Splitter/Looper/Mixer REDUX

[jayfroo]

Got some good info when I posted but it was all lost a few days ago after the server issue. To review...I'd like to add feed back to each channel of this circuit http://www.runoffgroove.com/splitter-blend.html

I remember someone had suggested I put volumes on the returns which I liked. They also discussed a possible issue with phase inversion on the feedback part of the circuit. I will have the feed back controlled by a 100k pot and but I don't anticipate any phase issues if i tap the return jacks and send the signal right back into the send jacks.

I am also planning on adding a feedback switcher to send Red output to Green input and vice-versa. Should be fun.

If you remember replying to me a few days ago, and recall what you shared I would appreciate it.

*Can anyone explain what the Q1 part of the circuit does? If its a buffer why doesn't the other side of the circuit have it?

[slacker]

Here's the schematic I posted the other day.



Q1 is a buffer that provides a high 1Meg Ohm impedance input for that return. This is because U2A only has an input impedance of 100k or 200k depending on what mode it's in. The other side doesn't need a buffer because it already has a 1Meg input impedance, set by the 1Meg resistor.
If you use the red channel you could get phase related issues using feedback, if the pedal in the loop inverts the phase then your feedback signal will then be out of phase with the loop input so the feedback might not behave like you want, or it might do something interesting. The green channel avoids this because you can flip the phase if you want, for maximum flexibility you could build both returns like the green channel.   

[jayfroo]

I was planning on using a passive feedback circuit (just a pot with one lug connected to the return jack, the wiper to the send jack, and the other lug to ground). Does it need to be more complicated?

[slacker]

No that will work fine, you probably still want the mixer resistors like I added though, the only real advantage of taking the feedback after the buffers like i drew is that on the green channel you can flip the phase.

[blackieNYC]

I had offered up the mixer idea, and the input attenuators on the returns. And stompswitches on the returns as well.  I pointed out that this prevents noise from unused pedals from getting into your signal path.  But then again, if you happen to have output controls at the end of each pedal chain, you can skip the attenuator pots on the returns.  Know what I mean?  But if it's chorus or delay, you'll want those inputs adjustable. Yeah, what happened to this thread? 
Oh, hey, there's my mixer in the little picture there.  Too many switches. Unllike your circuit, my polarity reversers were on the sends, and my summing op amp inverted polarity, using the inverting input. That ended up interfering with the feedback.  Slacker is right - with this design you can compensate for one reverse-polarity pedal chain, and your passive feedback won't be out of phase and you shouldn't have any problems.  80% of what I do with the 3ch mixer can be done much more simply with the circuit you've got.  Highly recommend the splitter looper mixer.

[jayfroo]

Awesome, thanks for getting back to me

[jayfroo]

Alright, I built it an am working through the troubleshooting process. The Green channel works but the Red channel is nasty and distorted. I traced the signal through end up getting a nasty output from U2B (pin 1/2).

http://postimg.org/image/io4vu2jbh/

I connected Red send to Red Return with a patch cable and same with green. Ran a test signal through (500hz 1.5Vpp 0VDC offset) I get a perfect signal out of the green channel side of U2 (pin 7) but from pin 1/2 I get a large square wave. Is this a sure sign that the chip is bad? I didn't use a static wrist strap when soldering/handling, should I have?

[blackieNYC]

Weird.  Are you sure pins 1 and 2 are shorted?  For all your inverting inputs, only pins 6&7 should have resistance to them.  If one of those 220k resistors snuck its way into the feedback loop of 2B, that might do it. Might swap the op amps around too.

[jayfroo]

And now is when I kick myself for not using sockets

[jayfroo]

All the solders look good. No shorts that I can find. Is it possible for half of an IC to go bad?

[blackieNYC]

To be clear, the outputs of three op amps should measure a short to the inverting inputs. I think you know this, just saying you could double check with the ohmmeter.   Half can go bad.   They are really two different operations going on.

[jayfroo]

Opamp is good, i put a socket in and tried another, same issue. Is it possible that the Red channel needs a high impedance buffer like the Green channel? Here is was ive got (minus the volume pots on between the return jack and 220uF caps)
http://s2.postimg.org/tnhjjxfkp/Untitled_1_copy.jpg

[jayfroo]

Figured it out! The input of my Red Channel Feedback Pot was between the cap and op amp, this was throwing off the DC bias provided by the 1M resistor to Vref. So can I tap in just before that 220uF cap or should I add another cap and isolate the connection point?

[blackieNYC]

There ya go. Yes, move the cap. Between the wiper of the pot (where your input goes to the top, and your output comes from the wiper) and the 1M bias resistor.