Continuous Tone (Self oscillating filter)

Hi,

I’m the proud owner of a freshly assembled Ambika, everything was working fine until I mounted the 2 last 4P Voicecards.
When I plug them they make a a continuous tone as if the filters were self oscillating.
I checked, double checked, triple checked that they were no short in those 2 voice cards. I also checked that every circuitry was in the proper place, it seems they are … Then I tried to switch the ICs between my working 4P and one that was self oscillating … They still make the annoying sound so I guess it’s not coming from the IC.

Here is a picture of the voice card, I only took one picture of them since I soldered them at the same time and they both have the same problem.

Is there any possibility that this problem come from the PCB ?
What test should I run to check it ?

You can find here an archive with the picture and a recorded sample of what it make (I’m playing with the cut off and I played a few note at the end to show you what’s happening.

Here is a Mediafire Link (file was to big for uploading) : http://www.mediafire.com/?v2d4vjq80jy61hc

Hi there,

the picture is looking good. No obvious swapping or something like that.
The sound is kind of interessting…^^
Did you test the faulty voicecards alone or in another slot on the mobo? Still the same problem?
In your sound example… did you also play with resonance? Does reso have any effect?

I tested both of the faulty cards alone, together, and with the 4 other voice cards, with different port ! The resonance do have an effect : it affect the tone in different ways (mainly the volume).
I’m almost certain that this noise come from the filter which is self oscillating, and this is strange since it’s not comming from the VCA itself (I tried it with the VCA of my working 4P and it still make this noise).

If I didn’t miss something while testing my guess is that there’s a problem with the PCB, but that’s strange !

Thanks for your fast answer :wink:

PCBs are factory tested, I haven’t seen a single evidence of a faulty PCB since the beginnings of MI. At least once a week I encounter a problem which I think is a faulty board or damaged IC, before realizing than it’s a bad firmware upload script or compilation flag, a weak solder joint, or something mounted backwards :slight_smile:

Things to check:

  • Signal at pin 6 of the DAC (should be the digital oscillators’s signal).
  • Voltage at pin 8 of the DAC (0V when no note is playing, 2.5 to 4V when a note is held - depending on envelope and velocity settings).
  • Pins 14 and 15 of the MCU should not be continuous with ground (they are on the SVF board, the chip detects that and scales its CV signals appropriately).
  • Pin 11 of MCU: 0V to 5V when you sweep resonance.
  • If you remove the LM13700 you should hear the oscillators signal at pin 1 of IC5 (and previous cascaded stages at pin 7 of IC5, 7 of IC3, 1 of IC3). No self oscillation can occur if the LM13700 is removed.
  • While the LM13700 is removed, measure current between pins 1 and 8 when resonance is at its minimum and maximum values.
  • Little sanity check: voltage on IC1 pin 7 should go from 2.5V approx to 0V when you sweep cutoff.

Let me know what you find…

Hi ! Thank you for the support, I just ran the tests here are the result :

  • I have the oscillator signal at pin 6 of the DAC

  • pin 8 DAC : 0V when no note, 2.96 when note held

  • I have a continuous 5V at pin 14 of the MCU and 0V at pin 15

  • The voltage range at pin 11 of MCU is ok.

  • I removed the LM13700 and I have the oscillator signal at each step.

  • When LM13700 is removed : pin 1 to 8 are 0V except pin 6 which is - 7.86V when minimum resonance. When maximum resonance pin 1 is at 4.26V, the rest are the same.

  • Voltage at IC1 pin 7 is between 1.95V and 0.59V

I guess that’s all !

I wasn’t that clear in the first message : the 2 new 4P voicecards have the same problem, the 2 old one works perfectly, since the old one have SSM chips and the new one have the coolaudio I thought the VCA was faulty but after switching them they work perfectly on the old (working) 4P voice cards.

Sorry for the inaccurate english …

> - I have a continuous 5V at pin 14 of the MCU and 0V at pin 15

That’s not what I am asking. Unplug the board, and check if there is continuity between pin 14 of the MCU and ground ; and pin 15 of the MCU and ground. This is not a voltage measurement.

> - I removed the LM13700 and I have the oscillator signal at each step.

No filtering?

> When LM13700 is removed : pin 1 to 8 are 0V except pin 6 which is – 7.86V when minimum resonance. When maximum resonance pin 1 is at 4.26V, the rest are the same.

I want to know the current flowing between pin 1 and 8. This is a current, in mA, you need to put the meter in a special mode for that and put the probes between pins 1 and 8.

Sorry, I’m kind of new to this !

So there’s no continuity between ground and pin 14 & 15. I do have the filter in the signal at pin 1 of IC5, but I have no current at all between pin 1 and 8 of the LM13700 …

I just tested the current between pin 1 and 8 of LM13700 on my working 4P and didn’t find any current ! But if I try pin 1 to 6 then I have something which change with resonance setting …

You’re right about pins 1 and 6 - these are the ones that required testing! Is the current null when resonance is set to 0 - increasing and approaching the mA range when resonance is at its maximum?

Will test it asap !