FM Crystal Radio Receivers
Radio, 2002, 7
The notion of "crystal radio" is strongly associated with huge antennas and radio broadcasting on long and medium bands, in this article, the author describes the experimentally tested detector circuits of VHF receivers designed to listening to a FM stations.
The very possibility of receiving VHF FM detector was discovered accidentally. One day I was walking in the Terletskiy park in Moscow, Novogireevo, I decided to listen to the broadcast - I had a simple crystal set without resonant tank (this circuit is described in the "Radio", 2001, N 1, Fig. 3). The receiver had a telescopic antenna with length of about 1.4 m. Wonder whether it is possible to receive radio broadcast with this short antenna? It was possible to hear, but weakly, simultaneous operation of two stations. But what is surprised me is the volume of receiving was rise and fall periodically almost to zero after every 5...7 m, and it was different for each radio station!
It is known that in the LW and MW bands, where the wavelengths are hundreds of meters, it is impossible. I had to stop at the point of receive with maximum volume of one of the stations and listen attentively. It turned out - this is "Radio Nostalgie", 100.5 MHz, broadcasting from the near city Balashikha. There were no line of sight between antennas. How does the FM transmission could be received by using the AM detector? Further calculations and experiments shows that it is quite possible and is not depends on the receiver.
A simple portable FM crystal receiver is made exactly the same way as an indicator of the electric field, but instead of measuring device it is necessary to connect a high-impedance headphones. It makes sense to add an adjustment of coupling between the detector circuit and the resonant tank to adjust the maximum volume and quality of the receiving signal.
The simplest Crystal radio
The circuit diagram of the receiver suitable for these requirements is shown in Fig. 1. This circuit is very close to the circuit of the receiver mentioned above. Only the VHF resonant tank has been added to the circuit.
Fig. 1.
VD1, VD2 - GD507A - an old USSR Germanium high-frequency diodes with the capacitance of 0.8 pF (at the reverse voltage of 5V), the recovery time of reverse resistance is no more than 0.1 uS (at the Idirect pulse=10 mA, Ureverse pulse=20 V, Icutoff=1 mA)
The device contains a telescopic antenna WA1, directly connected to the resonant tank L1C1. The antenna is also an element of the resonant tank, so to get the maximum power of the signal it must be adjust both the length of the antenna and the frequency of the tank circuit. In some cases, especially when the length of the antenna is about 1/4 of the wavelength, it is useful to connect the antenna to a tap of the tuning coil L1 (find the suitable tap of the coil by finding the maximum volume of the signal).
The coupling with the detector can be adjust by trimmer C2. Actually the detector is made of two high-frequency germanium diodes VD1 and VD2. The circuit is completely identical to the voltage doubling rectifier circuit, but the detected voltage would be doubled if only the trimmer capacitor C2 value is high, but then the load of the resonant circuit L1C1 would be excessive, and its quality factor Q will be low. As a result, the signal voltage in the circuit tank L1C1 will be lower and the audio volume will be lower too.
In our case, the capacitance of the coupling capacitor C2 is small enough and voltage doubling does not occur. For optimal matching the detector circuit with the tank circuit the impedance of the coupling capacitor must be equal to the geometric mean between the input resistance of the detector and the resonant resistance of the tank circuit L1C1. Under this condition, the detector is getting the maximum power of the high-frequency signal, and this is corresponding to the maximum audio volume.
The capacitor C3 is shunting the higher frequencies at the output of the detector. The load of the detector is headphones with the dc resistance of not less than 4K ohms. The whole unit is assembled in a small metal or plastic housing. The telescopic antenna with the length not less then 1m is attached to the upper part of the housing, and the connector or the jack for the phones is attached th the bottom of the housing. Note that the phone cord is the second half of the dipole antenna (a counterweight).
The coil L1 is frameless, it contains 5 turns of enameled copper wire with diameter of 0.6...1 mm wound on a mandrel with diameter of 7...8 mm. You can adjust the necessary inductance by stretching or compressing the turns of the coil L1. It's better use the variable capacitor C1 with an air dielectric, for example, type 1KPVM with two or three movable and one or two fixed plates. Its maximum capacity is small and can be in range of 7...15 pF. If the variable capacitor has more plates (the capacitance is higher), it is advisable to remove any of the plates, or connect the variable capacitor in series with a constant capacitor or a trimmer, it will reduce the maximum capacity.
The capacitor C2 is ceramic trimmer capacitor, such as a KPK or KPK-M with the capacity of 2...7 pF. Other trimmers capacitors could be used too. The trimmer capacitor C2 can be replaced with a variable capacitor, similar to C1, and it could be used to adjust the coupling "on the fly" to optimize radio receiving capabilities.
Diodes VD1 and VD2, can be GD507B, D18, D20 (it is old USSR Germanium high-frequency diodes. This diodes can be replaced with modern Schottky diodes). The shunting capacitor C3 is ceramic, its capacity is not critical and can have a value in range from 100 to 4700 pF.
Adjustment of the receiver is simple. Tune the radio by turning the knob on the variable capacitor C1 and adjust the capacitor C2 to get the maximum audio volume. The tune of the resonant tank L1C1 will be changed, so all operations must be repeated a few more times, and at the same time find the best place for the radio receiving. It is doesn't necessarily the same place where the electric field has maximum strength. This should be discussed in more detail and explain why this receiver can receive FM signals.
Interference and conversion of FM into AM
If the tank circuit L1C1 of our receiver (Fig. 1) will be set up so that the carrier frequency of FM signal falls on the slope of the resonance curve, the FM can be converted into AM. Let's find the value of Q of the tank circuit. Assuming that the bandwidth of the tank circuit L1C1 is equal to twice the frequency deviation, we obtain Q = F0 /Δ2f = 700 for both the upper and the lower VHF band.
The actual Q of the tank circuit in a crystal radio probably will be less than 700 because of the low Q-factor of its own Q (About 150...200) and because the resonant tank is shunted by the antenna and by the input impedance of the detector. Nevertheless, a weak transformation of FM into AM is possible, thus, the receiver will barely work if its tank circuit detune a little up or down in frequency.
However, there is much more powerful factor contributing to the transformation of FM into AM, - it is an interference. It's very rarely when the receiver is in the line of sight of radio station, in most cases the line of sight is obscured by buildings, hills, trees and other reflective objects. A few radio beams scattered by these objects comes to the antenna of the receiver. Even in the line of sight to the antenna comes some reflected signals (and of course, direct signal comes too). The total signal depends on both the amplitudes and phases of summing components.
The two signals are summed if they are in phase, i.e., the difference of their ways is multiple of an integer of the wavelength, and the two signals are subtracted if they are in opposite phase, when the difference of their ways is the same number of wavelengths plus half wavelength. But the wavelength, as well as the frequency varies at FM! The difference of the beams and their relative phase shift will vary. If the difference of ways is large, then even a small change in frequency leads to significant shifts in the phases. An elementary geometric calculation leads to the relation: Δf/f0 = λ/4ΔC, or ΔC = f0/λ/4Δf, where ΔC - the difference of the ways of the , it's required for the phase shift ±Π/2, to get the full sum of AM signal, Δf - frequency deviation. The full AM is the total variation of the amplitude signal from the sum of the amplitudes of the two signals to their difference. The formula can be further simplified if we consider that the multiply of frequency by the wave length f0λ is equal to the speed of light c: ΔC = c/4Δf.
Now it is easy to calculate that to get a full AM of the two-beam FM signal, the sufficient difference between the ways of beams is about a kilometer. If the difference of ways is smaller, the depth of AM proportionally decreases. Well, but if the difference of ways is more? Then, during one period of the modulating audio signal the total amplitude of the interfering signal will pass several times through the highs and lows, and distortion will be very strong when converting FM into AM, up to complete indistinct of the sound when you receive the FM by an AM detector.
Interference with FM broadcast reception is an extremely harmful phenomenon. It is not only produces a concomitant parasitic AM of a signal, as it is described above, but it is produces the parasitic phase modulation, what leads to distortion even if we got a good FM receiver. That's why it is so important to place the antenna in the right location, where the only one signal prevails. It is always better to use a directional antenna, because it increases the magnitude of the direct signal and reduces reflections coming from other directions.
Only in this case with a very simple detector radio receiver the interference played a useful role and allowed us to listen to the radio broadcast, but the radio broadcast can be heard weakly or with significant distortions, and the radio broadcast can't be heard everywhere, but only in certain places. This explains the periodic changes in the volume of the radio broadcast in the Terletskiy park.
Crystal Detector Radio Receiver with a frequency detector
A radical way to improve reception is to use a frequency detector instead of an amplitude detector. In Figure 2 is shown a circuit of a portable detector radio receiver with a simple frequency detector, based on a single high-frequency germanium transistor VT1. The germanium transistors is used because it's junctions works at a low voltage about 0.15 Volts, this allows to detect very weak signals. The junctions of silicon transitions works at a voltage approximately 0.5 V, and the sensitivity of the receiver with a silicon transistor is much lower.
Fig. 2.
VT1 - GT313A - an old USSR Germanium high-frequency transistor with hfe=10...230 (at DC: Uke=3 V, Ie=15 mA), hfe=3..10 (at f=100 mHz, Ukb=5 V, Ie=5 mA)
As in the previous design, the antenna is connected to the input tank circuit L1C1, the variable capacitor C1 is used for the tuning function. The signal from the input tank circuit goes to the base of the transistor VT1. The other tank circuit, L2C2, is inductively coupled with the input tank circuit L1C1. The tank circuit L2C2 is tuneble with the variable capacitor C2. Because of the inductive coupling between this two tanks the oscillation in the resonant tank L2C2 is phase shifted by 90° relative to the signal across the input tnak circuit L1C1. From the tap of the coil L2 the signal goes to the emitter of the transistor VT1. A bypass capacitor C3 and high impedance headphones BF1 is connected to the collector of the transistor VT1.
The transistor begins to turn on when its base and emitter has the positive half-wave of the signal, and the instantaneous voltage on the emitter is greater then its base voltage. At the same time the smoothed detected current passes through the headphone in the collector network. But the positive half-wave of the signal is only partially overlapping when the phase shift of the signal is 90° in the resonant tanks, so the detected current reaches the maximum value determined by the signal level.
With frequency modulation, depending on the frequency deviation, the phase shift is also changing, corresponding to the phase-frequency response of the tank circuit L2C2. When the frequency deviates in one direction then the phase shift decreases and the half-waves of the signal at the base and emitter is overlapped more, as a result, the detected current increases. When the frequency deviation goes in the opposite direction, its decreases the overlap of half-waves of the signal and the current decreases. So the frequency detection of the signal occurs.
The gain of the detector depends directly on the quality factor Q of the resonant tank L2C2, the quality factor Q should be as high as possible (in the limit of 700, as we calculated earlier), therefore the coupling with the emitter of the transistor is weak. Of course, such a simple detector does not suppress the AM of the received signal. In fact, its detected current is proportional to the signal level at the input, this is an obvious disadvantage. But anyway it's the very simple circuit.
Just like the previous circuit, the receiver is built in a small housing, on the top of the housing a telescoping antenna is mounted, and the headphone socket in the bottom the housing. The knobs of the variable capacitors is located on the front panel. These variable capacitors should not be combined into one unit, because a louder volume and a better quality of reception can be obtained with separate tuning.
The coils L1, L2 if frameless, they wound with the copper wire 0.7 mm (AWG 21) in diameter on the mandrel of diameter 8 mm. L1 contains 5 turns, L2 - 5+2 turns. If possible, the coil L2 wound with silver plated wire to improve the quality factor Q, the diameter of the wires is not critical. The inductance of the coils is adjusted by compressing or stretching of the coils L1 and L2 to get the FM radio stations in the middle of the variable capacitors tuning range. The distance between the coils L1 and L2 is in the range of 15...20 mm (the axis of the coils is parallel), the distance is adjusted by bending their terminals, soldered to the variable capacitors.
With this receiver can be done a lot of interesting experiments, exploring the possibility of reception of VHF radio broadcasts with the detector receiver, exploring the propagation of radio waves in urban areas, etc. Can be done experiments to further improve the receiver. However, the sound quality in a high-impedance headphones with membranes is poor. Because of it a better receiver was developed, which provides better sound quality and allows you to use a different external antennas, connected to the receiver by feedline.
Radio receiver powered by the energy of radio waves
Experimenting with a simple crystal radio set, repeatedly had to make sure that the power of the detected signal is sufficiently enough (tens or hundreds of microwatts) to provide a very loud sound in the headphones. But the quality of reception is not good because there is no frequency detector. This problem is partially solved in the second receiver (Fig. 2), but the signal strength is also used inefficiently because the transistor is powered by quadrature high-frequency signal. Therefore it was decided to use two detectors in the receiver: the envelope detector - to power the transistor, and the frequency detector - to improve signal detection.
Fig. 3. C1, C2 - 2.2...15 pF, C3 - 0.15 uF, C4 - 1 uF, C5 - 1 nF, R1 - 130 k
The circuit diagram of the receiver is shown in Fig. 3. An external antenna (dipoles) connected to the receiver by a two-wire line, made of ribbon VHF cable with the impedance of 240...300 ohms. The impedance matching between the cable and the antenna is performed automatically, and the impedance matching of the input tank circuit L1C1 is performed by selecting a suitable tap of the coil L1. Generally speaking, unbalanced connection of the feeder to the input tank circuit reduces the noiseproofing of the antenna feeder system, but because the low sensitivity of the receiver, it doesn't matter. There is a well-known methods of balanced connections for a feeder with the use of a coupling coil or a balun.
The author's folded dipole was made of a conventional isolated connecting wire, the dipole was placed on the balcony, in a place with a maximum field strength. The length of the feeder does not exceed 5 m. With such a small length the losses in the feeder is negligible, and therefore, the balanced line can be successfully used.
The input tank circuit L1C1 is tuned to a frequency of a signal, and a high frequency voltage across L1C1 is rectified by an amplitude detector, based on the high-frequency diode VD1. Since the amplitude of FM signal has a constant value, there is practically no requirements for smoothing the rectified DC voltage. However, to remove possible parasitic amplitude modulation in case of multipath propagation of radio signals (see above story about the interference), the capacitance of the smoothing capacitor C4 is selected sufficiently large. A rectified DC voltage is used to power transistor VT1. For the control of the current consumption and for a signal level indication is used an analog current meter PA1.
A quadrature frequency demodulator of the receiver is implemented with the transistor VT1 and phase shifter tank circuit L2C2. The high-frequency signal from the tap of the coil L1 is applied to the base of the transistor VT1 through the coupling capacitor C3, and it's signal is applied to the emitter of the transistor VT1 from the tap of the coil L2 of the phase-shifting tank circuit L2C2. The work of the detector is exactly the same as in the previous design. To increase the gain of the frequency demodulator, on the base of the transistor VT1 is applied an offset voltage through the resistor R1, and because of it the coupling capacitor C3 is used. Note that the capacitor C3 has sufficient capacitance (0.15 uF) - this capacitance is chosen to shunt the low-frequency currents, i.e., for grounding the base of the transistor VT1 for the sound frequencies. This increases the gain of the transistor and increases the volume of reception.
The primary winding of the output transformer T1 in the collector circuit of the transistor VT1 is used to match the high output impedance of the transistor to the low impedance of the headphones. A stereo headphones TDS-1 (8..16 ohms) or TDS-6 (8 ohms) can be used with this radio. Both the earpieces (left and right channels) are connected in parallel. The bypass capacitor C5 is used to filter the high-frequency currents in the collector circuit. The button SB1 is used to short the collector circuit of the transistor VT1 while tuning the input tank circuit and the search for a signal. The sound in the headphones at the same time disappears, but the sensitivity of the indicator PA1 is significantly increased.
The design of the receiver can be very different, but anyway it needs the front panel with the knobs of the two variable capacitors C1 and C2 (each capacitor has individual knob) and the button SB1. To reduce hand effect on the tuning, it is desirable to make the front panel of a metal plate or a copper clad laminates. It can work also as a common wire of the receiver. Rotors of the variable capacitors should have good electrical contact with the panel. The antenna socket X1 and the phone jack X2 can be placed either on the front panel or on the side or back of the receiver. Its dimensions are dependent on the available components. So let's say a few words about them.
The capacitors C1 and C2 is KPV type with a maximum capacity of 15...25 pF. The capacitors C3-C5 are ceramic.
The coils L1 and L2 are frameless (see Figure 4), wound on a mandrel of diameter 8 mm, L1 contain 5, L2 contains 7 turns. The length of the winding is 10...15 mm (do some tuning by adjusting the length). The enameled copper wire of 0.6...0.8 mm (AWG 20..23) is used, but it is better to use a silver-plated wire, especially for the coil L2. The taps are made from 1 and 1.5 turns (L1) and from 1 turn (L2). The coils can be arranged coaxially or axis parallel to each other. The distance between the coils (10...20 mm) is adjusted. The receiver will work even in the absence of inductive coupling between the coils - the capacitive coupling through the junction capacitance of the transistor is enough. The audio transformer T1 is TAG-3, it has a winding ratio of 10:1 or 20:1.
Fig. 4.
The transistor VT1 can be replaced by any germanium transistor with maximum operating frequency ft not lower than 400 MHz. A p-n-p transistor can be used too, for example, GT313A, in this case the polarity of the indicator PA1 and the diode VD1 should be reversed. The diode can be any germanium type, a high-frequency. As the indicator PA1 any ammeter with a current range of 50..150 mA can be used.
Tune the tank circuits to the frequency of a radio station, adjust the taps of the coils and the distance between the coils to get the best result (maximum volume and best quality of the reception). It is useful to adjust the value of the resistor R1 for maximum volume.
On the balcony the receiver with the antenna described above provided high quality reception of two stations with the strongest signal from the radio center at the distance not less than 4 km and with no direct line of sight (obscured by buildings). Collector current of the transistor was 30...50 mA.
Of course, the possible design of VHF crystal radios is not limited to described above. On the contrary, this circuit should be considered only as the first experiments in this interesting field. When using an efficient antenna, placed on a roof and targeted at a radio station, it is possible to obtain sufficient signal strength, even at a considerable distance from the station. This provides a high-quality reception on a headphones, and in some cases, you can get loudspeaking reception. It is possible to improving this receivers by using a more efficient detection circuit and using a high-quality resonant tanks, in particular, spiral resonators as resonant circuits.
V. Polyakov, Moscow