I have a design question. Working on a buffer and tone control.
I have two dual-op amps U1 and U2 for a total of 4 devices.
Should I design my circuit so that U1 does everything left channel related, and U2 does right channel...
OR should U1 do the buffer for both channels and U2 do the tone control for both channel, thereby splitting them by function.
I'm tempted to have each op amp do their own channel to minimize the possibility of channel crosstalk and therefore imaging improvements, however, I can't help but notice a nagging feeling that there's a pitfall somewhere with this theory...
Please advise!
I have two dual-op amps U1 and U2 for a total of 4 devices.
Should I design my circuit so that U1 does everything left channel related, and U2 does right channel...
OR should U1 do the buffer for both channels and U2 do the tone control for both channel, thereby splitting them by function.
I'm tempted to have each op amp do their own channel to minimize the possibility of channel crosstalk and therefore imaging improvements, however, I can't help but notice a nagging feeling that there's a pitfall somewhere with this theory...
Please advise!
Either will work just fine.
I prefer the first option ("U1 does everything left channel related") because it lets me use identical PCB layouts for left and right channels. Cutting in half the opportunities for error.
I prefer the first option ("U1 does everything left channel related") because it lets me use identical PCB layouts for left and right channels. Cutting in half the opportunities for error.
In less "corksniffy" applications, I would just do whatever was easier to lay out on a PCB, however, i'm being a little vain for this project lol
If you use one dual op amp per channel, you can lay out the left channel,
and then block copy that layout for the right channel. Less work, and better
channel separation. There's no downside.
and then block copy that layout for the right channel. Less work, and better
channel separation. There's no downside.
If you want a lot of gain (such as microphone or phono preamp) its best to use one dual opamp for the input stage and another for the next stage, simply to allow extra power-rail filtering for the input stage to isolate it from the higher signal volages further down the signal path which might lead to inadvertent positive feedback and howl-round/oscillation.
If everything's line level and no large gain factors are involved this is unimportant.
If everything's line level and no large gain factors are involved this is unimportant.
Another possibility is that by grouping by function instead of by channel you have the option of using different op amps for the input side versus the output side. Could be interesting to play with.
Whatever works to ensure the minimum signal path length is achieved.
Are you tapping off the signal to reach the potentiometers (volume and tone control), or are you planning to have the potentiometers on the PCB?
Maybe draw what you want to do, so that we can provide a better suggestion, one that is applicable to your exact implementation.
Are you tapping off the signal to reach the potentiometers (volume and tone control), or are you planning to have the potentiometers on the PCB?
Maybe draw what you want to do, so that we can provide a better suggestion, one that is applicable to your exact implementation.
Bit confused by this, we're talking audio frequency, not microwave. Where circuit impedances are relatively low and the standard precautions applies (signals routed with shielded cable), there's no particular problem with signal path lengths for audio. A few _metres_ of shielded cable might be a whole 100pF which at impedances of 10k or so starts to be a problem around 150kHz. Routing lengths within one chassis is thus not a concern you need have.
True, true. That maths does seem like iron-clad proof that minimising the signal path on a PCB and/or minimising the hook-up wire length, and possibly using an unshielded wire, does not make any difference.
Last edited:
Do not forget the crosstalk with in a chip, so all left, all right is probably safer. However, in most cases for stereo, who cares if crosstalk was only 60dB. Really.
Pay attention to what Wrankin said.
Pay attention to what Wrankin said.
Well -30dB crosstalk is the threshold of audibility anyway, crosstalk is never an issue for stereo in practice, only for separate signal sources in the same box.
Thank you everyone who replied - I ended up grouping them by channel. Given that the pots were on the board, the layout was certainly a bit messier than I would have liked with a few unavoidable jumpers (single-sided board). Had I known the layout headaches, I would have grouped by function instead, But I was already too invested to start over. In the end, the new circuit sounds significantly better than the cheap 1970s single-supply single-transistor one it replaces.