A dying OP-AMP: problems with BIAS and voltage

[Yoshi]

Hello, Im doing a litle experiment but is going horrible.
Im tryng to make a circuit that works with +18v and -9v, but when I hit the strings of my guitar, the sound sounds like a dying effect, like is chouking  with the sound. This effect dons´t happend with a +9v and gnd configuration.

what's happening? can the sound be fixed?.

My pickups are a Lundgren M8 for a 8 string guitar.

[Rob Strand]

I don't know what your test setup is.   What's after the opamp/C13?
Try putting a 1k resistor in series with the output (ie. in series with C13).   I might be oscillating.

Beyond that check your connections, maybe try a bypass cap on the negative rail to ground, located near the opamp.

[Yoshi]

After C13 there is nothing, is only the output that goes into a audio interface.

[PRR]

what's happening?

What actual DC Voltages do you have around the opamp inputs and outputs? Do they change when you play? Floating bias connections will give "farting" or "dying" effects.

[Rob Strand]

After C13 there is nothing, is only the output that goes into a audio interface.

Try the 1k, it's easy to do.

There could be a weird interaction between the audio interface and the ckt when the voltage swing is high.  Ideally you would want to listen to the output of the buffer alone, not connected to the audio interface.  Another angle is do a test (not a fix) with a 10k resistor in series with C13 then a 1k to ground and see what happens.  That drops the input levels to the audio interface by a factor of 11 and reduces any interactions.  Kind of a sanity check the make sure it's the buffer alone.

[Yoshi]

After C13 there is nothing, is only the output that goes into a audio interface.

Try the 1k, it's easy to do.

There could be a weird interaction between the audio interface and the ckt when the voltage swing is high.  Ideally you would want to listen to the output of the buffer alone, not connected to the audio interface.  Another angle is do a test (not a fix) with a 10k resistor in series with C13 then a 1k to ground and see what happens.  That drops the input levels to the audio interface by a factor of 11 and reduces any interactions.  Kind of a sanity check the make sure it's the buffer alone.

Thanks... your technique is working. And yes it is only a buffer, How can I maintain the level of the input? with your configuration the output drops significality

[Rob Strand]

Thanks... your technique is working. And yes it is only a buffer, How can I maintain the level of the input? with your configuration the output drops significality

Yes the drop will be severe.  It's only good for a test to identify the problem.   The way I read it is the instrument output is too high for the audio interface and the audio interface doesn't clip gracefully.   Maybe you could try only a 10k in series (no higher though).  If that works it would indicate some sort of interaction.  If that doesn't work go back to the divider and try increasing the 1k until you find the point where it doesn't overload.

If you think about even a 9V pedal it probably clips at 3.5V to 4V peak.   For an 18V system it's going to be higher.   If the audio interface can only handle say 2V peak then it will overload.    So you might need to drop the level by 1/2  or 1/3.

[antonis]

If you don't mind for signal phase reversal, try an inverting buffer..
(ratio of R_F/R_G set according to your taste..)