2013-02-06 Coil And Crystal Tester

I had already built one of these several years ago, and since I’ve used it to test the crystals that I’ve salvaged.  It’s a simple circuit, using only a single transistor to make a Colpitts Oscillator with the two leads for the inductor run to terminals so that various inductors and coils can be connected up for testing.  The circuit I built years ago had a rectifier and second transistor to drive a LED to show that the crystal was oscillating.  In the one I built yesterday, Feb 5, there is only the single transistor for the Colpitts Oscillator. There are two capacitors connected in series, from the base to negative common.  The top lead of the top capacitor is connected directly to the base.   The bottom lead of the toip capacitor and the top lead of the bottom capacitor are connected together and to the emitter of the transistor.  The bottom lead of the bottom capacitor is connected to negative common.  I used two 470 pF capacitors; the values equal.  But I see other schematics where the top capacitor is much larger value compared to the bottom capacitor.

Using The Circuit

I connected the output of the oscillator to a frequency counter so I can compare various coils to see what frequency they oscillate when combined with the internal capacitance of the circuit.  In this case, the two capacitors are in series and when combined total 470 pF divided by two, or 235 pF.  There is about another 10 pF stray capacitance in the transistor and wiring, making the total about 245 pF.  What I liked about this circuit is that when I combine the 245 pF with a 100 uH coil, the resulting resonant frequency is almost exactly 1 MHz.  I thought that was very interesting.  Right in the middle of the AM broadcast band.  I put the 100 uH choke on the circuit and the frequency meter can be seen reading about 34 kHz higher than 1 MHz.  This allows me to estimate the frequency fairly accurately by tuning an AM radio to the frequency and then estimating the frequency by comparing it to the other adjacent radio stations on the dial.  In this case, when the oscillator is turned off, I could hear another radio station at 1020 and 1070 on the dial, so I knew the oscillator was about halfway in between.  Once I know the approximate frequency, and that the combined capacitance of the oscillator is 245 pF, I can use the resonance calculator (choose one from this search) to find the inductance of the coil.

Alternately, I can use the frequency of the radio to tell if my toroid coil is higher or lower than 100 uH, and add or subtract turns to adjust it to about 100 uH.  It doesn’t have to be exact, but if the frequency is higher or lower than about 1100 or 900, some turns should be added or subtracted.

When I looked at the waveform on the scope, it was distorted on the top.  I suspect this was because the capacitors were the same value, instead of being unequal as seen in many other schematics.  Some schematics show the upper capacitor as being ten times  the value of the lower capacitor.

The reason I used the 470 pF capacitors is that they are 1 percent silver mica, and are very accurate and don’t change with temperature.  I calculated that I could use six of these capacitors and get 1410 pF (three in parallel) for the upper capacitor, and 313 pF for the lower capacitor (three caps, two in series with a third one across either of the others).  This would give me an equivalent of 256 pF for all of the capacitors.  The ratio is not ten to one, but it is much better than the equal values and it is close to the original 245 pF.

Update Feb 6 afternoon – I changed the capacitors to the six as explained above.  The waveform is still distorted, but maybe not as much as before (I don’t have a way to directly compare them).  The distorted waveform becomes less distorted when the supply voltage is below 2 volts (the circuit oscillated down to 1.6V).  The frequency is lower, it measured 938 kHz.   But the frequency is sensitive to what is connected to the output.  When I connected the scope probe, the frequency dropped well below 900 kHz.  I suspect the frequency counter is also causing it to drop.  I disconnected everything from the output and found the frequency on the AM radio.  I found out the frequency of the station it is next to: 980.  The circuit is oscillating at about 970 kHz with a 100 uH inductor.  According to my calculations, it was supposed to be at 977 kHz, so the actual frequency agrees closely with the calculated frequency.

One way to reduce the frequency sensitivity, I believe, is to take the output from a tap on the emitter load resistor.  Another way is to put another transistor on the output to act as a buffer stage.  If the coil tester is used with an AM radio and no load on the output, then nothing would have to be done.

Additional Circuits

I was considering adding buffer if needed, a rectifier and filter to the output.  The reason for this is to make a frequency sensitive DC output.  The DC output would vary with the frequency, so that for example if the coil tester was oscillating at 938 kHz, the output might be 0.938 volt (the output would be read with a DMM).  This would not be as accurate as listening to the radio, but it should be good enough to give a good idea of the frequency, and thus the value of the coil.  For right now, I’m happy to be able to find the frequency with the AM radio, and confirm that the inductor is about 100 uH.