Parallel Op Amp Stages

[DrT]

I am working on a design that would involve splitting an incoming signal to two separate op amp based gain stages that end up getting mixed at the end of the signal chain.  The input to both op amps currently goes through a similar resistor and capacitor pair before going to the non-inverting input.  Could I instead split the signal after the resistor and capacitor to reduce part count or is it better to keep them separate?  Also, is crosstalk on a dual op amp an issue for gain stages?  If so, is there a rule of thumb for when this will become an issue?  I've breadboarded a couple configurations, but think the overall noise of the breadboard could be masking any differences.

Thanks

[aron]

You can take a look at the Quadrafuzz.
https://www.paia.com/ProdArticles/quadrafz-design.htm

[idy]

Most designs, like the Quadrafuzz put the signal through an opamp before splitting, either as a buffer (unity gain) or with some gain. Either way it is buffering and making sure that the parallel stages won't load the signal.

I think cross talk is specified on the data sheets for opamps and is really negligible. It still wouldn't "be a thing" in your idea even if there was some.. Cross talk would only be bad if you had two different signals, like a stereo L+R channel and didn't want ghosts appearing as faint echos in the wrong speaker. If the same signal is going into two opamp stages, how would cross talk be a problem?

[ElectricDruid]

My experience has been that crosstalk in duals is much lower than in quads, or at least circuits based on duals are less prone to it. The datasheet numbers are definitely low, but there are situations where it becomes an issue. Think of all the "LFO ticking" threads on here that recommend keeping the LFO on a different chip from the audio, for example. If your gain stages are high gain, I'd certainly be careful about it.

If I understand the first part of your question correctly, you're talking about splitting the signal after the AC-coupling cap and input resistor of an inverting op-amp. If that's what you meant, it won't work. Although it seems like you've got two sets of cap+resistor doing the same thing (and you sort-of have) that doesn't mean that one is redundant, unfortunately. If you left one out, you'd have the -ve inputs connected together. The op-amps aren't going to like that since it will remove their ability to balance their output and their input (they'll be fighting).

[amptramp]

If I understand the first part of your question correctly, you're talking about splitting the signal after the AC-coupling cap and input resistor of an inverting op-amp. If that's what you meant, it won't work. Although it seems like you've got two sets of cap+resistor doing the same thing (and you sort-of have) that doesn't mean that one is redundant, unfortunately. If you left one out, you'd have the -ve inputs connected together. The op-amps aren't going to like that since it will remove their ability to balance their output and their input (they'll be fighting).

I believe he said non-inverting amplifiers and non-inverting input, so this would be OK.  The feedback goes to the inverting input so there would be no conflict.

[ElectricDruid]

I believe he said non-inverting amplifiers and non-inverting input, so this would be OK.  The feedback goes to the inverting input so there would be no conflict.

Oh, so he did. I missed that the first time. Sorry, my mistake.

However, in that case, there isn't much point having two op-amps doing the exact same thing. You'd be better off feeding one non-inverting op-amp and then splitting the signal *after* it, rather than before. The two paths don't need individual buffers.

[DrT]

I'd like to be able to set a different amount of gain and very different caps to ground from the inverting input on the two parallel signals, so I need to split before the two op amps.  I will take a look at the Quadrafuzz.