Yusynth VCO problems

[carlo64]

Hi, I'm building the famous Yusynth Vco
but I ran into some difficulties
I can't get regular waveforms by adjusting the trimmers t3 / t4
any idea what i could check?
Thanks



about 1000 hz
saw.. it looks ok


triangle.. horrible


if I increase the frequency, it gets much worse

[ixtern]

Looks like saw short edge is too long. You can try:
- Q1 FET switch with small ON resistance like PN4391(2N4391),
- U5 comparator: smaller open colector output resistor R20 (even down to 1k),
- faster buffers (OPA4134 or so),
- fiddle with C11 value, may be needed smaller or bigger depending on the Q1 type.

[carlo64]

Thanks for the reply,
I will do some tests, hopefully well :-)

Carlo

[carlo64]

I tried to vary the values of both r20 and c11 but I didn't get any improvement; unfortunately I have no fet to replace Q1. Any other ideas?

Thanks

[Thalassa]

Check that the 5V voltage reference is correct ( U7a). Also I would lower the frequency to 100hz or so for calibrating the waveforms. If I'm not mistaken, the volts/div in the oscilloscope is in 0.1ms per division setting so your VCO is more of less at 1Khz. When you rise the frequency the effects of the filtering capacitor C12 is more noticeable and that will distort the waveform.

You can see in here how the frequency can affect waveform shapes.

https://modularsynthesis.com/moog/901/901.htm

[carlo64]

Hi
here are some images at about 100 Hz and the corresponding one at about 1000 Hz
in fact my problem seems like that of the link you posted, only that it occurs at much lower frequencies and more

sine wave 100Hz


sine wave 1000Hz


triangle 100Hz


triangle 1000hz


saw 100hz


saw 1000hz


pulse 100hz ac ??


pulse 100hz dc ??


pulse 1000hz ac (same on dc)




Volts pin1 di u7 = 4.57

I tried to increase it to exactly 5v by changing one of the two resistances r50 / r49 but I did not find significant changes in the output waves

Hi
here are some images at about 100 Hz and the corresponding one at about 1000 Hz
in fact my problem seems like that of the link you posted, only that it occurs at much lower frequencies and more
I tried to increase it to exactly 5v by changing one of the two resistances r50 / r49 but I did not find significant changes in the output waves

On his website http://yusynth.net/Modular/index_en.html he shows perfect waveforms at all frequencies!

Carlo

[Thalassa]

You should check all your waveforms in DC mode. As you could see with the pulse the AC filtering affects the shape wave form.

[carlo64]

I have already done that test (dc mode) but the only wave that changes is the pulse, the others are the same.

[Thalassa]

The saw in the 100hz seems that is not centered and it doesn't have 10Vpp. The saw wave has to be more or less 10Vpp and go from +/-5V. Any unbalance between the positive and negative side will affect the triangle and the sine wave. But again this can be an effect of the oscilloscope been in AC mode. Put the oscilloscope in DC mode , VCO frequency at 100HZ, and check the output of U6A. If your don't have a more or less +/-5V saw wave, replace R23. Lower the value to increase the saw amplitude and vice versa. Once you have a nice +/-5V saw wave, use the T3 trimmer to center it. After that check the triangle output and calibrate T2 until you have a perfectly symmetrical triangle.

To reduce the spike in the triangle wave , you can rise C12 from 47p to 68p or 100p.

[carlo64]

I worked a bit on r50 to have 5volts (pin1 U7a) and on r23 to have about + -5v on pin1 of U6a; I adjusted a little T2 and T3 and these are the results .. really bad .. I'm about to give up :-(

pin1 U6a saw (ramp wave) 100Hz

pin1 U6a saw 10000Hz


triangle 100Hz

triangle 1000Hz

triangle 10000Hz

[Thalassa]

After seen the 10K saw I think that the problem that you are having is that C8 is not discharging fast enough. Check the Gate of Q1, you should see a pulse that should be something under 10 microseconds long. If it's longer or take to long to rise, the problem could be in the comparator U5. Try for example to lower R20 and R21( 10k and 2k for example). Also you could try change C10 for a smaller or bigger value.

If the pulse is ok them you could try to swap the discharging capacitor C8 for another type.