Finding Your Best Crystal Radio 'DX Diode'
![]() |
![]() |
SMS7630 Schottky Vf = .147V Id = 9uA Vdx = 61 |
![]() |
'95481' Vf = .246V Id = 13uA Vdx = 53 |
![]() |
Vf = .252V Id = 12 Vdx = 48 |
![]() |
Vf = .335V Id = 13uA Vdx = 39 |
![]() |
![]() |
SMS7630 Schottky Vf = .147V Id = 9uA Vdx = 61 |
![]() |
'95481' Vf = .246V Id = 13uA Vdx = 53 |
![]() |
Vf = .252V Id = 12 Vdx = 48 |
![]() |
Vf = .335V Id = 13uA Vdx = 39 |
The first time I was involved with a crystal radio DX contest was about 20 years
ago when I built a well-performing crystal receiver for the Yahoo Crystal Radio Group's
annual DX contest. It was a great learning experience and taught me much about
circuit losses and how to overcome them. I originally built several sets but was unable to
hear anything other than local stations until I eventually figured things out ... the system was only as good as its weakest link or links!
Fast
forward to more recently when I obtained and wrote about the Heathkit CR-1 Crystal Radio, a simple but very well-designed tuner that has become popular
with collectors. Using the CR-1 re-kindled my interest in the DX contest
activity of years ago and when talking with two other amateurs that had an
interest as well (one had also been in the earlier contests sponsored by the now defunct Alabama Crystal Radio Group), we decided to bring the contest back once again.
The Facebook Crystal Radio DX Contest Group was formed last fall, a set of rules drawn up and the contest date set for the first week of January of this
year. This gave interested participants plenty of time to build something
they could use in the contest.
I spent all of December designing and
constructing a new contest radio, hopefully one with enough selectivity to get around
the 15 local flamethrowers (10-50kW S9++ signals) that plague the band for me
and eventually drove me from crystal radio activities.
The new radio
makes use of several 'traps' to null strong signals ... two are in the antenna
line while one is loosely coupled inductively to the detector tank circuit. The
two inline trap coils are wound with Litz wire on ferrite toroids (R40C1) while the
third is a basket-wound Litz coil (660/46) on a 4" diameter form.
The
antenna tuning stage also uses the same ferrite material but in the rod / bar
form. I wound a low-end as well as a high-end coil for the tuner using the same
high-count Litz as on the big trap coil. The low-end coil is wound on a bundle of three rods while the high-end coil uses a single rod.
![]() |
Antenna tuner |
![]() |
Detector stage |
The detector also has provisions for comparing various diodes as not all diodes are created equal ... not even all diodes with the same number! When testing and comparing diodes of the same type such as the popular 1N34 germanium, every once in awhile one of them will turn out to be noticeably more sensitive than the others. In my built-in B-A-C diode test module, the hottest diode is always mounted in middle-position A, making it easy to quickly compare by switching to the left for B or to the right for C. So far the best one I have found is the vintage Russian D18 germanium diode but an old 1N34 removed years ago from a 1950s-era Heathkit has given it a good run for the money! I've still several hundreds of early germanium diodes, pulled from old diode matrix boards years ago, to test against the D18 as well as numerous Schottky diodes.
Also on board the detector module is a Selectivity Enhancement Circuit (SEC) that increases selectivity by unloading some of the diode's effect on the detector coil, similarly to tapping the diode further down the tank coil. It uses a small butterfly capacitor seen to the right of the main tuning capacitor in the photo above. I found it extremely effective when needed and is well worth the addition to a high-performance tuner.
The detector stage is followed by an impedance-matching transformer for the sound-powered headphones. This stage also houses a 50uA meter to measure diode current / signal strength levels.
The meter can be switch-bypassed to prevent needle-bounce on stronger signals. It is particularly helpful when using the traps to null a signal to the minimum level.
The three traps utilized have been very effective in eliminating what I had originally perceived as an impossible DXing situation.
Here are the daytime-power signal strengths of my 15 line-of-site blowtorch stations that, without trapping, very effectively block most sections of the band. Anything over 50uA is ear shattering and problematic, usually requiring the use of all 3 traps:
KVRI 1600 50uA
KRPI 1550 100uA
CJVB 1470 40uA
CFTE 1410 350uA
CHMB 1320 100uA
CJRJ 1200 400uA
CKWX 1130 300uA
CKST 1040 90uA
CKNW 980 150uA
KGMI 790 100uA
CHMJ 730 450uA
CBU 690 650uA
CISL 650 200uA
CKSP 600 100uA
KARI 550 100uA
Overall I was very pleased and surprised at the good performance of the new radio. During the contest period I identified and logged 92 unique stations in 16 states / provinces. More than one station was logged on 9 different frequencies as the propagation varied from night to night.
Highlights of the DX Contest were hearing WHAS in Kentucky (2,007 miles), WJR in Michigan (1,970 miles), KXEL in Iowa (1,556 miles), WCCO in Minnesota (1,423 miles) and CBW-990 in Winnipeg, smack up beside local blowtorch CKNW-980! Additionally, hearing Washington state 250 watter KFLD-870 and 250 watt KWBY-940 in Oregon were great surprises.
I found the use of a spotter radio (Sony ICF-2010) to be very useful in locating signals to target and to zero-beat with an RF signal generator. The generator’s tone-modulated signal can then be tuned in and the xtal radio and antenna / detector stages optimized.
From here, any pest signals are then tuned to and individually nulled using the traps while watching the signal meter. Antenna and detector stages are then re-tweaked before disabling the generator and listening for the desired signal.
Often it is heard immediately following the above tuning procedures but if not, monitoring the frequency for several minutes often allows time for the weak signal to fade up to audible levels.
Comparing programming audio with what is heard on the spotter radio will confirm hearing the correct signal as will comparing audio to the station’s own live-feed on the internet.
Due to the larger and much better antenna (inverted-L 70’ x 100’) on the crystal radio, I would often hear good audible signals on it and not on the spotter (something that I found surprising) so often times it was productive to just tune around the band on the crystal radio, tweaking stages as required.
I’m looking forward to further improvements of the tuner as well as to the next DX Contest whenever that will be scheduled ... hopefully you can join in as well!
FREQ UTC STN LOCATION MI