Antenna 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.

Ground 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.

Coils 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.

Rook coils 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 works.

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.

Variable capacitors 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 dx radio.

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.

Detector Circuit 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.

Diodes 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 F215 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 F215 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 F215 for sensitivity. Selectivity isn't a big difficulity 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 T725 transformer.

Headphones 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 of them.

Ears Make sure your ear canals are clean. Please check with your health professional concerning this.

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 Jim.

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 coil.

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 caapcitor 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 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 fopreparationation of this page.

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 this design.

Jim Frederick's 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 operate.

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.