At the time I'm writing this, United States I bonds are paying more than 3% interest. While they're not as easy to get as when I was a kid, savings bonds are still as good of a way as any for kids to learn about savings. The gang has been pestering me for two weeks, (pestering at my request, or I'd forgett to do it), to use a $100 windfall they came upon to buy savings bonds for them. I finally did it! Here's what I learned.

Kids have to have an account managed by an adult

Actually, one really big note

As a parent, you'll create a 'linked account' for each kid

Actually buying the bonds

The last weird thing

 Halibut Electronics is working on a new satellite antenna kit!

This is kinda cool for two reasons, first because we've recently started attending the high altitude balloon meetups at Noisebridge. Satellite antennas came up during one of the meetings.

More tactically importanly though, the EggNogs docs inspired what be a better tuna can feedtrough for Projct TouCan's antenna! For notes, here's my original EggNogs documentation review reply:

The documentation looks great so far! I've made it to page 17/22. One thing:
 
 
For those of us with partners heavily into fountain pens, those of us who like to print out manuals on JIS B5 paper and then store them in Kokuyo Campus binders, page numbers in the table of contents would be very cool. (I know, I know, such a niche group :) )
 
Mostly though, I wanted to thank you for jogging my memory into a, (I hope), better solution for Project TouCans antenna ports. At present, they're inverted bananna plug posts. Banana plug screw terminals are tiny, and therefore somewhat problematic in outdoor environments. Here's an idea of how tiny
 

 

 
Suspending a two pound rig 6 meters up, eventually the screw threads begin to strip. We stuck with banana plugs though because the insulator around the conductor makes them perfect for mounting in a tuna fish can. Yes, I've read about how Yagis and dipoles are balanced and therefore have 0 volts across the antenna center, so theoretically you don't need insulators, but TouCans frequently hangs at angles to the ground and/or very close to the ground,  so, insulators.
 

 
Anyhow! The last figure on page 17, the one with the cool weather-proof washer, reminded me that gromets exist! With the correct sized gromet, we can put any size bolt pointing upward as an antenna connector.
 

 
And, even better still, we can source them from our local hardware store at the bottom of the hill!
 
Thanks Mark!
72 de KD0FNR Hamilton

 Last night was a shorter run at things in general because KO6BTY, (the 13 year old known as Diaze here), and I got a later start.

We spent what felt like forever, but what was actually only 12 minutes trying to share files from the 'usual' file side of the kid's Chromebook with the Linux side using file folders and whatnot. Nothing worked. For whatever reason, the Linux folders weren't visible in the machine's 'My Files' app. Sharing folders led to the machine basically hanging. Then! Then, we handled the issue like a couple of programmers, and instead of downloading in one system and trying to copy to anther, Diaze just ran the following from her Linux terminal

wget https://data.cosmic.ucar.edu/gnss-ro/cosmic2/provisional/spaceWeather/level2/2024/203/ionPrf_prov1_2024_203.tar.gz

That was snazzy! It just brought the file right in because, well, command line interface tools are just... snazzy. 

Having a chat record of our work together is also really helpful. I remember feeling like we'd wasted so much time trying to copy files over. We literally spent 12 minutes. Not. The. End. Of.The. World. Happily!

Then! The kid used tar to unzip the compressed data tarball. Watching her as the 4000 files fold out onto the drive was a lot of fun. This is a good project! I also loved the part where she caught on that this was the same Python with the same constructs she'd been using on the Project TouCan's remote Morse code key.

From there, we got back into using the netCDF4 package to look at data. We talked about Python Dictionaries and then looked at the latitutude values of a radio occultation satellite pass that measured the electron density of the ionosphere. 

The goal at this point is for KO6BTY to create a map of the latitude and longitude values in Datasette to see where the satellite pass took place with respect to Earth.

 KO6BTY and I are making another run through Python.

Diaze has learned a bit of Python in the past when she set up our QSO mapping app to pull in pertinent ionosonde data from the Digisonde ionosondes. Now, we're working with Python again to analyze data not from ground-bound ionosondes, but from the COSMIC2 constellation of satellits that provide ionospheric data includihng electron density profiles.

That was the intro, and the application, but this post is more about how to informallly teach Python. What will work, and what won't? With unschooling, a lot of learning is initiated by something called strewing. Strewing as it's comonly defiined is, essentially, keeping things a little cluttered around the house. It's leaving reading material, projects, web sites, and so on, out where everyone in the house, including and especially the kids, can see them. I've widened the definition to include our entire indoor and outdoor lives, and the city and world at large. For us, it works in more ways than one.

But, it's a very informal activity designed to lead, perhaps, to more formal educational activities depending on the kid's interests. Here's the question: can a topic also be taught informally? I've had varying results with this. Sometimes the kids here pick up things hanging out with my partner and I. Diaze picked up writing by attending writing groups with me. Mota, who's manual dexterity is through the roof, picks up physical activitiese by hanging out with other people doing them as evidenced by me getting nailed in the chest by a perfect spiral football pass when he was six. Tawnse has picked up and lost physics more than once just soaking it in from my partner's online lecture productions.

I enjoy showing the gang how to do things but, especially for me, and especially with Python, there can be a lot of unexpected set up that can start to slow progress and sometimes drive everything to a halt. Last night, it was working in a venv vs not workign in a venv on a Chromebook with respect to installing packages with pip.

I, of course, have setup my Chromebook at sometime in the past to let me do pretty much whatever the heck I please. I don't remember how I did it. Meanwhile, Diaze's Chromebook was adamant that she not install netCDF4 with pip as it might ruin the entire Python install.

I'm happy to report that we did arrive at the venv solution mostly through our experience with Datasette plugins. The development section of most plugins mentions using venv. That did the trick. Once Diaze was within a virtual environment, pip felt comfortable enough to allow the install.

We stopped there for the day.

I'm going to try to keep a lab notebook of how this goes. I caught Covid this week, which  might seem irrelevant, but Diaze and I are doing all of this work through Google Meet so I  don't expose her. Consequently, I have our entire chat streams as a result. I'm going to see if I can somehow weave those in as data for how to improve or change or show what works about this 'teaching/learning' method.

It outghta be fun.

I've been referred to a lot of indie web blogs of late, and it's paying off quite nicely for me Here's the latest example.

Simon Willison--co-creator of Django and creator of Datasette--has a blog that led me to Maggie Appleton's site. Once there, I found a very nice, and very pretty primer on databases. Within the primer was a perspective I'd never seen before. There was an emphais, (certainly not the only emphasis, but an emphasis nonetheless), placed on columns, like so:

from "A Shelfish Starter Guide to Databases"

I, frankly, had never considered coluimns in any way except, as 'fields' that contributed to rows, and that could have conditions placed on them. A column as a whole entity unto itself? I'd never considered such a thing.

A few days later though, while studying the netCDF format used by COSMIC2 missions among projects, I suddenly needed that column perspective, and I had it! netCDF files from COSMIC2 are very much arranged as coluimns. There's a dictionary entry for longitudes and a completely different dictionary entry for lattitudes. The entire dataset was built out of columns that, at the time of this writing, are assembled with each other merely by making them adjacent, and then accessing whatever data one needs by a pseudo-row index across the columns of interest.

So, reading little independent blogs made thinking about my new-ish project simple rather than the larger mental leap it would have been if I'd made the undertaking from a clean slate. Kinda cool!

It also led to an interesting though. Should there be a netCDF-extract plugin for datasette? 

 Image quality compare from Simon Willison and Claude!

One of the many aspects of Simon Willison's blog that I've enjoyed is the set of posts about coding tools with LLMs (AIs.) The latest one was handier than most for me. It takes an image and downsizes more and more, presenting the different version on a web page so you can judge which one will work best for your website's view while cutting down on the amount of data your web site serves for that image.

So, here's the faster version of this blog's occasional header

Chosen from a variety of options:

You might wonder if I went meta on this and used the tool to reduce the size of the screenshot of the tool, and I aboslutely did! 

Cool stuff!

Nine year-old Tawnse and I got half-way to activating the park before the ML-300 Bluetooth transmitter gave up the ghost. I knew I forgot to charge something. 

Tawnse was so entranced with the walled Philadelphia park she thought we should have stayed the extra half hour we would have probably needed to activate it. 

What led Tawnse to this scheduling priority decision? Turns out the park doubles  as the neighborhood dog park. There's only one way in or out though a small and, of course, historic  cemetery. By the time the pups got to the wander-around-unleashed bit we were in they were far to transfixed to try to leave. Turns out two of the pups were aspiring radio engineers to boot .We’ll get to that.

Because our antenna was low, propagation wasn't great on 20m at 15:30 UTC on the East Coast next to an interstate in Philadelphia. Even so right after I self-spotted a fellow ham immediately called in from South Dakota. After that, the QSOs came in every five minutes or so on average. We talked to Louisiana, two hams from Florida (whose QSOs came in back to back, and a ham from Tennessee. 

I'm curious if the QSOs from Florida might have come from the same physical station. I’ll find out when I map out the activation. In the meantime, I really, really need to think more about local-first data and how to download map tiles onto this device.  I remember that Simon Wllison has written about this more than once on his blog, so Im basically leaving a reminder to myself to go look that up. It would be nice to map things when I’m offline at least down to geographic regions like, lets say, cities?

The other folks who stopped by the park were very, very nice. Also, they had dogs. Tawnse and I met two daschunds, one of which was very fluffy, and two Golden Retrievers. All of whom introduced themselves at our picnic table.

Did I mention we had a low dipole placement? This would be the bit about our two radio-engineering dogs.

The Golden Retrievers immediately caught on to our dipole positioning being just to low, and they were certain they could fix it. They first sniffed out the rolls of tarred twine laying on the ground below each of the two trees that supported the dipole. They considered pulling on the twine. Perhaps I just hadn't really put my back into it. Then, they each decided it was probably a problem with the tree, and started to climb the tree following the twine. Until they remembered that they were dogs and heh, dogs don't climb trees. Abashed, one of them gave up. The other one though… That guy! He followed the antenna out of the tree till he saw TouCans which was only about six feet up, putting the dipole seven feet up. 

He. Could. Fix. This. He leapt at the radio to see if maybe he could adjust it. Fortunately, for all of us, his vertical leap was just a bit too short.

Tawnse just giggled and giggled and giggled. You gotta love engineering dogs.

Like I mentioned before, the Bluetooth transmitter's battery died right after the 5th QSO, and that was that. We were off to pick up KO6BTY from the library and get some of the best Italian food in the world from Paesanos on 9th St.

QSO map coming soon.

 The gang and I are very lucky to live in a San Francisco neighborhood where we have markets, a family owned pharmacy, and family-owned restaurants, coffee shops, and delis all about a ten block walk from the house. Here’s the haul from our Farmers Market that we walk to along with our local butcher shop and dim sum bakery a few weeks ago.

Consequently when I saw a mention of local-first software,[via] I was intrigued. (As a side note, the presenter, Maggie Appleton, at one point in the not to distant past, worked with Elicit, a research paper summarizing AI startup with an office in Oakland, so kind of surprisingly local, but I digress.) 

It turns out that local-first development advocates for keeping the data for an app offline, i.e. keeping the data with the person who created or is using that data by default. But, what happens to collaboration? Well, the local, offline data is synched to the cloud when a connection is available allowing for collaboration while also creating a bit of a task to keep everything synched.

Another aside: reading this local-first article turned me on to why my Bluetooth keyboard locks up when on airplanes. Google Docs synchs on a keystroke by keystroke basis. I have discovered this before as proven by a dim memory that I used to keep a local, (there’s that word again), html page on my personal devices. That page contained only a html text input control. I could free-type into that control at breakneck speeds, and then cut and paste blocks of text at a time into more formal location such as Google Docs.

Point being, local-first is a good idea, an idea I have believed in for quite a while as proven by implementation history if not by my knowledge of what to call it. I do love what its’s called though.

I’m slowly but surely working though what local-first looks like with respect to our QSO mapping apps. Here are some thoughts

• F2 data from ionosondes (probably good since it’s for a specific date, time, and location, and caching it at least momentarily will speed up our QSO mapping apps.)
• Elevation profiles (doesn’t seem useful yet since each QSO path is pretty unique)
• QSO physical addresses and locations (reduces geocoding calls and QRZ lookup calls)

Also of Interest:

 I finally landed at a quiet tape vs. noise  configuration for TouCans on Saturday afternoon. Here’s how.

In other parts of the project, the Pico-W has started burning through pairs of AA batteries rather quickly.

 Slowly but surely, KO6BTY, Tawnse and I are making our way back to the West Coast. Last night, we stopped in Quincy, MA for the flight we thought we were going to take this morning...

It looks like we are gonna call the East Coast our headquarters for a few more days, This has however, opened up POTA opportunities. Last night, I had the chance to activate Saint Pope John Paul II Park, US-8422, in Boston just across the Neponset River from where were staying in Quincy. The park is relatively new. It was opened to the public in 2001 and commemorates the Pope’s visit to Boston in 1979. It was constructed on the site of a landfill and a drive-in movie theatre.

Setting up near the river turned out to be a wash. The park is arranged so there is fairly thick vegetation between park denizens and the river.

I did, however get to set up about 10 yards back from the river in a pleasant little copse of trees. 

First though, I had to get the spools of twine up into the trees. For that, I was happy to find that there were plenty of branches under the hedge tree row

Here is the map. For the moment, it has F2 skips. Please ignore them, I will get a map without them in place soon.

Other US-8422 References:

The campground, where LobsterCon was held, Thomas Point Beach and Campground, was clean and well appointed, maybe too clean. There were no rocks laying about that would fit in the end of the tarred twine spools, so we wound up launching the antenna using pine cone stoppers. We had plenty of nice, flat ground for our two tents, and managed to get Project TouCans about 20 ft up between a pair of pine trees.

KO6BTY and Tawnse diagrammed TouCans on the nearby beach the first evening. Later that night, the rig reached KF9VV in Wisconsin.

TouCans is back up and running! The culprit did in fact lie in the path between the power supply and ground in the PA chain. I didn’t see it coming though. Here’s the problem:

Anyway, a few minutes later, we’d gone from this

To this

Which, in turn, led to this later that evening.

 We headed from Boston down to Philadelphia. Yeah, I know, that's no way to get to Maine, but when my partner and I were doing grad studies at Brookhaven on Long Island, Phillie was one of our favorite hangouts, so here we are!

With all the mapping Cesium has enabled of late, we were pretty tickled to get to visit their headquarters here in town. We got to demo our mapping tool that works with Datasette, and then got to ask questions about Cesium as well. Turns out the small object we frequently notice on the horizon is the Moon!

From there, we headed to Isgro near Christian and 10th. Armed with delicious pastries, we wandered up and down 9th St. checking all the other delicious things! A few hours later, Paesano's became our favorite place in town for pasta. We've had lunch there twice now, and just, oh my gosh, the food is so good!

We're on the East Coast! KO6BTY, myself, and the 9 year-old, (known on the internets as Tawnse), flew out to Boston from San Francisco yesterday. We walked out of the airport to the ferry terminal! That's so cool! You can walk to a ferry from the airport. But do you know what we did then? We didn't take a ferry, we took a water taxi to a different ferry terminal! So many cool things already!

Here's the view from the water taxi.

We landed at Rowes Wharf which seems to be one of the fanciest wharf's in Boston, so.. yeah. From there we caught the ferry southeast to Hingham 

and from there made it to our camp site at Wompatuck State Park.

We're now making our war around via public transit taking the T back in to Boston South Station to catch a train from South Station into Philadelphia.

Meanwhile! KO6BTY and I have had Cesium maps built into our QSO log for a few months now—ever since the day Simon Wilison nonchalantly pointed out that Datasette queries are URLs in an office hour we attended with him—but haven't put up a demo video! Here's a video of a few of KO6BTY's QSOs from one of our recent camping trips along with their associated F2 ionospheric skips.

 KO6BTY took over debug implementation last night. She wired the RF out from the Rockmite directly to Project TouCans antenna out and... Nothing.

Looking at the schematic, that leaves a few choices for what's going on.

My favorite for the moment, because it's easy, and because the part is actually very bent, is the T/R switch transistor:

The transistor is a 2N7000 MOSFET. Wikipedia lists its maximum current as 200 mA and I can see where we could have exceed that when the Rockmite was shaking loose in TouCans (several of the nylon spacers sheered after a fall.) Also, keep in mind that the power bump has more current traveling through this part of the circuit in any event.

After that, we'll be looking at whether or not the oscillator is still oscillating. But actually! Good news! If the oscillator weren't oscillating, the receive branch also wouldn't work, and it most certainly is working!

Since we returned from camping, Project TouCans has been pretty much off the air. We still have receive, a keyer, and a sidetone, but the rig just isn't transmitting.

Last week I thought the issue might be the output transistor of the RockMite. It is not. I removed the original transistor, replaced it with a new one, and to no avail! Here's the before

And after

There was no change in the operation of the rig. No signals were spotted by the RBN or either of the Utah or Half Moon Bay SDRs.

 Just a quick note. 

My partner and I went to see Robot Dreams with the 11 and 9 year olds about a week and a half ago. First, the movie is awesome! It stays almost true to the book in that there are very, very few words at all. More in the movie than there ever were in the book, but still. The whole thing was a lot of fun and I highly recommend it.

After the movie, my partner and I went off to do errands in one direction, the 11 and 9 year old headed  towards the house on BART. And! Guess what?

They ran into their older sib on the bus home. There routes coincidentally linked up for the last leg and they hopped on the sam bus she was on. About a stop later, they all realized it.

Public transit and independent kids are pretty awesome! We got to do our errands. The kids got to go do what they wanted, and they ran into each othere anyway! Without transit, I doubt we could have even seen the movie because we'd have been making camp pickups and whatnot. Woooo MUNI and BART!

 On one of the most interesting radio days of our recent camping trip, Project TouCans made QSOs with Australia, Japan, Columbia, and Argentina, all on the same day! The QSOs to Japan and Australia were made in the middle of their night. The Japan, Australia, and Columbia QSOs were all made in a sixteen minute window beginning with VK3YV at 12:40 UTC.

What was the Propagation Mode?

Chordal Modes Introduced with Villard

Carter Manny Jr. worked with Stanislaw. Here's his Chicago Tribune obituary.

Finding Our Antipodal Point

To find the anitpoidal point, we can follow our longitude over the North (or South for that matter) where it will become the same longitude minus 180 degrees, or pi radians if that's the unit you prefer. You can see this in the diagram below where our longitude of about -107 degrees traced over the pole becomes about 73 degrees.

Meanwhile, our latitude above the equator will be used to find the same number of degrees below the equator:

37.82275 becomes -37.82275.

More precisely, we get 

37.822754°N 107.717935°W -> 37.822754°N 72.282065°E

And our anitpodal is shown below near the 70 degrees East label.

Pretty excellent discussion of anitpodal points.

Conclusions for the moment