The crystal radio gets its name from the galena crystal (lead sulfide) used to rectify the signals. A "cat's whisker" wire contact was moved about the surface of the crystal until a diode junction was formed. The 1N34A germanium diode is the modern substitute for galena and most other germanium small-signal diodes will also work well. Silicon diodes are not a good choice because their much higher barrier potential requires larger signals for efficient rectification. Certain silicon Schottky diodes with low barrier potential will work well but most small-signal Schottky diodes will not perform as well as a garden-variety germanium diode.
The circuit is quite simple but many pitfalls await the novice. The first precaution is most important! The crystal radio works best with a long, high outdoor antenna but the beginner may not fully appreciate the danger of bringing such a wire into the house. Lightning strikes to the antenna will probably destroy the crystal radio but if precautions are not taken, much more damage will result. The best strategy is to incorporate a commercial lightning arrestor with a straight, heavy gauge ground wire leading down to a buried water pipe. It is not sufficient to disconnect the antenna from the receiver during thunderstorms.
Other pitfalls are less dangerous and relate to the receiver's performance. A common mistake when building a crystal radio is to load the tuned circuit excessively. The Q of the tuned circuit must remain high to give selectivity or strong radio stations will all mix together. A good design will usually have low-impedance taps on the inductor for connections to the antenna and diode as shown in the schematic. A long wire antenna with a good ground connection will connect to the lowest impedance tap whereas a shorter antenna with no ground connection may connect to a higher tap. The diode may be experimentally moved to different taps and even across the whole coil for maximum sensitivity. The antenna and diode connection may be made with alligator clips for easy experimentation.
Another potential problem area is the earphone. Not all crystal earphones are sensitive and the experimenter should test a few to get a "good" one. High impedance dynamic earphones are a bit more reliable and can give excellent results. Try an old telephone receiver or a modern portable tape player headset (some are high-Z and fairly sensitive). Low impedance earphones like those used with many portable radios will not work at all. A simple test is to hold one earphone wire between the fingers while scraping the other lead across a large metal object. If static is heard in the earphone it will probably work well with the crystal radio.
The variable capacitor is often connected incorrectly. Make sure to connect the rotor to ground and the stator to the "hot" side of the coil. Otherwise, the radio will detune when the capacitor knob is touched. If detuning is noticed then try reversing the connections.
Some experimenters are tempted to omit the 82k resistor which discharges the capacitor on the theory that it wastes precious signal power. With a typical germanium diode, this little "improvement" may work somewhat but only because the diode has significant leakage and the performance will not be predictable. A dynamic earphone may be DC coupled eliminating the need for the resistor.
The coil may be wound on a 1.5 inch PVC pipe coupler as shown in the drawing. Drill two small holes at each end to secure the ends of the coil. The wire type is not particularly critical but select a gauge and insulation so that the 65 turns cover about 2/3 of the coupler. A "loopstick" coil may be used in place of the coil shown. These coils have an adjustable ferrite core for tuning so a fixed value capacitor may be used in place of the variable capacitor shown. The coil, capacitor and a terminal strip for the other parts may be mounted to a small wooden board. (See photo of receiver with transistor amplifier below.)
If a metal chassis is used then the coil must be mounted horizontally and above the metal to prevent unacceptable loading.
An amplifier may be added to boost the audio level as shown below. The current consumption of this amplifier is quite low and a power switch is not included. Disconnect the battery when the receiver is stored for long periods.
© 1995, Charles Wenzel
Note: You may use the transistor above as a sensitive detector eliminating the need for the 1N34A diode. Simply leave out the diode, the 0.001 uF, and the 82k resistor. Connect the negative side of the 1 uF directly to the coil. Change the base resistor from 10 meg. to 1 meg. and change the collector resistor from 100k to 10k. Now add a 0.01 uF from the collector to the emitter and the modifications are complete. This detector is quite sensitive and will be overloaded by very long antennas! Use a shorter antenna or a coil tap very near ground if significant distortion is noticed. The circuit draws about 1/2 mA.
Crystal Radio Audio Amplifier
Here is a simple audio amplifier using a TL431 shunt regulator. The amplifier will provide room-filling volume from an ordinary crystal radio outfitted with a long-wire antenna and good ground. The circuitry is similar in complexity to a simple one-transistor radio but the performance is far superior. The TL431 is available in a TO-92 package and it looks like an ordinary transistor so your hobbyist friends will be impressed by the volume you are getting with only one transistor! The amplifier may be used for other projects, too. Higher impedance headphones and speakers may also be used. An earphone from an old telephone will give ear-splitting volume and great sensitivity! The 68 ohm resistor may be increased to several hundred ohms when using high impedance earphones to save battery power.
To use the circuit as a general-purpose amplifier, apply the input signal to the top of the potentiometer. (Leave out the diode and .002 uF capacitor.) A higher value potentiometer may be used for a higher input impedance.
© 1995, Charles Wenzel
When building crystal radios or other simple receivers, the experimenter often wonders about the relative performance of the different diodes in the junk box. Here are the results of several experiments using the typical types available to the hobbyist. The types include germanium, silicon, Schottky, and even a light-emitting diode! The test setup uses an accurate RF synthesizer, a homemade AM pin diode modulator driven by an audio generator, a simple test fixture, a DC power supply for adding bias current, and a sensitive audio voltmeter. The setup shown below was used to test the diodes at several frequencies with a low modulation index (about 20%) and the near optimum bias current was determined by varying the DC supply.
Initially an RF level of about -15 dBm was used but this level was dropped to -25 dBm without much change in the relative results. The best performance was provided by an H-P (Agilent) diode, the 5082-2835 with a tiny 10 uA of DC bias. The -25 dBm (35 mV p-p) results are shown below using the 5082-2835 Schottky diode at 20 MHz as the reference point. The dB numbers are the audio level at the audio meter for the different RF carrier frequencies. Some variability is due to the test setup.
|Diode Part Number||DC Bias||20 MHz||60 MHz||100 MHz||130 MHz||Notes|
|5082-2835 Schottky||10 uA||"0" dB||0 dB||-2.5 dB||-4.5 dB||quite good!|
|1N5711 Schottky||10 uA||-0.5 dB||-0.5 dB||-2.0 dB||-3.5 dB||better at high freq.|
(similar to 1N914)
|20 uA||-8.5 dB||-9.5 dB||-10.5 dB||-11.5 dB||pretty bad!|
|1N277 (Ge.)||None||-3.0 dB||-4.0 dB||-6.5 dB||-8.5 dB||not great at high freq.|
|1N34A (Ge.)||None||-3.0 dB||-4.0 dB||-6.5 dB||-8.5 dB||.|
|1N32||10 uA||-1.0 dB||-1.0 dB||-3.5 dB||-5.0 dB||microwave diode|
|Red LED||10 uA||-4.0 dB||-4.5 dB||-8.0 dB||-11 dB||not bad for low freq.|
The Schottky diodes are the all-around winners if bias is used but they do not perform as well as a germanium diode without bias. Other small-signal Schottky diodes gave nearly the same results as the 1N5711. Other germanium diodes were tried but the results were nearly identical to those shown. (The 1N60 was not available for testing.) The 1N4454 is similar to other ordinary silicon diodes and the results were poor, as expected. Several LEDs were tried and a bright type red led gave fairly good results as shown. Zeners were tried with dismal results in both directions. Large rectifiers like the 1N4001 were similarly poor.
|The optimum bias current is tiny and a very small battery can be permanently wired into a "crystal" radio without a switch. A little photocell battery like those on tiny calculators could do the job.|
|Schottky diodes don't need bias when heated by a soldering iron! (Hardly practical information.)|
|Germanium diodes are hard to beat for low frequency crystal radios if no battery is used.|