preventing op-amp self-oscillation?

[sneakthief]

i'm adding polyphonic CV inputs to a Korg Mono/Poly and am having a bit of op-amp trouble:

1. the mono/poly vco's receive 4 CV signals which come from the CPU

2. i installed switched jacks that interrupt these CV's so you can plug in your own

3. i used tl074's as input/output buffer's - but they self-oscillate like crazy and mess up the CV signals!

4. if i take out the tl074 and stick in jumpers between the CV IN and OUT pins it works fine.

any thoughts here? is there a better way to buffer these signals?

[goldenechos]

Have you tried a 358 or 324 opa?


TR

[sneakthief]

re. 358 - not yet. the only other quad opamps i had on hand were a lf444 and an opa404 which both did the same thing.

i just tried a tl084 and there was less oscillation but it was still noticeable. i wonder if it's a power issue - should i try tapping the +/-15v from somewhere else in the synth?

other option: i guess i could also put some dual opamps on a board as i have a bunch more of those lying around.

[frijitz]

[sneakthief]

i'll whip up a rough schematic tonight

[goldenechos]

I am not sure I understand the modification... you want to be able to cancel the CPU CVs and insert your own, correct. You are using switching jacks to do so. With this in mind, why do you need to buffer the incoming CV?

Please understand that I am just assuming alot of things as I do not have a schematic in front of me... BUT aren't all of the CV from the CPU going to a SUMMING point via a higher value resistor (100k or something I would guess)? If you are opening this connection so that an external CV could be inserted, would it not be possible to just feed the an external CV to each resistor attached to the summing node? No buffer needed. You could add a Pot to allow for attenuation of the CV.

Any buffer you add also increases the chances of errors in the form of DC offsets and what not... so 1V/Oct. can be skewed.


If you must use the the buffers:

-What frequency is the oscillation?
-Do you have Bypass caps neear the opa supply lines?
-Do you get oscillations with external CVs inserted?

TR

[sneakthief]

1. I tried 100n bypass caps on the +/-15v supply lines - didn't work.
2. Self-oscillation occurred with both external and internal cv's
3. The oscillation is crazy, unstable and quite high. Makes all the pitches warble a lot!

Some pics:




As I said, it works fine without the op-amp but I'll risk blowing the 4051. I already killed one during my experiments so I put a socket there

[sneakthief]

note: I don't even want to get into the crazy rigging I did to get each vco to trigger individually. hahah! There are 4 gate inputs for each SSM2033 VCO, but you also need to trigger the main VCA EG (the monopoly only has one EG for all 4 notes so it not really a "real" polysynth)

I basically had fun with a 4071 cascading the 4 trigger inputs down to one output to fire the main VCA.

[blue hell]

I'd try 100 Ohm resistors close to the tl084 outputs, they could oscillate because of the wires on the outputs being long or maybe you used shielded wire?

[goldenechos]

If you are concerned about blowing the 4051 you could put protective diodes on the the outputs.


TR

[sneakthief]

i've been reticent about adding protective diodes or resistors because dropping voltage decreases the range of the vco's and makes them more difficult to scale properly.

nevertheless, it's better to have protection than not - so i'll try some of these ideas out

[frijitz]