VCF issues [WILLZYX KIT]

[hywlen]

I've tested everything up to the VCO using both oscilliscope and headphones, i am definately getting saw and square waves from probing the waveform switch.
Probing the potentiometers results in a sine wave of +- 100mV and frequency 50 Hz. Turning the pots does nothing and the frequency doesnt change with input voltage so im assuming the signal is lost at some point in the VCF.
Are my transistors (Q23 and Q14) in the correct way, theyre different to those used in the manual so i just used common sense (or not).
http://postimage.org/image/qnfvode23/

It could just be me probing wrong,can anyone confirm its the end point of vr3 you have to probe?

Thanks for any help!

[MMM]

The orientation of the transistors looks ok.

Probe point is actually VR4, not VR3. Please check again.

[hywlen]

Thanks for your reply,
I have found the same waveform as described probing vr4. Perhaps I'm not understanding the term probing correctly. When I ground the negative and probe the end pin of vr4, I get no signal but without the ground, I get the signal I have described. I think I'll check my DC voltages throughout the filter if this doesn't lead to any conclusion.

Thanks again

[MMM]

When I ground the negative and probe the end pin of vr4, I get no signal but without the ground, I get the signal I have described.

If this happens with an ungrounded probe (the ground doesn't have to be obtained directly; sometimes it's obtained indirectly over the earth connections) you're probably NOT getting the filter signal here; I've experienced this myself quite often.
I suggest at this point tracing the signal through the filter ladder. Please keep in mind that the signal is going to be VERY weak compared to the signal you've found at the waveform switch.

Perhaps you could start at C17, then continue through C18, C19, C24, C26.

[hywlen]

I've traced the signal starting at the waveform switch. On the right side of R62 the signal is still there but on the left it becomes 0.8 VDC. There is also no more ac in any of the VCF including the ladder. I understand the signal should be MUCH smaller and have zoomed in on my oscilloscope to no effect.
I've also checked my voltages against the set kindly uploaded by someone on here. The C17 anode is correct and the cathode is 7.24 VDC (instead of 5.34). In the area around Q12, voltages are higher than expected and all voltages at Q12 itself are wrong. Any ideas would be much appreciated

Thanks

[MMM]

Ok, this needs a bit thorough debugging.

To really find the reason for what goes wrong here you need an oscilloscope that can be AC coupled. And then we need to establish where is left and right because "the right side of R62" had little meaning otherwise.

So I assume we're looking at the component side of the PCB and the processor (or its socket) is in the lower right corner. In this case the right side of R62 is the side that connects directly to the waveform switch. So here'll you get the same signal as form the middle pin of the switch. This should be for sawtooth wave a signal with about 5v peak to peak travelling on a DC offset of approximately 5V. This one you can easily measure with a DC coupled scope.
On the other side of R62 things should look vastly different. The DC offset is probably still there (albeit a bit lower) however using the same setting for the scope as before will not show you any meaningful results. The signal itself is about 100mV or less peak to peak; this is only visible with an AC coupled scope and the appropriate zoom level (the signal will also look somewhat different; that's ok).
The signals in the filter ladder get even smaller and will look very "fuzzy".

One last thing: When looking at the picture it could be that Q22 and Q12 are not "correctly" oriented (but I may be wrong about this, it's a bit hard to see). If this is the case it should not be much of a problem (those transistors are symmetrical); however please doublecheck if you used the correct parts for each location.

[hywlen]

So I've now looked again using ac coupling.
The signal coming out of R62 is similar to what you mentioned (100 mV peak to peak), then smoothed by C17 correctly. The signal into both R70 and Q12 is maintained then the output of R70 and Q12 both become fuzzy beyond recognition. Do you think it could perhaps be faulty transistors? I've checked all the resistors and capacitors several times now.

Here's another photo to double check the transistor orientation:
http://postimage.org/image/krsnniwfp/


Thanks again

[MMM]

Can you measure the voltages around Q12 and Q22 and post them here?

Make sure to set Cutoff to max and Resonance to min.

[hywlen]

Heres are the voltages taken from the component side (opposite to the pots side)
http://postimage.org/image/eewigwc2b/
Specifically for the transistors:
From left to right on the component side
Q22 - 10.99, 11.59, 11.56, 11.6, 11.18
Q12 - 7.2, 9.08, 8.51, 9.84, 9.07
That's with the pots set to what you said

Thanks!

[MMM]

So the problem is likely Q12 or its surrounding parts. What irritates me most is the 9.07V on the right hand side of your drawing. That's something that shouldn't be there.

Can you remove C23 and measure again? This is the part where the resonance path passes through so the filter should (theoretically) work fine without it (without resonance of course).

[hywlen]

Without C23, everything I measured before is the same. Does this help narrow down the problem?

Thanks!

[MMM]

Yes, it helps.

Please measure the voltages on both sides of

C18
C19
C24
C26 (this should theoretically be identical to the 10.99 and 11.18 measurements you got on Q22)

The ladder is somehow out of balance and we need to narrow down which part is responsible for this.

[hywlen]

10.99 - C26 - 11.18

10.35 - C24 - 10.74

9.72 - C19 - 10.30

9.09 - C18 - 9.86

edit - I discovered that C25 and C27 were using 104 capacitors rather than the 2A104's. I've changed them to what they were supposed to do but nothing else has really changed.
Thanks

[MMM]

I believe Q12 is faulty here. The values within the ladder don't look too bad.

You may want to do one last test: Lift one side of R69 and measure again all the voltages at Q12.

[hywlen]

R69 lifted, measurements from the component side at Q12:
7.22, 9.86, 9.68, 9.11, 9.11

Thanks