Farfisa oscillator repair help, please!

[drabb1983]

I am working on a Farfisa Mini compact which has a problem with the 'A' natural oscillator/divider board. The symptom is that the note generated by the card is closer to B natural. I've replaced all the transistors for 2N3906 and replaced the electrolytic capacitors with no improvement. I used an audio probe and found that the fundamental note is not being produced, the 1D and 2D have the same pitch and then the note divides down as it should. Oddly, the voltage reading (to ground) on all 3 of the transistor legs in the oscillator circuit measure very similar voltage whereas I expect to see a drop ( I verified this on schematic and also tested the another good card). I also noticed that the tune coil does not provide any change to the pitch - I tested the tuning coil for continuity and get 120Ohm A-B, and 350Ohm C-D. I am stuck at this point and wondering if this could all be related to a bad tuning coil, though I am not convinced the tuning coil is bad because the Ohm reading seems ok.

Thanks in advance for any help!

[JQA]

Welcome to the forum! As much as I'd like to begin on a positive note, I'm sorry to say that you've gone about fixing the issue in a way that has likely caused more problems than you have solved, and with already greater effort than necessary. Step 1 should have been signal tracing, to identify which stage was at fault. There is no chance that every divider stage needed new transistors, and 2N3906 is not a proper substitute for OC71 (or whatever germanium PNPs were in there). It is silicon instead of germanium, and thus has much higher base-emitter forward voltage: about 0.7V instead of 0.2V. The current gain (i.e. "beta") is also higher. So without circuit modification, the bias will be off, maybe enough to silence the master oscillator circuit, even if not the dividers. Bulk-replacing electrolytic caps is also no substitute for real troubleshooting, and of course can cause problems too if the caps are installed backwards or with poor soldering, etc.

What I suggest is that from here, you focus on making each stage work one-by-one, starting with the master oscillator. Find a good OC71, and put it back in the oscillator circuit. Make sure your new electrolytic (replacing the 25µF across the 10kΩ resistor) is the correct polarity and capacitance, presumably 22µF using modern parts. Then see if the master oscillator comes to life. Assume the coil is good for now, given it has continuity in both windings. If the oscillator still doesn't run, check that you are getting both the 8V supply and the same vibrato voltage as the good cards. Check the voltages on the transistor; the base should be about 0.2V lower than the emitter, and the collector some greater amount lower than the base; compare to a good oscillator. If necessary, check the resistors and the tank capacitor ("Co" in the schematic). Most likely it will be fixed before you need to even think about the coil.

Once the master oscillator is running, tune it to correct pitch (assuming it is possible). Then the dividers can be addressed. The dividers are the "free-running" type, meaning they can oscillate on their own, as you have found. The free-running frequency must be slightly lower than that desired. This frequency does not depend much on the parameters of the transistors, nor the 1µF capacitors; rather, it is determined mostly by the film capacitors used to cross-couple each transistor pair, and by the base and collector resistors. These may have drifted out of spec.

Hope this helps - feel free to ask any questions, and let us know how it goes!