Op-Amp Overdrive with Bipolar Suply

[liquids]

I feel like I've done this before...

However, I have a simple op-amp gain stage on the breadboard, running off +/- 15v via AC rectification/regulators etc.

When I put a diode or two in the loop, it seems to sound particularly awful, somewhere between gated and octavia-esque, though different....

If I re-configure things to run with just the positive supply rail, ground as V-, and a voltage divider as the center of the supply...things go back to sounding normal.

Now I am wondering - do op amp+diode circuits actually 'work' better off of a single sided supply because the 1/2 V+ reference foward biases the diodes at all times, compared to when the said op amp stage is using ground as a center supply?

Or is that not typically an issue and it's likely that something else at work here?

[CurtisWCole]

A certain "Boutique" pedal with the initials C.M. and rhymes with Flexicone has a quad op amp that is biased at 0v and seems to work well. Madbeans has a version of it called the Boneyard. I've wondered the same thing and what the advantages or disadvantages would be.

Curtis

[Chugs]

I have made a number of standard opamp/diode in the feedback loop overdrives but running off a bipolar supply at +/-9V. I prefer them that way compared to a single supply with VR setup.

[liquids]

Thanks for the confirmation that it's just me.      I do feel like I had done this before...needed a second opinion.   

[PRR]

> a simple op-amp gain stage ..... When I put a diode or two in the loop, .... awful.... with just the positive supply rail

Post a picture.

I may know what you are doing but there are far too many variations on "simple".

[earthtonesaudio]

Sounds like a wiring error to me.

[liquids]

I'm not in front of my breadboard...but here's the scenerios I've tried just last night...



I always assume it's a wiring error...but I can't understand why it works so well on a single supply.  And if I use the bipolar supply but remove any diodes, it's 100% normal.

And to make matters worse, the whole thing has me so confused that I can't decide if the sound if the Bipolar + LEDs sound weird or normal.

[liquids]

I forgot to put the 1k to Vref in the drawing on the single sided supply image...but that's a drawing mistake, not something I'm not doing.

[PRR]

> the 1k to Vref in the drawing

That's the only flaw I see.

What is "OK?"? Really OK, or no-OK?

[earthtonesaudio]

Did you measure dc at all pins? 

[liquids]

Thanks for the help.  I did measure voltages and they were all mV from ground...from my recollection.  Which is hazy.

I'll be back in town tuesday and possibly able to go back to scratch then.

'OK' means it was working normally...as a boost with no diodes, it was clean-ish outside of the rail clipping...I think I only tried that with a 47k resistor in the loop instead though...that was another variable in there...

Im just going to have to go back to scratch I guess.  I was assuming it was a diode issue there for a while...and maybe it is...but having two separate clipping scenarios not working seemed to disuade me from pursuing that further.

Anyhow, if nothing 'obvious' is wrong, and there's no good reason clipping shouldn't happen as usual with +/-15v rails, my questions has been answered more or less and resolved to be a debugging/error issue, not an issue of 'circuit design' for lack of a better phrase, and that was what I was more or less trying to confirm or deny more than just getting it working...

Thanks for the extra pairs of eyes.

[earthtonesaudio]

The idea itself is sound and should work as expected.  Good luck bustling the gremlins out of their hiding places.

[mrmoo1337]

First, what are you driving with the potentially +/- 15 volt output signal?   Maybe it sounds wrong because you are clipping the input of the next stage?  Or do you have a lower signal level and are using the available power rails?  

Secondly,  won't the diodes do their job at +/- one diode drop?  won't this result in +/-0.8 volts of output with gain and then effectively unity gain once the diodes conduct?  More like crossover distortion than clipping?

[liquids]

First, what are you driving with the potentially +/- 15 volt output signal?   

Noting, I have it isolated so as to reduce/eliminate the variables and as you can see, there is a volume control at the output, which I turn down accordingly.

Maybe it sounds wrong because you are clipping the input of the next stage?  Or do you have a lower signal level and are using the available power rails? 

If so, the diode-less option wouldn't sound perfectly normal.

Secondly,  won't the diodes do their job at +/- one diode drop?  won't this result in +/-0.8 volts of output with gain and then effectively unity gain once the diodes conduct?  More like crossover distortion than clipping?

LEDs and mosfets allow a much higher voltage swing before they clip.  The regular silicon diodes only came into the picture when I had issues with the mosfets.  The +/- 15v is not necessary if this were the circuit entire, but I've isolated it on it's own as a part of a larger circuit because it was a problem area.