DX Crystal Set Building Notes
In writing this page, my first thought was to call it
"conclusions". Conclusions is such a final sounding word and while these
are my conclusions, I better be careful about calling them final.
I have built several what I call "DX" or "high performance sets" (HP).
These include my #20,
and 70. In the past, if I heard a half
dozen stations in the earlier morning or late afternoon hours, I thought
I was doing fine. I just heard 18 stations after sun up on a September
morning. How did this happen? It happened because over the last couple
of years I have made improvements in every department, from antenna to
headphones. Below I will describe each section of a good dx receiving station.
I will use my #35 set as a basis for
my discussion as this is the set that
I have used quite a bit. It was also my
2004 Crystal Radio Contest entry.
This page won't be real technical. Building high performance
crystal sets is mostly picking the right components and using good
rf construction practices. I always keep in mind the technical
background when I design and build a set. I then use my experience
when it comes down to the actual design and construction.
A final warning .. Building a set like this is going to make
your wallet hurt like hell. Your Frau might even threaten to give you
the boot. Don't worry. Just do what you can.
DX Crystal Set Building Notes
Here is where it starts, folks. Try to make
your antenna as high and as long as possible. The wire should be thick enough to
stand up to your local weather conditions. Use insulators where the wire will be
connecting to another object. Some have used multiple wires strung about 12 inches
(30cm) apart. This is a "flat top" antenna. The feed line should be kept away from
other objects as much as possible.
In reading the old QST magazines, I came across several stories of hams
getting killed when their antenna systems came in contact with the power lines.
Folks, this doesn't have to happen if you just think! Think of the worse possible
scenario and put the antenna further away.
The ground is just as important as the antenna. I use my water service pipe that
comes into the house. It is metal and probably the best that I can do. I use a
piece of tv twin lead as the ground wire to my radios. The dual conductor gives
the illusion of a larger wire. All this minimizes noise and some of that pesky
short wave image problem.
Litz is the word here, friends. Litz = $. Big Litz = $$$$. The state of the art
is 660 strand, 46 gauge. To improve past this, you have to have many more strands.
I found that the 660/46 provides very good selectivity in a dual coil set. Under
reasonable conditions, I can separate two stations that are 10 khz from each
other in any part of the band. I recently made some
bandwidth measurements on my
#50 set, which also uses the 660/46 litz.
If you aren't quite this flushed with extra cash, 165 strand is still pretty good.
It is a quarter of the number of strands. Having litz that is 4 times as big
does not make it work 4 times better. But if you are really interested in the
best performance, the 660/46 is recommended.
There are three types of coil winding methods that you should consider.
Spider, rook and space wound cylinder coil. You can use what you feel most
comfortable with and how much space you have. The best material for coil
forms is styrene. Please don't use cardboard. Wood can work too if it is
sealed from moisture. Look at my
spider coil page for additional information.
work as well as spider coils. Cylinder coils are fine too, but the windings should
be spaced to increase the Q. Space wound
coils require more wire than close wound coils.
Rook coils work well with 660/46 litz but tend to be unstable with
165/46 litz. Additional supports made from styrene rod can help.
Try to make the inductance as small as possible but yet be able to
tune the entire band. I like to start with a slightly bigger coil and
check the bottom of the band. Then I remove a turn at a time so that the
tuning capacitor turns just past the bottom. This seems to give the radios
the best efficiency. Multiple coils can be used, but tapping the coil
is not recommended. A contra wound coil
has been recently found to provide very good performance.
The coils should be mounted away from other objects as much as
possible. I would avoid the use of metal boxes. Wood boxes are fine but please
give the coils a lot of breathing room.
When building a two coil set, make sure that you allow for variable
coil coupling. This can be done by using separate boards for the antenna tuning
and detector tuning as in my #35 set,
or turning the coil axis as in my #48 set.
Having an over coupled circuit will lower the sensitivity and selectivity.
Having an under coupled coil will lower the sensitivity. There is a sweet spot
depending on your conditions and where in the band you are listening. The
best coupling is never at one position. I use a coil spacing between
12 and 18 inches (30 - 45 cm) under normal conditions. If you use smaller
litz, the distance will be less. It kind of shows you how well the big litz
Some use toroidal coils quite effectively. My experience with toroids
is very limited. I would recommend that you talk with people from the
BAMA group. Some of the members have brought toroids to a new high level.
The selection of a good capacitor is as important as using large litz. The Q of
the entire coil - capacitor circuit seems to settle to it's weakest point.
The capacitor should be an
air type, and the insulators should be made of ceramic. This is a minimum. Other
important factors are the quality of the wiper arm. Silver
plating is also a positive.
In circuits where the frame is not grounded, the capacitor must be
isolated from the panel. Placing your hand near the capacitor (by grabbing
the knob) will detune the set. This is very annoying. By placing the capacitor
an inch or two behind the panel and using an insulated shaft, the capacitor will
work it's best. Don't forget to use ceramic or styrene insulators when attaching
the variable to the radio base. Any leakage will sap the sensitivity of your
Vernier Drives and Dials
Building a dx set that is very selective presents a new problem. If you try to
tune all 115 U.S. broadcast channels in a 180 degree dial rotation, you will miss
stations. Tuning requires minute tuning capacitor adjustments. A 6:1 vernier drive
will give you three full turns of the knob to cover the band. Low backlash and binding
are also desirable. All variable capacitors should have them.
I like the detector that is known as the "Hobbydyne". The original
Australian design used a small coupling variable capacitor while the
Hobbydyne uses a differential capacitor. Please look below for the
section called Selectivity Enhancement Circuits.
Make sure that connections that are on the tank coil side of the
diode are isolated on ceramic insulators or styrene.
In a very high impedance tank circuit, a standard 1N34A diode just won't cut
the mustard. The diode will load the circuit, causing loss in selectivity and
sensitivity. 1N34A diodes are alright for low and medium performance sets.
There seems to be two "best" types of diodes I have used. A schottky
(or several in parallel) and the ITT FO-215 diode. A good schottky diode
is the Agilent HSMS-286* series diodes. They are available in several configurations.
They are surface mount devices but wires can be soldered to them.
I found that having a selector switch (use a ceramic or styrene switch)
is very useful as the type of diode used depends on the reception circumstances
and the portion of the band that is being tuned. In general, I use the FO-215 diode
at the bottom of the band up to around 1300 khz. Then I switch to the schottky
diode for the rest of the band. The schottky is the best for selectivity and
the FO-215 for sensitivity. Selectivity isn't a big difficulty at the bottom
of the band, but I need all the sensitivity I can get.
Audio Transformer Matching
Magnetic type high impedance headphones have an impedance range from about
1000 to 4000 ohms. Sound powered phones are 100 to 1000 ohms.
Crystal earphones are around 10,000 ohms or so. The crystal
set impedance after the diode can be several hundred thousand ohms.
If you connect one of these headphones directly to the detector output,
you can have a giant audio mismatch. This will reduce the volume. Also
the load will destroy that great selectivity that you spent your hard
cash getting the big litz and great capacitors.
An audio matching transformer is essential in any high performance
crystal set. The old style very high efficiency transformers are rare and
expensive. Early in 2007, I had some custom audio transformers made.
They have a very high impedance and several low impedance taps.
Another good alternative is the Bogen T725 transformer. You will get
good performance from one of these in a medium design crystal set.
Ramon Vargas and I have prepared some information on the
The ear interface is no less important than the rest of the set. The most
sensitive headphones are the so called "sound powered" headphones.
Darryl Boyd has done a lot of research on this subject and has presented
it on his crystalradio.net site.
These phones are available all the time on eBay. This can be
an expensive item too. They have to be carefully adjusted most of the
time. Regular magnetic phones generally aren't sensitive for most dx
situations. My second choice is the little crystal ear piece, or two
Make sure your ear canals are clean. Please check with your health
professional concerning this.
Selective Enhancement Circuit
I would like to talk about selectivity enhancement circuits (SEC). Some
refer to these as the "Hobbydyne". Jim Frederick in Florida came up with the
Hobbydyne circuit. The earliest publication of a SEC I have seen was in
Australia in the early thirties.
Jim Frederick's Site
Here is the Australian page.
So you see that what appears to be new, sometimes isn't. But Jim's Hobbydyne
was new. Also his first (to my knowledge) recent use of the SEC is credited to
The SEC and Hobbydyne do the same thing. They allow unloading of the tank circuit,
thus increasing the loaded Q, which means better selectivity. It also provides a
better match between the tank and diode (and audio circuits further down the line).
This is better than the old fashioned way of matching and unloading - tapping the
The basic SEC consists of a small value variable capacitor (usually around 20-30 pf),
and a small RF choke. I found that a 27 millihenry value is good. Some have used
a reverse biased diode in place of the choke. The choke or diode is to complete a
DC path so the detector can work properly.
Jim discovered this feature while experimenting with his crystal sets in 2003. He
sent me an ie-mail and invited me to test this circuit. I tested it on my, at that
time current dx receiver, my #35 radio.
It didn't take me very long to see the value in this discovery. I knew that this was
going to be a regular part of my dx crystal radio designs, so I looked around for
some inexpensive chokes. I found some 27 mh chokes and found they worked great.
The next development from Jim was his Hobbydyne circuit. Using the SEC described
above had the issue of the resonant frequency shifting when the selectivity was
changed. It is easy to understand as the capacitance was being changed and this
affected the tuning. Jim's solution was to add a differential capacitor. A differential
capacitor is a variable capacitor with one rotor and two stators. As the capacitance
on one side increases, the other side decreases. This part was used mostly in the
high end variable oscillators for better stability.
Jim later made an enhancement to his basic hobbydyne by adding a small trimmer
on the ground leg of the differential capacitor. This allowed for better tracking
across the differential capacitor range.
I found there are two Hobbydyne hookups. One is what Jim came up with and the other
is what I came up with. I don't claim any discovery here, it was most likely a wiring
error. But, isn't that how some everyday items were discovered?
Below are some circuits showing various SEC configurations. Figure (a. is a
directly connected diode to the top of the tank. This is the worst possible way to
directly connect a diode for most builders. Only if a high impedance load is used
at the output, will this method work well.
Otherwise, you can tap the coil and connect the diode to that
tap. That way you won't end up with a bad mismatch and severe tank loading. This causes
loss of sensitivity and selectivity.
Figure (b. is a basic SEC. Just a small trimmer capacitor and a dc path rf choke.
This is a good place to start, especially if you don't have a differential capacitor.
The rest of the figures show various SEC and the Hobbydyne connections. It is kind
of a take-your-pick situation. Figure (e. has the additional tracking
capacitor. That is recommended if you can spare the capacitor.
Another hookup would be for you to use two small capacitors in a quasi
differential capacitor. You can manually increase one capacitor while decreasing
the value of the other one. This will give you that capacitive divider effect.
It should be easy to take your existing crystal set hookup and a SEC.
Last summer, after building my
#64 contra coil radio, I was asked to try a direct
top of the tank connection, then switch back to the Hobbydyne. Since I hadn't made
the comparison for nearly three years, I thought it would be interesting to see
if the SEC circuit really helped. I ran some listening tests at the high end of the
band and I was quickly convinced to keep the SEC.
These people and their websites were very useful for preparation of
Ben Tongue has done much
high level research on crystal set design and presents them in nearly 30
papers. The papers are updated frequently. This site is the basis for
my designs. It is quite technical so be prepared to read each several times.
Darryl Boyd's crystalradio.net
has great information on sound powered headphones. Also on his site is the
full description of Mike Tuggle's
Lyonodyne high performance crystal set. My #35 and 48 sets are based on
crystal radio page, featuring the Hobbydyne circuit. I would say that Jim's
circuit is the best detector for a high performance radio.
Ramon Vargas has provided
great technical articles on how the Tuggle Tuner and Bogen T725 transformers
Gary Johanson, WD4NKA for coming
up with the Tubepad schematic diagram drawing system.
There are other sites and people that have given me excellent information
and hints over the last couple of years.