[i]"Upper harmonics appear only when pin2 and C30 are in place while without R42 and R40."[/i]
This is strange! There should not be enough signal coupled back to Q8:C via R44-R43-C30 (R44 is 4M7 so should be almost irrelevant)
Two possibilities come to mind - IMO both are obscure and unlikely..
1.) If any signal appears on IC1:A pins 4 or 5, which is greater than 2/3 the supply to this IC, then the IC can go into phase reversal (a undesired condition) which would produce high current spikes and may cause (here I am guessing) current to be sunk or sourced from these input pins.
2.) There are likely to be fast edges on the square wave from pin 2, and these could couple capacitively to other tracks if the layout is poor.. A fix for this - [b]See correction at end of this posting [/b] But - oops.. you might find that slugging the edges will cause problems for the pitch-CV circuit..
In fact.. looking at this circuit fills me with utter horror! Everything is out-of-balance.. C40.R42 will charge via R51 during the +VE output, and discharge without any current limiting via pin 2 when this goes low.. this will generate a substantial (but extremely fast)current spike! .
If you do not need the CV out, remove C40, C28, Q5 (you can leave the other components as they are effectively then out-of-circuit) - This alone may fix some of the audio problems, otherwise try my suggestion in (2) as well.
Fred.
[b] ADDED CORRECTION ->[/b]
My original suggestion in (2) was flawed.. it would have produced a much slower rising edge than the falling one..
You would need to add a 15k pull-up resistor connected directly to pin 2, and have a 10k resistor between pin 2 and R15.R42.R44, and a 1nF from this point to ground. C40,Q5,C28 would need to be removed, and the signal level at R42.R51 qould be reduced by 50% and have a DC offset which you could eliminate (but I dont think you will need to) by having a capacitor (220n) in series with R42. You may need to decrease the value of R42 or R40 to increase the square wave audio level - do not reduce these to below 100k.. I would go for reducing R40 to 220k first.
Increasing the value of the capacitor (lets call this the filter capacitor) between R15.R42 to ground will soften the square wave edges - at 1nF you should not notice any filtering effects (other than, hopefully, curing the spikes) as it rolls off at above 10kHz - at 3n3 it will start rolling off harmonics above 3kHz.
This is strange! There should not be enough signal coupled back to Q8:C via R44-R43-C30 (R44 is 4M7 so should be almost irrelevant)
Two possibilities come to mind - IMO both are obscure and unlikely..
1.) If any signal appears on IC1:A pins 4 or 5, which is greater than 2/3 the supply to this IC, then the IC can go into phase reversal (a undesired condition) which would produce high current spikes and may cause (here I am guessing) current to be sunk or sourced from these input pins.
2.) There are likely to be fast edges on the square wave from pin 2, and these could couple capacitively to other tracks if the layout is poor.. A fix for this - [b]See correction at end of this posting [/b] But - oops.. you might find that slugging the edges will cause problems for the pitch-CV circuit..
In fact.. looking at this circuit fills me with utter horror! Everything is out-of-balance.. C40.R42 will charge via R51 during the +VE output, and discharge without any current limiting via pin 2 when this goes low.. this will generate a substantial (but extremely fast)current spike! .
If you do not need the CV out, remove C40, C28, Q5 (you can leave the other components as they are effectively then out-of-circuit) - This alone may fix some of the audio problems, otherwise try my suggestion in (2) as well.
Fred.
[b] ADDED CORRECTION ->[/b]
My original suggestion in (2) was flawed.. it would have produced a much slower rising edge than the falling one..
You would need to add a 15k pull-up resistor connected directly to pin 2, and have a 10k resistor between pin 2 and R15.R42.R44, and a 1nF from this point to ground. C40,Q5,C28 would need to be removed, and the signal level at R42.R51 qould be reduced by 50% and have a DC offset which you could eliminate (but I dont think you will need to) by having a capacitor (220n) in series with R42. You may need to decrease the value of R42 or R40 to increase the square wave audio level - do not reduce these to below 100k.. I would go for reducing R40 to 220k first.
Increasing the value of the capacitor (lets call this the filter capacitor) between R15.R42 to ground will soften the square wave edges - at 1nF you should not notice any filtering effects (other than, hopefully, curing the spikes) as it rolls off at above 10kHz - at 3n3 it will start rolling off harmonics above 3kHz.
