Excessive SMR4 - MkII resonance [SOLVED]

Hi folks,

It’s my own fault really, after being inspired by reading about the new Moog synth having a feedback feature for gnarly distortion by connecting the output back to the input level control. The output signal was the wrong polarity on the Shruthi-1, so as I turned up the input level the output went quieter instead of louder and more distorted, but somehow in the process I’ve damaged my SMR4

SYMPTOMS:

Filter still tracks beautifully over 10 octaves as before, and all my SDE filter modes still work perfectly, so the four stages of the filter appear to be fine. Most factory patches are still perfectly recognisable, but many sound harsher than before - the reason is that the filter now starts self-oscillating with the Res control turned up to just 6

This massive change to the resonance parameter made me think I’d damaged IC3OTA1 (LM13700N), but I’ve just swapped it for a new one with no change. The only other things I can think of that would affect the amount of filter resonance are Q4 (2N3906) or IC4 (TL074P), neither of which I have direct replacements for at the moment, although I wasn’t fiddling anywhere near that part of the circuit.

QUERIES:

1. How high do others have to turn up their SMR4 resonance before it goes into self oscillation?
2. Can anyone give you a few normal voltages around the resonance components to check against to more easily track down where the problem is?
3. Any other ideas from anyone please?

Many thanks!

Martin

1. On mine, it starts self-oscillating at 58

2. Remove IC3. Measure the current flowing between pin 1 and pin 6 with resonance = 0, resonance = 8… up to 63.

Thanks Olivier - that gives me something far more positive to measure.

I’ve just done that, and here are the results:

Res, I (microamps)
0, 0
8, 35
16, 69
24, 103
32, 137
40, 171
48, 206
56, 240
63, 270

That looks a pretty linear progression to me (with a slight droop on the highest result)!

Does that rule out IC4A and Q4 as possible culprits?

Martin

This is correct, the OTA receives the right bias current.

Try removing IC3. If you still get a well-behaved signals on the 2P and 4P points (with no resonance of course), then the suspects can only be IC3 (the socket?) or the resistors around it.

I’ve checked thoroughly (including all my mods) and everything sounds and measures fine (minus the resonance) on both 2P and 4P points, but as soon as I replace IC3 the filter self-oscillation is back at res = 6

Changing IC3 for another (previously unused) one still makes no difference at all., and I’ve examined all the joints around IC3 with a magnifying glass and can’t see any problems. This is bizarre!

It will no doubt turn out to be due to something really silly.

Since the other two LM13700Ns obviously work fine I’ll carry on by unplugging the current IC3, swapping it with one of the two used in the filter pole circuitry and then seeing if the 2P/4P test still works - if it does then all ICs are confirmed working and I’m down to the IC socket (which I’ve already checked as far as possible), four resistors and a capacitor.

I’ll get there in the end

Martin

And I did get there in the end.

The culprit? R18 (which should measure 1K) actually measured 14K

But there was nothing wrong with it, as an inspection with the largest magnifying glass I could find finally proved - the soldered joint at one end looked a bit dodgy, had worked perfectly for months, but suddenly decided to think it was a 13K resistor (delusions of grandeur)

Out with the soldering iron, and my Shruthi-1 now sounds as sweet as it always did, and starts self-oscillating at Res = 59

Yay!

Thanks again for your superb support Olivier!

Martin