2 wet + 1 dry fx loop

[i2k]

Dear all,

I need an fx loop with 2 wet & 1 dry for bass so I took circuit snipplet from all over the net and came out with this schematic.
It starts with buffer which then split signal into 2 wet & 1 dry.
The wet lines have the polarity reverser as found on splitter blend from ROG.
The 3 signal each has their own vol control before mixed then goes to a JFET buffer at the end.

Could somebody review my schematic below and points out error and/or improvement before I etch the PCB ? Thanks a lot

[PRR]

You are really using bi-polar + and - 9V supplies? (This is unusual in pedal-world.)

If so, you might rig the last stage bi-polar. Otherwise the 10K mix resistors will drag-down the gate bias.

Good Practice:

Put some resistance between the Big Bad World and chips/amplifiers. If bad voltages happen (they do), then zero resistance means infinite current, some resistance means current can't be infinite. For small audio, 5K-50K at inputs and 50r-500r at outputs.

Cap-couple all Ins and Outs. There may be stray DC from your sources. And you never want your stray DC to upset your loads.

OTOH....

Your return-amps run at unity gain and modern TL07x have quite low DC error. It may not be essential to cap-couple the return amps to the pots.

The dry signal comes only from a TL07x which is fairly strong. No second buffer needed. This frees-up one opamp which can be your sum-buffer.

Don't over-size capacitors. Your 1.0uFd+1Meg at input is good to 0.17Hz. Such a large cap may leak more, may cost more, may be physically large. A tenth that is plenty for driving 1Meg load. (1uFd at output is reasonable because you may drive loads somewhat under 50K.)

The passive summer is probably fine, but there is "some" interaction between knobs. My inclination would be to make the pots smaller (10K) and the resistors larger (50K).

Most changes in pink: http://i.imgur.com/zyQQDin.gif

[i2k]

You are really using bi-polar + and - 9V supplies? (This is unusual in pedal-world.)

If so, you might rig the last stage bi-polar. Otherwise the 10K mix resistors will drag-down the gate bias.

Good Practice:

Put some resistance between the Big Bad World and chips/amplifiers. If bad voltages happen (they do), then zero resistance means infinite current, some resistance means current can't be infinite. For small audio, 5K-50K at inputs and 50r-500r at outputs.

Cap-couple all Ins and Outs. There may be stray DC from your sources. And you never want your stray DC to upset your loads.

OTOH....

Your return-amps run at unity gain and modern TL07x have quite low DC error. It may not be essential to cap-couple the return amps to the pots.

The dry signal comes only from a TL07x which is fairly strong. No second buffer needed. This frees-up one opamp which can be your sum-buffer.

Don't over-size capacitors. Your 1.0uFd+1Meg at input is good to 0.17Hz. Such a large cap may leak more, may cost more, may be physically large. A tenth that is plenty for driving 1Meg load. (1uFd at output is reasonable because you may drive loads somewhat under 50K.)

The passive summer is probably fine, but there is "some" interaction between knobs. My inclination would be to make the pots smaller (10K) and the resistors larger (50K).

Most changes in pink: http://i.imgur.com/zyQQDin.gif

Hi Paul, many thanks for the suggestions. In fact, apart from the resistors on opamp inputs, my initial schematic looks a lot like your suggestions where there's nothing except vol pot on the dry side. But after some reading, I concluded that a buffer is necessary there to prevent feedback. Although I also saw other design without buffer on the dry side.

Could you confirm that there will be no feedback without buffer on the dry side ?

About the coupling cap, I was putting one after each return but then I though that almost all pedal circuits I've seen has output coupling caps (which at the end acts as input coupling cap for the return side), that's why I remove them. But I do agree that it is lot safer to have a cap on any ins & outs, I'll add caps on the return. I'll keep the coupling caps before the wet pots as it's easier for me with existing layout with that caps.

The reason why I use bipolar supply voltage maybe stupid, because I found that it is easier to layout the PCB that way. I plan to generate the bipolar supply from single supply using outboard circuit.

[blackieNYC]

Nothing but encouragement for you - I built this exact thing a year ago, with the same changes PRR suggests. I don't have a schematic that's accurate-too many changes, but I did use just 9 v.   It is frikkin' awesome.  You'll love it. Drop the FET for an op amp stage, even if you have to add another. I breadboarded the whole thing with J201s for months and gave up.  Could not get it clean.  Other options I added are toggles for fixed HP and LP filters on the two sends, pots and switches feeding the final output back to either return (eh), and a boost stomp for the final output, which boosts and rolls off a little low lows and high highs. (Ranting now, this was my very first pedal, which was not a wise decision)There's a switch to turn loop A and B from serial to parallel (DPDT), and Dry, A, and B each have 10k pots, 100k summing resistors and stompswitches.  So there's 4stomps in a big enclosure with 30 holes drilled in it.  If I did it all over again I'd probably keep it just as it is.  Blend distortions, bring in noisy effects, or have 3 completely different sounds available with all the pedals on fixed settings piled in a heap behind the amp and just this thing in front of you.  Then there's the slow stereo panner I connect to the two loops, then folding the two sounds back into this morphing mono signal.  Fun stuff. Mine's wired point to point with perf and countless off board connections, which almost killed me, but it is very quiet, clean, and neutral.

[i2k]

Nothing but encouragement for you - I built this exact thing a year ago, with the same changes PRR suggests. I don't have a schematic that's accurate-too many changes, but I did use just 9 v.   It is frikkin' awesome.  You'll love it. Drop the FET for an op amp stage, even if you have to add another. I breadboarded the whole thing with J201s for months and gave up.  Could not get it clean.  Other options I added are toggles for fixed HP and LP filters on the two sends, pots and switches feeding the final output back to either return (eh), and a boost stomp for the final output, which boosts and rolls off a little low lows and high highs. (Ranting now, this was my very first pedal, which was not a wise decision)There's a switch to turn loop A and B from serial to parallel (DPDT), and Dry, A, and B each have 10k pots, 100k summing resistors and stompswitches.  So there's 4stomps in a big enclosure with 30 holes drilled in it.  If I did it all over again I'd probably keep it just as it is.  Blend distortions, bring in noisy effects, or have 3 completely different sounds available with all the pedals on fixed settings piled in a heap behind the amp and just this thing in front of you.  Then there's the slow stereo panner I connect to the two loops, then folding the two sounds back into this morphing mono signal.  Fun stuff. Mine's wired point to point with perf and countless off board connections, which almost killed me, but it is very quiet, clean, and neutral.

Yes, I also plan to put series / parallel switch for the A & B fx loop. I've figured out how to do it but won't put it on the schematic as things like that is better with outboard wiring. On the PCB I added a switch before Wet A pot which I can use later for series/parallel arrangement.

The power will come from a charge pump with +/- 9v output for better headroom.

HP/LP will be later ... it will be a constant tweaking.