can build this.”

This article is broken into two parts. First comes the “Oscar” preselector/preamplifier. As I said, I intended it to be either part of the overall receiver or used as a standalone where desired. The second part deals with the WBR upgrades. Both designs were built using the same techniques and I’ve tested both and found that—especially in concert—they do about as well as some of my boat anchors! So if your soldering iron is ready, I’ll start by describing “Oscar.”

Part 1: Oscar’s Preselector/Preamplifier

Figure 3 (above) shows the circuit during testing on my bench. To the left I placed the coils, clustered around a small relay (Panasonic TQ2-12V). To the right is the amplifier portion of this circuit.

Figure 4 (below) shows the schematic.

Now before you say to yourself that it looks pretty complicated, let me point out how simple it really is. Each “band” has its own bandpass filter that it tuned by one or more Toshiba 1SV149 varicap diodes. The 1SV149 is a little gem that was developed for AM radios but is now obsolete. In spite of this, that diode is plentiful and inexpensive on the internet (I got mine from eBay, about 50 for $10, but Amazon sells them, too, as does Minikits.com.au.) An important item to note here is the value of the series isolation capacitors for the diodes, C1 and C3 in Fig. 4. I use 0.1 μF, 50V ceramics with an X7R stability rating. The large capacitance is actually a must; as the value goes down, the interaction between those caps and the rest of the circuit becomes a problem. The relay is something that I had in a drawer, and made the layout a bit easier. I’ve also built “Oscars” using rotary or toggle switches to switch bands, by the way, so while the relay is nice to have, it isn’t required. One thing to note, however, is that, with the relay, the overall design lends itself quite nicely to a remotely-tuned preselector that can be mounted right at your antenna. So far, nearly everything I’ve built along these lines has worked. The amplifier in Fig. 4 is based on a low-noise MMIC pair in a push-pull arrangement that keeps distortion and unwanted harmonics down a bit. The MMICs are Mini-Circuit Labs MAR-6SM+ devices. At a maximum of 16 mA per device they offer gain of about 20 dB and noise figure around 2 dB. Quite impressive for so simple a device. The 1:1 transformers are also from Mini-Circuits, their T1-1+. All of these components offer good technology for a relatively low price. The relay (from Digi-Key, for example) is $3.88 in small quantities. The MAR-6SM+ is $1.40 each, but the minimum is 20, so either consider a lifetime buy for $28 or split the batch with a friend. The T1-1+ is $3.25 in small quantities. If you’re up to it, go ahead and wind your own transformers on small type 43 ferrite toroids. A simple bifilar winding should work, and there are usually design guidelines in articles on baluns and transformers to help you decide on a target inductance. Actually, I found that building my first “Oscar” made me make sure I had enough parts—including transformers—to build more of them. The circuit has become somewhat my “go-to” front end for things. I should add that I’ve also built the amplifier section with a single MMIC and without the push-pull transformers, and it still works okay, if you want to minimize or simplify things. The circuit then looks more like the C5-U1-C6-R6 cluster in Fig. 4.

 

Table 1 (below) gives my winding data for the toroids, including those for the WBR receiver part. I like the website toroids.info to help me design the coils, then I use another best friend—a Peak LCR45 meter to verify results. I find that, even knowing that each time wire passes through a toroid counts as a “turn,” I wind up removing a turn or so once I measure things. Frankly, this whole project would be a lot more difficult without the LCR meter, and once you use something like this I suspect you’ll be hooked as well. The bandpass filters don’t do nearly as well if the coils aren’t either the correct values or reasonably well matched.

Good thing to remember: the MMICs and diodes are sensitive to ESD. Not horribly sensitive, but you will want to be careful, since it will save you headaches later on. Sometimes working on some aluminum foil or an inexpensive ESD mat is plenty, also making sure you touch the foil or mat before handling a part. Overall, in my own experience I’ve found these parts to be pretty robust.

Depending on the available PCB space you have, you might want to experiment with the general layout on perforated PCB material and then make a sketch or photo of your final design to guide you in construction. Part of the fun of this project is that we all tend to do things differently and put our own “fingerprint” in the final result. Remember, although I’ve done a lot of up-front work to guide you, what you build will be your own to be proud of.

My own usual RF build technique is to use 0.10” center perforated PCB material with plated through holes and cover one side with copper foil tape to make a ground plane. The boards I used in this build are 7 cm x 9 cm, and as you can see, I have plenty of room. (I should also add that the last revision of the WBR was done on a board about half the area.) Then I use an Exacto knife (or something like that) to carefully cut away portions of the ground plane where necessary. Figures 5 and 6 show this in more detail. Then I use tweezers and solid wire-wrap wire (28 or 30 gauge, stripped first as needed) for the interconnections. I solder components to the PCB and leave about ¼” of lead projecting on the underside of the board to provide for the interconnections. These are made using the tweezers (and maybe a magnifier) to wrap wire on one lead, anchor it with solder, and next do the same at the other end. (I call this “Compact Wiring.”) This keeps the wiring neat and compact and still provides an excellent RF ground. However, to be honest I’ve also used “dead bug” and even less glorious methods of construction (even using longer wires), and just about everything works as long as you’re careful.

Mounting the MMICs is the only thing where I needed to think hard about “how to do it.” In Fig. 6 you can see the MMICs, mounted on the ground-plane side (the “underside”) of the board, but mounted “upside down” with respect to the photo. I drilled a hole in the board for each MMIC, just large enough for it to rest in the hole, allowing the protruding pins to just touch the adjacent plated through holes and allow me to make good electrical connections.

Turning back to the schematic, you’ll notice that I use small voltage regulators to keep things stable and quiet. The LM317LZ is an inexpensive part (about 40 cents apiece from Digi-Key) that can handle up to 100 mA of load. U3 is used to provide the tuning voltage required, and U4 controls the gain of the MMIC amplifier by varying the voltage presented to the current-limiting resistors feeding the MMICs. I’ve taken to using these ICs in virtually every project. They simplify the design work and are very flexible and stable. Add to that, they offer some low-pass filtering effects that can reduce the hash from cheap power supplies. No miracles, but good engineering.

I recommend coaxial cable input and output for this circuit. I bought a bunch of PCB-mounted SMA connectors a while back, and you see them in Fig. 3. But there are so many different ways of making these connections—especially if you choose to do the switching via a panel-mounted switch—that no builder should be intimidated by the technology.

I used a 10-turn pot for the tuning and a single-turn pot for the gain. Both are linear in design and do not need to have a wattage rating over 1/4W. But a word about choosing potentiometers. I’ve used cheap ones from the Pacific Rim and higher quality ones from US suppliers. In my experience, especially in the tuning pots you tend to get what you pay for, although the cheaper ones can be used if you are careful how hot they get during soldering. It seems that the inner workings of the less expensive ones are more susceptible to heat and can “quit tuning” if you do a lot of “cut-and-try.” If you do buy the cheaper ones, buy several.

Results

Figure 7 shows my measured results, using a Rigol DSA815-TG spectrum analyzer and subtracting the tracking generator output level. That means that the circuitry was tested with a good, solid 50 ohms input and output, but the circuit still works well in my shack with antennas not very well matched. The bandpass curves were taken with a “gain” setting of about 10 dB. As you can see, the curves are sharper on the low end of each band, but there’s also more attenuation there, about 4 dB less on the low band, and about 5 dB on the high band. The sidebar on “Designing your own Oscar bandpass filters” talks a bit about this, but once you start designing for Q values above about 10 these effects are normal. But a preselector that maintains Q over 10 over this range should produce a workable unit that provides some selectivity without having to constantly readjust things every time you change frequency by a couple of kHz.

 

 

 

 

 

 

At the bottom of Fig. 7 is the gain range curve at 14 MHz. With the design as I’ve built it, I can bias the MMICs to be “just barely there” and provide some signal reduction or to blast things with an additional 15 dB or more.

Now I chose to do this in 2 bands, but you may decide on 1 or 3 or anything else. The MMICs don’t care, as long as they think they’re seeing roughly 50 ohms at input and output. But be aware that the MMICs also will amplify anything from DC to 6 GHz! Using them without some form of bandpass in front will suck the amplification headroom from where you want it to someplace you don’t. And if you decide to make things simpler, using a single MMIC is a straight-through configuration works very well also. I’ve tried a lot of different configurations, so these are just my guidelines based on experience.

About all that remains is to hook this circuit up to a receiver and see what happens. Don’t be afraid to experiment. The fun of building something like this is that there’s a lot of room for changes, improvements, and growth. For instance, the gain section in the Oscar design is custom-made for a homebrew AGC circuit, and maybe someday I’ll give that whirl. (If you don’t beat me to it!) The next part of this article will expand on what I’ve written here to build the companion WBR receiver, but I’ll also refer back to this design to cover some of the construction guidelines I’ve mentioned here.

Part 2: The Upgraded WBR Receiver Design

As I mentioned in the beginning of this article, both the Oscar preselector/preamplifier and the upgraded WBR receiver are designed to work together or be built as stand-alone units. Figure 2 shows both units assembled with standoffs between them right before integration into a chassis that has been the final home for all the previous design iterations.

So what’s changed? Well, if you look at the schematic for the RF deck in Figure 8, it might seem that not much has evolved since N1BYT’s article back in 2001ⁱ. However, a lot of tests and calculations led to some significant changes. But rather than just list the changes, let me take you on a brief excursion as to why they were made.

Online blogs and chat lines have talked about different variations on the WBR theme for a while. Often there are a few things that everyone seems to agree on, like the fact that AM sensitivity is very low and that the near-zero input impedance is needed to block strong signals, and that it’s difficult to get more bandwidth out of the design, even if mechanical tuning capacitors are used. Sometimes it seems to be nitpicking, but I decided to dive in and see how much pizzazz I could give the basic design.

First, about the sensitivity. The WBR isn’t a “normal” regenerative detector design, and this gets overlooked sometimes. It’s actually a regenerative Q-multiplier with an infinite impedance detector (IID). When the Q-multiplier is oscillating, the available signals to the IID are quite a bit stronger than when the Q-multiplier is set just below oscillation threshold, as in for AM reception. But look at the IID part of the circuit. It’s actually a source-follower and thus offers little chance for any amplification. IIDs have been around in tube circuits for years, favored by AM aficionados for their excellent low-distortion detection characteristics. So when you’re trying to listen to AM signals, you’re going to need a fairly good audio preamplifier. That’s Q3 in Fig. 8. I’ve looked at a bunch of WBR-like circuits online, and there are several good (and not-so-good) preamplifiers to consider. In the end, I designed my own. It has pretty good gain (about 120-150) and doesn’t use parts that aren’t easily available. I should mention that the value of C12 is important, though not critical. I use a rather large capacitor there, 100 μF, and that really helps keep the gain up over a nice audio range. Probably anything over, say, 47 μF should work.

Next, the “front end,” or lack of it. There are 2 points to make here. First, that’s why the Oscar preselector/preamplifier became part of my own design. Second, that 1” piece of wire in the original design by N1BYT helped “balance the Wheatstone Bridge,” but many builders have fiddled with that, adding a little inductance at that point. I tried that too, but finally got an idea that worked better. It’s a simple single-turn link immediately adjacent to the L1 center tap (which is connected to that 1 to 1.5 inch wire to ground). Feeding the 50-ohm sourced signal in through that tap turned out to be about a thousand times better than the “old” way. (See the addendum for technical details.)

I spent a lot of time working to get the design to work reliably over the roughly 10 MHz span that it finally achieved. The capacitance range of the Toshiba 1SV149 varicap diodes really did shine there, but early attempts to tune below about 8 MHz weren’t successful. In the end, though, what’s in the final design will probably tune below 5 MHz if I spent more time on it (and maybe even cover 80 to 30 meters or something like that). That’s where a lot of changes found their way into the design, but let me talk about them individually.

First, I was initially using garden variety (i.e., “flea market”) 2N3904 transistors for the Q-multiplier. Frustrated at the limited tuning range, I first increased the “gain” of the oscillator by increasing bias voltages from 5 volts to 12 volts. That helped, but not enough. In the end I found that the h_FE for that transistor (i.e., its DC gain) is vitally important. My weak oscillator used a 2N3904 with h_FE of about 85-90. I found a first-quality one with gain closer to 180. Aha! Much better performance. I got down to about 6.5 MHz at the low end.

But, to make a long story much shorter, I found an excellent, and even better, alternative. It’s the BC546CT from On Semiconductors. At about 10 cents apiece (about the same as the 2N3904) it offers the same basic qualities of the venerable 2N3904, but with an average (sample of 40) h_FE of 553! A batch of first-quality 2N3904s had an average h_FE of only 189. The pinout is the opposite of the 2N3904, but otherwise it was just a drop-in and now there’s no trouble with making a Q-multiplier that will oscillate easily just about anywhere, and I don’t have to measure h_FE endlessly to find a winner.

One experiment that also helped was to try different values of C2 and C3. You see, the feedback network in this oscillator (that’s the basis for the Q-multiplier) works fine for some values of C2 and C3, but as the frequency goes down, the losses in the feedback loop go up and can prevent oscillation. Dropping C2 and C3 to about a quarter (to 100 pF) of what I used initially (330 to 390 pF) did the trick. I’d recommend using small ceramic caps with an NPO or COG stability rating. Pushing this oscillator down further in frequency might entail re-tweaking those capacitor values, but that’s just part of the fun. One additional benefit from using the small capacitance is that the amount of regeneration bias required over the whole frequency span stays much more constant that when using the “older” values. Oh, and one last thing: in the original design the base of Q1 was biased using a more classic resistor pair, one carrying regeneration voltage, the other to ground. I removed the one connected to ground. What you see here works much better.

Now for my secret weapon. The little trimmer capacitor, C4, is unusually important. After staring at the original schematics for hours it occurred to me that the “balance” sought is almost impossible with the components used, no matter how carefully one measures things to force it to happen. Once I get a circuit oscillating, I tune to the low end of the frequency band and find where regeneration quits (you can hear it in headphones), even with maximum regeneration bias. Then I slowly tune C4 and make sure I have the headphone volume way down. Maximizing the oscillator strength with C4 is a set-once-and-forget adjustment, but it overcomes the last hurdles to giving this design its performance. Any trimmer that is small enough to fit and covers the 50-80 pF range should work. It just has to be near twice the value of C7 to be effective.

All That Bandwidth Presents a Problem

One thing became clear when operating this receiver. Even a 10-turn tuning pot made things dicey. I even found some Bourns “Digidial” counters on eBay and they helped, but not enough for this much bandwidth. I also played with several “bandspread” ideas using two tuning pots before deciding that instead of a bandspread control I could split the tuning range into, say, 6 pieces (I had a DP6T switch). This feature is included in Figure 9 along with the audio section built along with the RF deck. Again, I rely on the LM317LZ regulator, here in a paired arrangement, to set the ranges over which you can tune. Now signals are easier to separate, even with the smallish knobs on 10-turn counting dials. But taking this feature one step further (I encourage you to consider this), a builder might want to skip the “shortwave radio” aspect of the design and adjust the tuning ranges to cover only the 40, 30, and 20 meter bands, a very easy thing to do.

Table 2 lists the resistor choices I arrived at, both for the “shortwave” and “amateur-only” versions. I’d caution you to be prepared to build first and add resistors later. Your circuit might want resistance a little different from mine (and yes, small trimpots would make this a breeze).

 

Building It and Operating It

As you can see from the photos, the WBR section uses the same build technique (“Compact Wiring”) that the Oscar unit does. I encourage any builder to feel free to experiment with how the layout comes together and how the wiring gets done. What I’ve shown here is just my own way of doing things. Keeping the RF wiring reasonably short is a good goal, as is providing a good ground plane. Apart from that, all the versions I’ve built work (as long as I don’t forget a connection!), so there’s plenty of room for personal variations.

Figure 10 shows everything in my “WBR-O” box. You might note how the box has extra holes and knobs that are leftovers from all the previous versions. I included the ubiquitous 1k pot at the antenna input, but in hindsight, it isn’t needed when I use both circuits. I also decided to use a small 12 volt supply that fits in the underside of the chassis. It’s a good quality and lacks the persistent hash that some cheaper supplies produce. Right now I’m using a Delta PMT-12V35W1AA from Digi-Key. Small, quiet, safe, and at about $15 for a universal AC input it’s a bargain, really.

I’d recommend a 10-turn pot for the regeneration control. Regeneration at low frequencies is higher, dropping somewhat as you increase frequency. You’ll have to develop a feel for this adjustment. With the new biasing I’ve included, it might be possible to use a single-turn pot, but that’s just one more experiment for the future. The key thing when getting started with a regen is that you want as little regeneration as possible. I still have the bad habit of leaving it too high then wondering why things don’t sound right. Overbiasing the Q-multiplier just adds distortion, and even harmonics. Regeneration also varies the frequency a bit, so tuning in on a single CW station, for instance, will require a little practice, but in the end it becomes second nature.

Table 3 lists a few of the more critical parts and their Digi-Key part numbers. Although I use Digi-Key as a supplier in a lot of the electronic work I do for a living, they are also quite amenable to serving the needs of individuals as well. The same can be said for Mini-Circuits and Amidon. Prices are reasonable and you get to choose the quality you can afford.

All that being said, the final product was gratifying to use. Just about anything one of my other older high-quality radios can hear can be heard by this little gem. Of course, selectable selectivity, a noise blanker, and good AVC would help, but…that’ll come later, I think.

With all the changes I’ve made the one ingredient I hope I have added most of all is flexibility. The real possibilities of what this basic design can do have only been barely touched. I think it would be excellent, for instance, to see how small one can make the entire unit, so it fits into camping gear or such. Also, why not 2 oscillators (sharing the same tuning voltage) instead of one, and use the second to drive a small transmitter? Or a second oscillator to drive a frequency counter so you can actually see where you’re tuning? Or using what I’ve explored here on other designs, like AA7EE’s “Sproutie” regenerative? Or add some good audio filters? Or, as I said before, adding AVC via the gain control voltage on the Oscar circuit? Or…well, you get the point. This is definitely not a software defined radio, but an imagination defined radio, and, as Oscar would note, perched high above me on my equipment, the sky’s the limit.

—–

Sheldon Hutchison, N6JJA, has been a licensed amateur on and off since 1965 and currently holds an Extra ticket, a Ph.D. in Electrical Engineering (University of Illinois) and is an ordained Episcopal priest. He and his wife Eileen, KI6UZJ, live and work in the Silicon Valley with their cats, including Oscar and several other “helpers.” Dr. Hutchison works in the laser industry (while also a retired but active priest) and Eileen is employed in the Valley’s aerospace industry. An avid experimenter, Dr. Hutchison also enjoys restoring old “boat anchor” receivers, and currently—according to Oscar—needs to find homes for a few of them to give his helper more room to play.

 

ADDENDUM

File this under “Can’t leave well enough alone.”

There’s something in electromagnetics called “reciprocity.” Basically, it means that if signals get into an antenna or circuit efficiently, they get out just about as efficiently. Taking long walks at noon helped me find ways to use this phenomenon. For my “Oscar” receiver I found one. I’ve long used my spectrum analyzer, clipped to the low-impedance tap on L1, to indicate the health of the regenerative Q-multiplier by making it oscillate and observing the relative strength of the signal. Well, in the schematic below I experimented with adding a link right next to the low-impedance center tap of L1 as a way of improving coupling to the Q-multiplier from a 50-ohm source. Compared to the previous way of coupling to the center tap the signal measured on the spectrum analyzer shot up by 30 dB! Adding 2 turns is too much. Moving the link too far away from the tap makes oscillation much more difficult. And after adding the link, you’ll want to go back and re-tweak C4. The photo also shows the red wire link as I installed it.

Now Oscar’s even happier.

To many, this will be just another Si5351 VFO project, with nothing to distinguish it from the others. In fact, that’s exactly what it is. The “how to” of connecting an Arduino board to an Si5351 board, wiring up a display, and loading the firmware, is straightforward, and well established. To me though, it was a complete mystery. I have been very adept, my whole life, at studiously avoiding anything to do with digital electronics, computing, coding, and the like. When my friends in school were getting excited over Sinclair ZX81’s, BBC Micros, and Commodore 64’s, I was building a one-tube regen, an 80M DSB transceiver, listening to my British military R107 shortwave receiver, and talking to hams on the local repeater on my converted Pye tube VHF base station. I remember wandering into a Tandy store (what us Brits called Radio Shack) in the city of Worcester at some point in the late 70’s, and being greeted by the sight of a Tandy computer – probably a TRS-80 or something similar.

“What does it do?” I asked the salesperson.

“What do you want it to do?” he replied.

This seemed like a strangely non-committal response. Maybe he didn’t know what it did, and was merely throwing the responsibility for finding out back on me? I don’t remember anything about him now, but perhaps he was some gangly teenager who knew little about the stuff he sold, and whose main thought was getting off work so he could go to the pub with his friends? That’s it! He was just trying to appear knowledgeable by giving me a non-answer! This suspicious reaction was quite representative of the way I thought about computers back then. Just as it’s hard, if not impossible, to get the measure of a person if they willfully refuse to reveal anything of themselves to others, so it seemed with computers. These expensive boxes just sat there, doing nothing, except waiting for instructions. Such a disappointing lack of character! How is one supposed to respect a person or an object that sits quietly in a corner, waiting to be told what do? How feckless! Dedicated hardware, however, was different. When you bought a radio receiver, you knew that, on twiddling a few knobs and flicking a few switches, it would receive radio signals. A burglar alarm would alert you to the presence of burglars (well in theory, anyway), and those remote control cars that RS sold by the gazillion were guaranteed to quietly drive your family nuts in the days after Christmas before work, and school, resumed. Computers, on the other hand, promised everything but actually did nothing, until you told them what to do – and even then, there were a myriad of ways in which they could obstinately refuse to comply with your wishes. Not for me!

And so it was that, throughout my adult years, I deprived myself of exposure to things digital. I am not proud of my incurious nature about many things – though, when I am interested in something, I exhaust myself with the sheer intensity of focus. It’s an odd type of blinkered approach to the world that leaves others confused. I can’t say I blame them. The projects I built ran off anything from a few volts, up to 15-20V or even more. 9V batteries worked fine (we called ’em PP3’s), as did the old 12V lead acid battery that would no longer power the family lawn mower, but did a sterling job of powering the radio gear in my bedroom. My circuits weren’t picky about voltage, but these new-fangled digital chips just wanted to see 5V. Really? What kind of a voltage was that? They came with a surfeit of incomprehensible nomenclature too. Words that sounded like something John Lennon would have made up for a song post-1965. Words like NAND. You know, it wasn’t so much that this stuff wasn’t interesting – it was simply that I was really into building little radio receivers, and didn’t see how this digital stuff could help me (I wasn’t very imaginative). I had a small stash of ferrite rods, variable capacitors, resistors, and transistors, and some 9V battery snaps and with that, I had all that I needed. These were the days when loading a program onto a computer meant playing an audio cassette into the “line in” jack of your computer. I just didn’t see how any of that world full of DOS, NAND gates, tiny amounts of RAM, and the like, as well as really weird voltages like 5V, could possibly apply to me. Yes – I was that closed-minded. It’s not hard to see how a few years later, when we all graduated from University, my colleagues went on to successful careers with big technology companies, designing integrated circuits, and building the backbone of the internet, while I moved to Los Angeles and promptly became a DJ

A few years ago, a friend generously gifted me a Bare Bones Arduino board. I didn’t know what it was. It had header pins sticking out of it but, at that point, I didn’t really know what header pins were, or how to connect to them. I looked at it, and wondered what to do with it. What did it do? How did it do it? What was I supposed to connect to it? I placed it carefully in a box along with some other electronic things that confused my simplistic analog mind, and carried on with my life. Every now and again, I’d take it out of the box, blink at it a few times, and put it back. I knew that Arduino was the new big thing, and something that was going to play a big part in ham radio homebrewing in the coming years but I guess that, with my toroids and air-spaced variable capacitors, I wasn’t ready for it yet. Not that my experience level in this arena was completely non-existent. I had taken part in the beta tests of the Etherkit CC-20 and later, the planned CC1 series of QRP transceivers. These experiences had taught me that I could solder SMT devices, and even replace an SMT ATMega328P with nothing more than a soldering iron, soldering wick, flux, and a sharp blade. It was a small revelation to learn that I could do this stuff. Jason NT7S very patiently walked me through the process of flashing the firmware onto the ATMega328 via the ICSP header mounted on the board. This was a first for me, and quite exciting to gain a new skill, which proved handy when I built the SPT “Sproutie” Beacon, and needed to flash firmware onto the ATtiny13 in that little transmitter.

Then, recently, I took the Bare Bones Arduino board out of the box which had been it’s home for a few years and, this time, something clicked. “Goshdarnit” I thought, “I’m going to make an LED blink. If others can do it, so can I!

I spent a few days and nights with the LED blinking and pulsating at various rates, as I loaded different sketches, and adjusted the parameters. As fun as flashing lights are to a simple lad like me, it wasn’t the reason I wanted to resurrect this little Arduino board from it’s relaxing life in storage. I had an Etherkit Si5351 Breakout Board that needed to have life breathed into it. I wanted to generate RF, by golly!

This next stage was where things started to come into focus, and it began to dawn on me that using one of these little breakout boards to generate a stable RF signal wasn’t all that hard at all – well, from the point of view of the end user, at least. Once I did a bit of reading up on how to control the Si5351, I was just a little gobsmacked. You mean all it needs in terms of data input is 2 connections? SDA (serial data) and SCL (serial clock)? That’s it? I made those 2 connections between the Arduino and Si5351 board, uploaded the Etherkit Si5351 example sketch, and almost fell off my chair when the Si5351 began emitting RF on the frequency I had entered into the sketch just before uploading it. It was a moment of realization – that this little board actually was a programmable oscillator. How incredibly neat! No more custom-cut crystals – for ~$10, you can get a board like this, and program it to the frequency of your desire (within it’s specified limits), and it replaces both the crystal and the oscillator. For a single frequency, once you have programmed it, it doesn’t even need a micro-controller connected to it.  Fantastic!

I suppose that was the moment at which my mind, which moves at the speed of molasses, “got” that a VFO with this board is really a micro-controller which, with the help of a rotary encoder, is re-programming the Si5351 “in real time” as the encoder knob is turned. Every single click of the encoder sends a new instruction to the Si5351, to step up or down, in an increment which the firmware has already specified. I was hot to trot, so began looking around for a basic Si5351 sketch. The word “sketch” reminds me of the Etch-A-Sketch which I never came close to mastering as a child. I must admit that I think this association slightly trivializes Arduino programs (which are written in a type of C) in my mind, but that is what they are called, so that is what I will call them.

What I was looking for was a sketch that would allow me to vary the frequency of the Etherkit Breakout Board continuously in the HF region, from at least 3 – 30MHz. I was only concerned with one of the 3 clock outputs. At this point, I simply wanted to use it as an HF signal generator for testing purposes, or to control a general coverage direct conversion receiver. Perhaps at some point, I’ll begin fiddling around with code, and learning how to modify it for my own purposes, but at this point, I wanted a sketch that I could upload to the programmer, and immediately be in business. I also wanted to use one of those tiny little OLED displays, due to the enclosure I was considering.  This was the point at which I found Thomas LA3PNA’s sketch entitled, “A simple VFO for the Si5351 for either LCD or OLED.” Perfect!

This was also the point at which I discovered that the ATMega168 in my little Arduino clone board didn’t have enough memory to hold Thomas’ VFO sketch. I considered purchasing a newer Arduino board or clone, but most of the ones I saw had more stuff on them than I needed or wanted, in terms of inputs/outputs and programming ability. All I wanted was the ATMega328P, and a 6-pin ICSP header to program it with. Then I remembered back to my time working on the Etherkit CC-20 beta, and how I had expertly fried the micro-controller. Jason sent me a replacement and, wisely, included a few extras, in case my prowess at destroying delicate chips were to reassert itself. I still had those little SMT ATMega328P’s lying around, as well as a supply of breakout boards to mount them on. Problem solved! Building something from parts on hand is so much more satisfying than purchasing a ready-made solution – at least, for the first time, it is.

I sat down to scribble out a schematic, and it was during this process that the realization hit, as to what an Arduino board is. What makes Arduino, well, Arduino, is not the board, but the software platform that supports it. Apologies for stating what is well known fact to many readers, but this had all been previously unknown to me. The board itself is really just a micro-controller, with the power supply and input/output options either suited to the tasks at hand or, in the case of a larger and more general purpose board, such as the Uno, many different such options, to make it as versatile as possible. Ths was fantastic, because what it meant was that all I needed to control the Si5351, was a micro-controller (ATMega328P), a 16MHz crystal with the two associated capacitors, a 5V power supply, and some 0.1uF capacitors for bypassing. Oh – and a 6-pin header for programming. The schematic for the VFO is simple because, as far as the hardware goes, everything happens inside the micro-controller and the the Si5351 (which are both internally complex). The rest of it happens in the firmware. As far as hardware goes, we’re simply tasked with the 21st century equivalent of assembling a crystal set.

Here’s what I came up with. There are an awful lot of unused pins but, for this purpose, there are a lot of pins we don’t need. Without thinking, I was about to connect the AREF pin to +5V, because that’s what I was seeing in the various schematics I was using as references, until it occurred to me what AREF stands for. It’s an Analog REFerence pin. This application uses only digital inputs and outputs, so figuring that I didn’t need an analog reference, I didn’t connect it –

While planning this little VFO, a number of questions were presenting themselves to me. The main one concerned the issue of both the Si5351 Breakout Board and the OLED display being connected to exactly the same SDA and SCL connections. The I2C protocol does allow for multiple devices on the same line, but my understanding was that if more than one device is employed, then the firmware needs to include the unique address of each device. Would Thomas’ sketch work from the get-go, I wondered? As it happened, it did and, as of writing this, I don’t know if this is because

a) the address of the OLED was included in the library definition for this little display, or

b) with a setup like this that only has 2 devices connected, the instructions for the OLED are ignored by the Si5351, and vice-versa.

I’d like to be able to describe the exact steps I took when setting up the sketch, but I have, lamentably, forgotten them. I do remember installing the UG8lib library in the Arduino IDE, which supports the commonly available OLED displays. I also remember, at some point, uncommenting a line that specifically refers to devices that have an SSD1306 driver chip. If you purchase a cheap monochrome 128 x 64 OLED, this is probably the driver chip your display will have. These little displays are available for <$3 including shipping. Deal!

Here’s the ATMega328P mounted on the breakout board –

The controller part of the circuit constructed, with it’s supporting components. No power supply yet, as during initial testing, it will be powered through the ICSP header –

And with the Etherkit Si5351 Breakout Board fitted. The I2C control lines and 5V supply line are connected underneath the Etherkit board –

It took a while to figure out how to mount the OLED to the front panel. The 4 mounting holes are sized for #2 screws. I thought of running 4 #2 screws from the front panel, straight through to the OLED, and spacing the display away from the panel with 4 #2 nuts. The nuts were so close to the glass covering the display though, that they could have cracked it while being tightened. In retrospect, stacked #2 washers might have worked, though long before thinking of that, I came up with this rather more complex solution. It involved a small piece of PCB material, drilled and cut to size. #2 shakeproof washers and 3/16″ x 3/16″ nylon spacers were also employed. Their use should be apparent in subsequent photos –

I seriously considered fabricating a PCB enclosure for this little VFO, even getting as far as cutting some of the main pieces. The primary reason for wanting to use a custom enclosure was that the other case I was considering (which I ended up using) was a little too high. As a result, the front panel had, in my opinion, too much empty space. A PCB enclosure, about 4″ x 4″ x 1.5″ high would have looked mighty spiffy. However, I didn’t have the mettle to go through with it. I just couldn’t get quite inspired enough to put all that extra work into making a custom enclosure, and fell back on my favorite ready-made enclosure, the 143 from LMB Heeger. It is 4″ x 4″ x 2″ high, and available in plain aluminum finish, smooth light grey paint, or a sort of wrinkled black finish. They are also available with either an undrilled cover, or a perforated cover. The encoder was connected using header. The main reason for this was that I wasn’t sure if the cheap Chinese encoder ($1.68 each, inc shipping) was up to the task, so wanted to facilitate easy replacement –

A close-up, showing a little more detail of the mounting of the OLED to the front panel. Like the encoder, the display was also connected using header, making installation, and any dismantling for repair or upgrade purposes, easier –

Amazingly, it worked!

With the perforated cover you could, if you wanted, add some internal LED’s, for a splash of light to brighten up the shack, and add some flair. Being frugal energy-wise, I left the LED’s out for the time being. As it stands, the VFO already consumes 86mA at ~12V, which is not an insignificant amount –

Although the hardware side of this little project is finished (or very close to it), I am still very much fiddling with the firmware. As well as using Thomas’ code, I have been trying out sketches from other folk too. I began by trying to find a sketch that would do exactly what I wanted it to do, and fast discovered that, at the very least, an ability to modify code was required. That lesson led to a desire to actually develop a more complete understanding of C, so that I can at least do intelligent re-writes, if not write my own from scratch. This is all a bit overwhelming, and I vacillate from having fun, to being very grumpy, and back again

Thank you Thomas LA3PNA for the sketch – and also to the many others whose code I have been borrowing, and will no doubt butcher. I view this little VFO very much as a learning platform, from a programming point of view. Also, a big thank you to Jason NT7S for the Etherkit Si5351 Breakout Board, and the very useful libraries, which are seeing much use from homebrewing hams.

PS – I just started reading “Beginning C For Arduino” by Jack Purdum. Great stuff.

UPDATE (Jan 30th 2021) – I made a few small changes to the sketch, to improve the display a little. I found a font to display the frequency, that is a little larger. It is in the U8G library, and is known as helvB18. I also removed the word “step” from in front of the step size, as I felt it was self-explanatory. The text needed to be moved around on the display a little to fit it in, and this was adjusted in the sketch.

Since building this little clock generator, I noticed that, although the sketch set the Si5351 to start up on 7030KHz, it would nearly always start up 100Hz higher. Unsure whether this is a software or hardware issue, I did the clumsy workaround of altering the sketch to start up on 7029.9KHz. As a result, it now starts up on 7030KHz. I’d like to find the reason for this discrepancy, and fix it at the source.

I also added a 5mm red LED inside the case, with a 1K resistor in series with the +ve lead, connected to the 12V DC power connector on the back. The LED lights up the whole inside of the case, though this is not so obvious in the daylight shot below – 

However, when used indoors, the red light is visible through all the perforations in the case, and looks great. It’s almost reminiscent of the old tube days. The red light effect actually looks a bit better in real life than in the photo. This Si5351 VFO/signal generator draws 105mA, of which 10mA is the LED –  

Although I’ve had quite a few thoughts about using it as a stable signal source for other projects, it’s very handy to have around the shack, just as it is. Calibrating receivers is a lot easier, with this accurate signal source. I didn’t run the calibration sketch but, through a process of trial and error found a correction integer to place in the sketch, that places the VFO to within a few Hz, as verified by beating the output against WWV.

My build of the K7TMG HF Morse Code Thermometer was fun, and it inspired me to use the same circuit to create a new HiFER beacon to honor my neighbor’s cat Boris. With some of my indoor cat-owning neighbors in the past, I have acted as caretaker when their parents are out of town or at work. I don’t have that kind of relationship with Boris though, so she and I stare longingly at each other through the window when she is out. (Edit – I’ve looked after Boris a few times since first writing this post, so kitty-human relations have definitely progressed!) We are two beings sharing a mutual admiration, but separated by a sheet of glass –

When there’s a kitty who you want to hang out with but can’t, the obvious thing, of course, is to build a little HF beacon to transmit their name in Morse code. It’s the thing to do and so, I found myself building another K7TMG HF thermometer, but without the temperature-sensing circuitry. I also added a 2-section LPF to attenuate harmonics. I used the capacitance and inductance values that Chris Smolinski uses in his HiFER beacon kit, but recalculated the number of turns on the toroids so that I could use T37-6 cores instead of the larger T50-2 ones he uses. I think that the tuned tank circuit in the collector of the oscillator transistor must also help reduce the harmonic output of this stage as the level of the 2nd harmonic at the antenna is further down on the fundamental than I would normally expect from a 2-stage low pass filter like this –

 

For the firmware, I found a very versatile and useful piece of AVR beacon freeware written by Nick SV1DJG. If you use the circuit above, in which pin 2 of the ATtiny85 keys the oscillator, you’ll need to change the line

#define ledPort 1

in Nick’s code, to

#define ledPort 3

If you leave the output port as port 1, you’ll need to make pin 6 of the ATtiny85 the keying line instead of pin 2. If you want to make pin 3 the keying line, just specify “ledPort 4″Also, in the code, you can specify the message (or callsign/ID) you want to send, the keying speed, and the length of pause before it repeats. My beacon sends the letters BRS at a speed of ~5wpm, with a pause of 2 seconds. If you want to send QRSS with this program, there is also an option to specify the dot length in milliseconds. It is currently set at 1200. The dash length is derived from that, being specified as 3 times the dot length. Inter-character and inter-word spaces are also defined in terms of the dot length, so when you change the dot length, everything else follows.

The build went smoothly, and there’s not too much to say about it. As always, Rex’s MeSQUARES and MePADS did a sterling job of making the process of building Manhattan-style a lot easier. I cut the board to fit into a specific enclosure, and it worked straight off the bat. The trimcap had very little effect on the amplitude of the output signal, and the oscillator started perfectly at all settings. A fixed capacitor of around 47pF probably would have worked as well. There is also room for experimentation with the values of the 2 feedback capacitors, which are 470pF and 330pF in my circuit. Lowering those values will shift the oscillator up in frequency. Two 100pF capacitors should work. You may even be able to go lower in the value of these capacitors. My oscillator came up on a nominal 13556.9KHz. It was a good frequency for me, and didn’t seem to conflict with any of the HiFER beacons listed over at the LWCA website page of MedFER, BeFER and HiFER beacons. Great – no need to change any components!

Unfortunately, a different enclosure was sent from the one I ordered. The one I wanted had 2 external lugs for fixing it to a wall, post etc. The lack of these meant that I had to drill holes and fix it with screws protruding from the inside of the enclosure. It wasn’t ideal, as it meant more holes needed to be sealed to prevent moisture ingress. At this point though, it was the enclosure that I had, so it was the one that I used. It’s a nice weatherproof enclosure, available from China for as little as $3.41 inc shipping, or just a couple dollars more if you want it quicker from a supplier within the US. There are versions with external mounting lugs and clear tops too, if you like that sort of thing. An eBay search for “85x58x33mm waterproof plastic box”, or similar, should show plenty of options –

I wanted to mount this little beacon outdoors and power it exclusively from a small solar panel with no battery. This meant that it would only operate during daylight hours, of course, but I’m thinking that some grey-line action should still be possible, as the beacon will still be operating when locations just to the east are entering their grey-line phase. Living in a rented multi-unit building means that I need to be cognizant of the wishes and sensibilities of others, and I didn’t want to take the chance of a battery exploding inside a very hot enclosure in the summer heat. It’s probably unlikely, but with little previous experience in this area, I didn’t want to take the chance. Besides, the idea of a little circuit that is entirely dependent on the sun in a very direct fashion appeals to me. The panel I used was an old one that I bought cheaply as a lot of two, from a fellow on eBay who decided he didn’t want it, immediately after purchasing it. When drilling and filing the hole in the enclosure for the cable from the solar panel, I was careful to keep it as small as possible, so that sealing it against the elements would be straightforward –

Boris seemed to like it –

The plan was to install it on top of a fence on the property line of the building I live in. Sitting on top of the fence, the solar panel would receive light until fairly close to sundown with little obstruction from nearby buildings. My tube of silicone marine grade sealant had dried up, so I decided to try using a product called Plastidip, a can of which I had on hand. It’s a black rubbery solution that comes in an aerosol can. You spray it on, and it forms a weatherproof seal. I’ve used it successfully in the past for sealing the ends of coax at dipole center-feed insulators, so figured it should be usable in this application too, though perhaps not quite as easy to keep to a small area as squeezing silicone sealant out of a tube. Here’s a close-up of the beacon just below the top of the fence –

I sprayed the the screws that fixed the enclosure to the fence with Plastidip, and at this point began to wish that I had either held out for the enclosure with the external lugs, or at least used silicone sealant. I had forgotten how very liquid Plastidip is before it sets. Much of it dribbled down to the bottom of the enclosure and pooled. You can see it oozing out from the bottom of the board in the next shot. I’m not sure whether it conducts when in the liquid state, but I didn’t much like this. Plus, it just looked messy –

All this time, I had been monitoring the beacon signal with my K2 on a battery, to make sure that I didn’t break any connections during installation. Strangely, at this point, the beacon had stopped, and was just emitting the occasional dit or dah. I guesssed (incorrectly, it later turned out) that perhaps the liquid Plastidip was conductive, and was the cause –

I pulled the board out, cleaned up the oozing mess with Q-Tips, then reinstalled it in the enclosure. Poking around the micro-controller with my fingers, the beacon sprang back to life. I wasn’t able to determine exactly what had caused the problems, which concerned me. Unfortunately, I was pushed for time, as I was trying to complete the installation before one of my neighbors returned, a woman with whom, sadly, relations have completely broken down. It’s a long and uninteresting tale but at this point, nothing I can do or say will help things. It seems that I have been identified as a mortal enemy.  The fact that she doesn’t like cats doesn’t help either   At this point, I decided to press on with the installation as swiftly as possible. I stapled the dipole antenna just underneath the top of the fence in both directions, and mounted the solar panel on top with two short screws –

This is the type of install I was aiming for – unobtrusive. My neighbor on the other side might see the solar panel, but I was hoping that they wouldn’t mind. You never really know with folk what will bother them and what won’t. It’s at times like this that I can see the advantage of owning my own place with a big plot of land in a lesser populated area. The dipole is horizontal and only about 8 feet above ground level, so it’s probably a bit of a cloud-warmer. Definitely a compromise installation –

I went back inside and, monitoring from indoors, was happy to hear a good signal coming from the beacon. The letters “BRS” were being sent at 10wpm (my initial setting) with a 2 second pause before repeating, and absolutely no chirp on the signal. Monitoring the signal on and off throughout the rest of the day, I was happy to note that it stayed on the air for about 2 1/2 hours longer than it had when located indoors with the solar panel mounted in the window. It continued to transmit until about 48 minutes before local sunset. Exposure to direct sunlight makes a huge difference to solar panels. If I had been able to angle the panel toward the sinking sun, no doubt I could have eked out a bit more time on the air. All was good. I was happy, and fell asleep that night with the K2 on, expecting to wake up to the next morning to the sweet sound of the letters “BRS” singing from my K2.

Instead, I awoke to the sound of a minute or two of dits, with the occasional pause, a few meaningless dits and dahs, then another minute or so of dits. Perhaps as the sun rose higher in the sky, the situation would correct itself, I thought. It didn’t however, and at 10:30, with the sun fairly high in the sky, and more than enough light to power the beacon, it was still sending out long series’ of dits, punctuated by occasional pauses, a few dits and dahs, and then the next long series of dits. Not a BRS to be heard anywhere. This was disappointing, and not what I expected. I decided to dismantle the installation, so that I could take my time trouble-shooting inside, as opposed to at the top of a step ladder. Bringing the beacon inside, I put the solar panel in the window and – lo and behold, the beacon started up perfectly, sending out it’s BRS callsign exactly the way it was supposed to.

So – why wasn’t the micro-controller starting up properly in the morning? At this point, I did what any modern 21st century renaissance man would do, and Googled it. A few others had experienced this exact same issue, of an ATtiny not starting correctly when powered just by a solar panel with no battery. One explanation offered in a forum seemed very likely – that when the solar panel is beginning to receive light, as the voltage gets to the point that the micro-controller wakes up, a small panel still isn’t capable of supplying much current. Anything else in the circuit that draws current, such as the crystal oscillator, will cause the voltage to drop below the point at which the micro-controller can operate properly. At this point, I was using a 78L05 regulator, which was drawing ~4mA of quiescent current. It’s not a lot, but when light is low, and the panel is only supplying a few volts, that extra current draw was most likely enough to cause the voltage to sag when the oscillator kicked in. Listening to the beacon, it seemed very likely that this is what was happening. The ATtiny, in the low light, had enough current to operate, so it turned the keying pin high, at which point, the oscillator began drawing current. However, that extra current draw caused the voltage to sag below the point at which the ATtiny could operate. As a result, the ATtiny turned off, the keying pin went low, the oscillator turned off, the voltage went back up, and the whole process started over. This gave way to the transmission of a constant series of dits, instead of the beacon callsign. Unfortunately, as the sun rose higher, and the light level also rose, the ATtiny was not recovering.  What was needed was to set the BOD (Brown-Out Detection) to a voltage level such that when the voltage from the panel equals or exceeds this voltage, it is also capable of supplying enough current to the entire circuit without the voltage dropping below the BOD voltage. I reset the BOD to either 2.7V or 4.3V (I forget), from it’s previous level of 1.8V and this seemed to solve the problem. However, with the beacon in it’s new (temporary) position indoors, with the solar panel in the window, the higher BOD meant that the beacon often didn’t come on until late in the morning, due to the fact that a) it was a small panel and b) panels in windows tend to generate much less power than panels outdoors.

I wanted to make this little setup as efficient as possible before putting it back outside, so swapped out the voltage regulator for the one shown in the schematic – a 5V LP2950. This series of regulators has a much lower dropout voltage than the 78L series – between about 0.04V and 0.38V, depending on current draw. They also have much lower quiescent current, at less than 0.1mA, compared to around 4mA for the 78L series. My final version of the Boris Beacon had an LP2950, and the BOD on the ATtiny85 set to 1.8V. You can do this with the 10PU version. By contrast, he lowest BOD on the 20PU version is 2.7V. It worked like a charm! I’ve had the beacon in the shack, powered just by the small 1.8W solar panel sitting in an east-facing window. It starts running early in the morning, even on overcast days, and stays on until fairly late in the afternoon. It would run for even longer hours if the panel was mounted outside. This was a very encouraging result.

You’ll notice that the capacitor on the input side of the voltage regulator is shown as a 100uF part. Normally, I’d use something in the range of 1 – 10uF, and I did start with a 1uF cap in that position. A larger value capacitor in that place helped to smooth out the voltage swings when the light level was marginal. When the ATTiny was beginning to send a semi-random series of dits, due to the sagging voltage issue previously described, a larger value capacitor helped to mitigate that somewhat. A 330uF, 470uF or larger part could help a little more but ultimately, when the light level falls, it falls. At most, I doubt that a large cap here would buy you more than an extra few minutes of operation at the very beginning and end of the day. Another idea for experimentation would be to try a different transistor. I’m wondering if, say, a 2N2222A would provide a little more power?

At this point, I feel that the experiment is complete, and am not feeling the need to mount it outside again. It would be interesting to see if the mighty 1mW RF output could earn me any spots, but that was really not the main purpose of doing this. My primary motivation was an interest in the circuit – building it, and optimizing it. I did order another enclosure, with lugs, which would be perfect for outside mounting. Alternatively, this case could, as well as housing the circuit board, effectively act as the center part of a dipole, with the lugs acting as strain reliefs. The wire carrying the power could hang down from the center of the enclosure –

The above case was bought on eBay, from a US seller, for $4.83 including shipping. I saw what looked like the same case from a Chinese seller, for the lower price of $1.56 + $2 shipping, so bought that as well, to compare the two.  Interestingly, the cheaper one directly from China looked like exactly the same case, but of inferior quality. It looked like it had been cast from essentially the same mold, but wasn’t as nicely finished. One imperfection almost made for a bad seal with the lid.  I intend to purchase a few more of this case, for future use, but will make sure to get them from the US seller. For sealing the holes where the wires enter, a silicone sealant should be more practical than the rubbery Plastidip that I had used earlier. This stuff should do the job nicely –

Another idea for an outdoor beacon enclosure would be an electrical junction box. I found this in my local hardware store. It’s 4″ x 4″ by a little over 2″ high. It has a rubbery seal around the lid, and is certified for burial, so should certainly withstand the weather above ground. It also has 4 lugs on the outside for fixing the enclosure to a wall, fence, or post. For securing the circuit board, the adhesive standoffs pictured should work well, so that you don’t have to drill holes in the enclosure. They are available for a 5/16″ hole, and in several different lengths. The ones pictured hold the board 3/8″ away from the box, and were part # 91443A130 from McMaster Carr. Someone in one of my discussion groups, when talking about outdoor enclosures for transmitters, suggested that if you have a completely sealed box, with no ventilation, it might be an idea to add a silica gel packet or two, to prevent condensation from forming inside the enclosure. I haven’t had any issues with the completely sealed non-metallic enclosure that houses my outdoor Part 15 AM transmitter, but it’s a good idea, and definitely worthy of consideration –

This little beacon has been greeting me in the mornings for the last few days, with the letters “BRS” sent at 5wpm. I rather like the fact that it comes on every day with the light, and goes to sleep at night – the way that we all did before gas and, especially, electric lighting came on the scene.

UPDATE – The Boris Beacon is now on the air. Details here!