The detectors used for crystal radio sets include many rare, unusual, old, and even homemade devices. Most common are the few germanium diodes still in production and the remaining stock of old germanium devices, though some modern Schottky rectifiers and low-threshold MOSFETs are gaining in popularity. In addition, the traditional catswhisker mineral detector still has its uses, and other even stranger detectors can be pressed into service to demodulate radiofrequency signals.
- Cat's-whisker and Perikon Mineral Detectors
- Germanium Diode
- Tunnel and Back (Backward) Diode
- Thin-film Metal Oxide (Razor Blade)
- Low-threshold MOSFET
- Zero-bias FET
- Schottky Rectifier
- Light-emitting Diode (LED)
- Electrolytic Aluminum Rectifier
A detector may be viewed as a rectifier.
A more selective detector will have higher impedance at the lowest current sufficient to power your headphones; a line on an I/V graph will have a shallow slope near the origin.
Now, I measured 30 µA (average) being drawn from a strong local station on a mediocre antenna. The peak current of a sine wave through an ideal rectifier is about three times the average.1) Since detectors are less than ideal and enough current must remain in the tank to keep sufficient voltage oscillating, 100-200 µA seems a reasonable upper limit to the peak current through the detector. (With a better antenna, the maximum average current is as much as 150 µA at an average voltage of 0.4 V for 60 µW using a modern germanium 1N60.2) With a vintage black-band 1N34A (or a “mystery” yellow-band diode), 100 µA at 0.6V is still 60 µW at higher impedance.)
The shape of the curve will also effect the overtones produced by the detector when it demodulates the signal, but a sharp cutoff or a flat curve is not necessarily the best shape to cleanly demodulate the signal. If the I/V curve is flatter at lower voltage (concave upward), the resulting RF pulse will be more rounded off by the higher resistance at the beginning and end of the rectified half-wave being shaved off of the top of the original signal. This will reduce the high-frequency overtones produced. However, if the curve is too sharp the leading and trailing edges can be rounded too much, the half-wave become attenuated until it begins to resemble an impulse train, and overtones may even dominate the signal.
Another useful way of looking at the detector is similar to the function of a charge pump or a resonant transformer.
AC In AC+DC Out (~)--+--UU--+-->|--+--UUUU--o | :: | | :::: | | | +--||--+ +---||---+ | | --- --- - -
Even for modern devices, the data relevant to their use in unpowered radio sets is sparse and scattered. Perhaps the information provided here will help others in their pursuit of this rewarding crystal radio hobby.
Hypotheses to test:
- Is a detector's selectivity related to its average impedance to a typical crystal radio signal? Can this be predicted from the Shockley equation?3)
- Is a detector's sensitivity related to it's saturation current Is4)
|Package||Part||Type||IF(AV)||VR||Rated VF*||VF@20mA,25ºC||Measured VF†||CT|
|DO-7 Green Band||1N34A||Ge||0.050||75||>1.00||>1.00||0.382||0.8|
|DO-7 Red Band||1N34A||Ge||0.050||75||>1.00||>1.00||0.360||0.8|
|DO-7 Radio Shack||1N34A||Ge|
|DO-7 Black Band||1N34A||Ge||0.050||75||>1.00||>1.00||0.305||0.8|
|DO-7 Yellow Band||“Mystery”||Ge||–||–||–||0.293||–|
|DO-35 Orange with Green Band||1N60||Schottky||0.030||40||0.320 @1mA||–||2|
|DO-7 Gray Band||FO-215||Ge||0.276|
|DO-7 Brown, White, Red, Blue Bands||1N192||Ge||0.350|
|DO-7 Black Band||1N198||Ge||0.500||100||1.00||>1.00||0.278||low‡|
|DO-7 Black Band||1N277||Ge||0.500||110||1.00||0.335|
|DO-7 Opaque Gray||1N277B||Ge||0.310|
|DO-7 Black Band||1N270||Ge||0.325||80||1.00||<0.5||0.285|
|Large Glass Red and Green Bands||1N452||Ge||0.450|
|DO-7 Double-black Band||“Unknown”||Ge||–||–||–||0.364||–|
|DO-7 Opaque White, Black Band||OA47||0.320|
|DO-17 Gold||BD-3||Ge Back||-|
- *Usually @IF(AV), TJ=125ºC
- †Measured with cheap multimeter, probably at about 1.15 mA
- ‡Observed to be similar to 1N34A.
The purpose of collecting this data is to learn how to build a better crystal radio. To this end some mathematical analysis may be helpful.