Pulse Counting FM Receiver with Tubes


In this page I describe my Pulse Counting FM Receiver. The very first thing I insist on is to say that I was inspired to run this project by the articles found at at cool386.tripod.com/pcrx/pcrx.html. I want to thank John, the owner of this site, for his, in my opinion, really very cool work and his still more cool contributions. In fact my pulse counter project originally is John's idea.


The second thing I want to say is that this is only a preliminary documentaion of the project. It shows just the schematic and some pictures of the finished receiver. I will later add some more documentation and the most important of all I will try to understand how this receiver really works and consequently to add some theory of operation part.


Well then, let's start!


The Schematic

As already said, the original schematic can be found among a lot of other very cool schematics on John's site. Here I post a schematic I draw using EAGLE (tm) in order to show how I implemented Johns design and in order to show all the technical documentation on this page.


Pulse Counting FM Receiver Schematic

Figure 1: Schematic of the Pulse Counting FM Receiver


I just want to add here some comments:

Short Functional Description

The first ECF80 tube is at the same time an RF-amplifier (triode part) and an oscillator / mixer (pentode part). The oscillator is producing a relatively pure sine wave between 80 and 110 Mc. It does not nicely or not a all swing below or above these frequencies. The incoming RF signal is mixed with the oscillators sine wave and produces a IF (of about 300 kc). This implicits an image problem, you will be able to hear a station twice very near together (station frequency - oscillator = 300 kc and oscillator - station frequence = 300 kc). This is not really a problem in case of wide spaced stations, but a severe problem in case of narowly spaced stations. I will try to fix this problem in a next version of the receiver (:-)). The output at the ECF80 looks like this:


Ocillator Output Signal

Figure 2: Output of the oscillator at RF frequency

As shown by figure 2, the radio is tuned to a station at around 104 Mc. It can also be seen, that the sine wave is distorted a little bit at the faling edge, but not on the rising edge. This is quite typical when a station is tuned in. The sinewave looks totally undistorted when no station is tuned in. If the oscilloscopes time axis would be set to around 300 kc, then the mixing product could be seen, but this looks just like nothing, at least on this oscilloscope.

In any case, the FM modulation found at RF frequency can be found exactly the same way now at the mixing products frequency which is about 300 kc. Why 300 kc? Well, this is because of the fact that you need some 270 kc bandwidth in order to hear undistorted audio when listening to an FM station. Therefore, the subsequent IF stages are limited at at a band with of around 300 kc. All other frequencies are filtered out. The FM signal at 300 kc is the amplified until clipping by the the first and the second EF80 tube.This can be seen in the following pictures:


First IF Tube Output Signal

Figure 3: Output at the first IF tube


Second IF Tube Output Signal

Figure 4: Output at the second IF tube

It must be said that the AF signal at the input of the first IF amplifier tube is very week (some mV PP), at the input of the secoond one it is already quite bigger (some 100 mV PP) and finally at the input of the third EF80 tube, it is quite big (some 5 V PP). If you look closer, you can see that the output of the second EF80 is already clipped. This is quite normal and not at all a problem, because the information is in the frequency changes and not in the amplitude of the signal. If would even be better if the signal was a bit more clipped, somehow approaching the form of a square wave signal.

The third tube plays, together with the EAA91 tube and the subsequent filter the role of the demodulator. I will try to explain now how the demodulation is done.The third EF80 tube does have some different function than the first two ones. It takes the clipped FM signal at 300 kc and differentiates it. The differentiation is done by biasing the tube in a way (the components around the third EF80 tube are different to those around the first two ones) that it is easily driven into saturation (Thanks to John for the explanations). Figures 5 shows the pulsed signal at the output of the third EF80 tube. Figure 6 shows by comparison of the input and the output signal, how the differentiation is done: One can easily see that the pulses of the output signal are exactly there where the input signal reaches or leaves the clipping region in the positive have wave. This behavior is that one of an differentiator.

Third IF Tube Output Signal

Figure 5: Output at the differentiator tube (third EF80)


Third IF Tube Input and Output Signal

Figure 6: Input of the differentiator tube in comparison with it's output

The rectifier tube removes the negative spikes of the pulse signal as shown by figure 7. Then it passes it to the subsequent filter. This one converts pulsed signal into a AF signal which again is fed into the AF amplifier. The principle of the the conversion of the pulsed signal into an AF signal is very simple: A condensator is charged by a diode. A lot of pulses produce a high voltage, less pulses produce a lower voltage. Thereby, the condensator is discharged by the potentiometer. This is alos the reason why the potentiometer should be of a relativly high value, otherwise the filter would be overcharged and not behave as it should. Some additional remark to the filter: It consits of two parts: the first consisting of a 47k - 1n combination is a deemphasis,the second one consisting of a 47k - 100p combination is a low pass with its cut of frequency at 1/(2*pi*R*C) (34 kc).

Third IF Tube Input and Output Signal

Figure 7: Output of the rectifier tube

Well, this explanation of the demodulation is a little bit simple. The truth is slightly more complicated. It is more like this: The clipped signal at the input of the differentiator tube contains the following informations: among a lot of frequencies forming the square wave (clipped) signal it contains the the carrier (about 300 kc) and the the AF (0 - 20 kc). The AF is frequency modulated into the carrier. The same information is still contained in the differentiated signal,but the distance of the AF frequencies and the (very high) frequencies forming the pulses is considerably big. Therefore passing this mix through a low pass filter with a cut off frequency of about 30 kc will resut in a quite clean AF signal. I will do two experiencies in order to prove this theory: One is to implement a MATLAB (tm) script showing this facts and the other is to use PSPICE to simulate a circuit doing the same as the demodulator. I will post both of them on this page soon.


The following pictures show how I have constructed the receiver.

Radio and Powersupply

Figure 8: The Receiver and the external Powersupply

Radio top view

Figure 9: The Receiver seen from top

At the left side, the RF part is hidden below the punched aluminium sheet. It was necessary to prevent touching the wiring, because my sons are more used to transitors and therefore not common to the high voltages of tube designs. The aluminium sheet is NOT for shielding. The three tubes that are aligned are the IF tubes. The small one is the rectifier and the one besides the speaker is the AF amplifier tube. The output transformer is out of a VERY old radio,that was built in Switzerland during WW2. If gives quite a nice sound. I measured it with a home built impedancemeter which told me that it has a input imoedance of 11k againts 8R.


Radio front view

Figure 10: The Receiver seen from the front


The reduction drive is home built. I drilled a 20mm diameter hoe into a wooden plate and pressed the button into it. I had to reduce the buttons diameter slightly on the lathe. afterwards, I rounded the wooden plate using an abrasive belt and by fixing a 6mm shaft at about 50 mm distance from the belt. turning the plate round the shaft results in a perfect outer contour. The small knob at the right side is a standard construction element with a lego tire (sorry Jan and Frederik :-)) on its shaft. With the friction of the tire against the wooden plate, I got a perfect 1:5 reduction. The variable capacitor, on whichs shaft the woden plate sits, has itself a reduction of 1:3. This results in a total reduction of 1:15, which is extremly fine for tuning in stations exactly. The grey knob at the left is the oscillators gain potentiometer. When using an external antenna (the receiver works local stations without antenna), the gain must be reduced strongly. The numbered knob at the right side is the volume of the AF amplifier. The 6.3mm jack is for the headphones.

Figure 10 shows the radio from the bottom.

Radio bottom view

Figure 11: The Receiver seen from the bottom


At the right side, one can see the wires goind or coming from the RF section. The drilled blue wire is the filament voltage. The red and black one are VCC and ground. The white on is the AGC signal. The yellow one is the IF signal. The BNC connector at the top right is connected to the choke-resitor connection at the RF triodes cathode. at the right side, one can also see the build in telescope antenna whos shaft is connected to the BNC's center pin (blue wire).

All construction and wiring is done around the center metal tubes of the sockets. This allows compact and ground-aware construction. Each center tube is connected with a 1mm silver wire to a solderlug that is fixed by one of the socket's screws.

At the top left, one can see the VCC and ground bananajacks. Soldered to the VCC jack, there are two 470R resistors in order to ajust the voltage. The jacks aside are for the filament voltage. Two 100R resistors are used to balance against ground.

The socket in the middle at the top is a cinch socket I rubbed from some old electronic. I used it to test the AF amplifier which I build first (warm up exercide :-)).

In the middle, one can see the three IF tube sockets an related components. At the left, just below the third IF tube sits the rectifier tube and at the left bottom one can find the AF amplifier tube socket and its components. The yellow and the green twisted wires go respectively come from the output transformer. The potentiometers and the phone jack can be found at the bottom of the picture.

One can also see that all filaments are bridged by 0.1 uF capacitors and that NONE of them is connected to ground. The only filament that is connected to ground is the one of the RF tube. If have no picture of the RF part. The RF part can be seen in figure 11.

Radio RF part

Figure 12: The Receivers RF section


A picture of the powersupply cani is shown by figure 12. It consists of a EZ81 tube and a 0A2 tube, that is not inserted in the actual picture. The choke is from the same radio as the output transformer. The capacitor (2 x 47uF) is suspect since it is also from the same radio. I believe, that is not the original one, but already onece repalced. I will replace it as soon as possible.


Figure 13: The Powersupply


Further work

The list of todo's below shows,what I plan to do with this receiver in the very next future:


Any comments and suggestions are welcome. You may write to medard at jfam dot ch or medard dot rieder at rhone dot ch