Problem with svf card


I have a problem with one of my 6 svf card.
When I turn the synth on, this card is whistling (when playing a note) like the filter is always fully open and lots of resonance. There is also some harsh noise coming through. (Actually its kinda cool sounding but not expected :wink: )
Now, after 5-10 minutes of warming up, it fixes itself and works like the rest.
I have resoldered all joints, switch the ICs with others, move the card at the different spot in the ambika, but I still have the problem.

What component could be the one that is depending on warming up?

I can upload a video with the sound it makes if it can help, let me know.

If anyone has any idea what I should test to pin point that problem that would be great, as I dont know much more than soldering in electronics.


The best way is to walk along the path of the signal and trace it with a scope.
Do you have access to some sort of oscilloscope? Something super cheap will do.

If not, we can also use an audio cable to "listen" to the signal at various points in the circuit.
  1. Connect a 1k-10k resistor to the tip of the cable (exact value doesnt matter).
  2. Connect the ring of the cable to GND.
  3. Connect the other end of the cable to a cheap soundsystem (old PC speaker or something else that doesnt go very loud. Depending on where you probe you may get very loud signals and you should make sure not to destroy your sound system or your hearing).
  4. Now you can use the other leg of the resistor to “listen” to the sound.

okay so here are some points to probe:

  1. check the point between C21 and R20. You should see (or hear) the raw oscillator signal.
  2. check IC3 (TL074) pin 14 with the resonance at 0. You should see (or hear) nothing.
  3. check IC3 (TL074) pin 14 with the resonance at max. You should see (or hear) the BP signal.
  4. Check IC3 pins 1, 7, 8. You should see the HP, BP and LP signals there respectively.

let me know how this works out.
Recordings of the sounds and images of the board may be helpful for further debugging.

Thanks a lot for helping me with this.

I havent had the time to do the sound test, but I have made a video showing the problem. You can hear how it sounds.
You’ll see it fixes itself after a couple minutes. So its hard to debug after that. Yesterday it actually worked right away when I started the synth.
Sorry for the bad video moves, but its hard to play a note, turn a pot and film at the same time with 2 hands :wink:

I’m not sure about the video. I find it quite hard to tell what sound is expected and what is unexpected. It would be great if you could do the same sound on a working card and the not working card - alternating between them so that we can directly hear the difference. Because a lot of what I hear in the video could just be a cool patch with some filter FM or some other weird modulation.

It’s in cyclic mode, so every note i play are a on different voice card one after the other. So you can hear the correct sound on the first 5 notes i play at the very begining and then when it hits voice card 2 you can hear it’s different:
the sounds before 0:03 are the other voice cards, and then when I hit the voice card 2 around 0:03 (and I hold it) you can clearly hear the difference.
Also watch the voices leds, you’ll understand what I mean.

Furthermore, at 0:27 I am closing the filter and you can hear some wrong noise going through, whereas the other voices are silent.
It’s a very simple patch, no modulation involved. Just init patch.

Sorry I didnt explain this correctly at first.
Does it make more sense now?

ah okay. So you can control the cutoff of the filter on the broken voice card, but it seems way more open than on the other voicecards. Right? Then maybe you have a problem on the control voltage generation for the filter. I suggest you look at the voltage at the junction between D3, R11 and R12. Measure that on the broken voicecard and a working one. Then open the filter statically to some arbitrary value (don’t apply any modulation, no env, no lfo!) and make a voltage measurement on IC1 pin 1 (again on the broken voicecard and on a working one).

Are there any differences between the two cards?

Ok, so I did the listening and voltages tests .
Voltages seems pretty much the same between cards: ± 0.04v (around -4.5v on the D3 junction, and around 1.2v on IC1 pin 1)

Now listening to IC4, i get a high pitch sound on the bad card on IC4 pin 1, 7, 8, 14 even when not playing a note.
So the high pitch sound is the problem. But where does it come from?

Do you get the high pitched sound even when all oscillators are disabled?

Yes i get the high pitch sound with osc disabled.
Its sound like the filter self oscillating. If I close the filter it goes lower in pitch. If I open it goes higher in pitch.
Now, if i turn up the resonance, it goes a bit lower in pitch, and lower in volume.

I made a video, listening to IC4 pin 1 and opening full and closing full the filter:
(careful with your hears its sounds nasty :slight_smile: )

edit: SORRY I KEEP WRITING IC4 BUT ITS IC 3 I meant (The TL074)

With resonance at 0, I would expect to have no signal at IC3 pin 14. Can you measure that?

Actually there is a signal on pin 14 with reso at 0 between freq 70 to around 100. It’s the same as on pin 1 7 and 8. Signals on pin 14 and 7 are about the same volume. Signals on pin 1 and 8 are the same but louder. (that’s with no notes being played and osc at none).

One thing I noticed is that I can hear that whistling when I turn the synth on, for one second during the time the synth boots.

Okay, great, we’re getting closer.
To my understanding, with the resonance turned to 0, there should be no signal at pin14. If you look at IC3D in the schematic, you’ll see that is simply converts the current out of IC2D pin 13 into a voltage that can be fed back into the input mixer that is formed by IC3A and the resistors around it. So the problem is not IC3 but IC2 or the circuit that generates the control voltage for IC2D.

Can you compare the voltage at IC2 pin 14 (the control voltage input of the resonance vca) between a working card and the bad card?

I get a nice 8.08 volts on all cards at IC2 pin 14.

All pins are the sames. Though when I play a note, i get 0.1 v at pin 4 on IC2 on the bad voice card, where the other cards are at 0v on pin 4.

Also on IC1 pin 7 i got 0.3v on the bad card and 0v on the other cards. (while playing the same note on all cards)

So I found out that when I push on C10 (styren cap) it fixes the problem.
I redid the soldering a bit on it. I can bend it so it’s working, but if I push it back its buggy again.
Can this type of component have a bad contact inside? I should probably buy a new one and replace it?

If C10 is buggy, replace it.

I wonder how this ties into our problem with the resonance VCA, though.

pin 4 @ IC2 is held in a virtual GND position by the opamp behind it. A voltage measurement doesn’t make much sense there. Also, the current out of this pin is at audio rates, so measuring with a multimeter is pointless.

What is the filter cutoff frequency set to? I remember your bad card sounded like the filter was a little more open than on the working cards. If (after fixing C10) you still find that this is the case, you may want to check the resistors around IC1A.

I’ll change the cap and check the resistors if it ever happens again.

Thanks you so much for all the help you’ve been providing, I have a fully working ambika again. We did find the problem in the end, and I learned that you can actually listen down the audio path on a circuit, which kinda blew my mind (even though it’s very logic and down to earth :slight_smile: )

1 Like