Comparator problem

[Arsimantur]

Hello

I've built little analog synthesizer. It works, but I can't get pure sinusoidal signal - it is mixed with square wave. There is a pot (RV3) for Timbre that mixes sinusoidal with square wave comming from comparator, but even it is set to pure sinusoidal I still hear square.

Sinusoidal wave comes from oscillators to X, Q8 amplifies the signal then sin goes through C29 or to comparator through C30 where it's converted into square. Then square and sin is mixed with RV3 and goes through C36 to VCA.


I tried to get pure sin. I've removed pin 2 - no change. I've removed C30 - square disappeared from sin. Is this comparator overdriven? I tried additional low pass filter on V+, no change.
Measures: Q8 C=5,57 B=0,6 E=0,02, U1 5=3,84 4=3,84 2=3,56. I checked also supply. It gives V+ 8,2 and Vr 4,1, each filtered. 8.2 is split and goes through separate filters to different sections of synthesizer. One goes only to comparator. This one shows 7,7 while other are ok and show 8,2. Vr is 4,1 everywhere.

I hope you know what's happening and can help me.

[Enzo]

You refer to square. Are you actually seeing square waves on a scope? Or are you just using the term square to mean you hear distortion?

What is the signal amplitude coming in at X? What is the signal amplitude at the collector of Q8?

Removing C30 stops the distortion? Scope the signal at the collector of Q8, and how does it change with and without C30?

[Arsimantur]

Sorry, but i don't have a scope. I can only say that voltage on collector doesn't change with or without C30. I recorded audio to see the wave and you are right - it's rather distortion not square. But how does it happen? Maybe it is mixing with square going from comparator through C30 to C29?

First screen is without C30, second with it.

[Arsimantur]

Ok, I've got an oscilloscope, done some measures and it seems that problem is a comparator polluting PSU rails. This pollution goes to VCA. I see nice waves on R39.R42.C36 with/without C30 or pin2. Changes appear on VCA output when square wave is being generated (C30 and pin2 in place). VCA and comparator sections have separate RC filtering on V+ and Vr. There was clipping without pin2 on sine wave, so I've incresed R39 to 220k. Here is VCA

[Arsimantur]

Now I've found that most noticeble change is happening on decoupling capacitor! Connecting pin2 of IC1 adds harmonics on pin6 of IC2, but waveform is just slighly changed. Without pin2 of IC1 there is a nice sine on pin6 IC2 BUT AFTER C38 I see waveform with added peaks and amplitude is lower. It is only when I plug something to output jack. Tried to use different capacitors, no luck. Can someone explain what's happening and how to solve it? When I connect pin2 more harmonics appear and it's even worse.

BTW I use opa134 in place of 748.