Effects loop circuit

[merlinb]

What about input caps?  I'm using 1uF here, but I could go with a 0.22uF cap easily.  Would there be any advantages to doing something like that?

Since you're input resistors are 1Meg, 220nF caps would give you a cut-off frequency of 0.7Hz. More than low enough! Even 47nF would be plenty.

[Sage]

Since you're input resistors are 1Meg, 220nF caps would give you a cut-off frequency of 0.7Hz. More than low enough! Even 47nF would be plenty.

What would be the benefit of using a lower value for the input cap?  Just to reduce load on whatever is driving the circuit?

[Sage]

More importantly, are there any drawbacks to lowering the value of the input capacitors (C2 and Cx), other than the higher filter cutoff frequency?  I'm planning on going with 0.1uF now, instead of 1uF.

[mth5044]

I'm planning on going with 0.1uF now, instead of 1uF.

For what reason?

[Sage]

For what reason?

Primarily because the spot where I'm inserting the effects loop is right before the reverb section, which has a 0.1 uF cap on it, and I thought it would be a good idea to use the same thing here.  Any reason I shouldn't?

[Sage]

Good news: it look like it works!  I had a brief scare when I lost power to the amp, but it turns out to have been a blown fuse.  I think that happened when I hooked up my little effects loop board to the wrong IC in the amp by mistake.  Once I replaced the fuse and reconnected my circuit to IC3, I was able to get it working.  The input and output are wired to a dpdt rocker switch for true bypass, and I can switch the loop in and out; it sounds pretty transparent to my ears.

Using R3 to attenuate the send and boost the return was definitely a good idea.  The loop had to meet three requirements and, optionally, a fourth:

1. It has to go after the overdrive channel's gain section
2. It has to work on both the overdrive and clean channels
3. It has to come back before the reverb section
4. Ideally it would also occur after the tone controls

There's only one place in the amp's schematic that it can go and serve all of those requirements.  Unfortunately, this means that it must also occur *after* the volume controls.  That's why the dual potentiometer is so important.  Without it, as you turn up the volume on the amp, the signal into the effects loop gets more and more distorted.  The potentiometer allows me to "clean up" that signal by increasing the amount of attenuation on the send while simultaneously boosting the signal back up on the return.  It's not perfect thanks to high tolerances - there's a slight audible volume bump halfway down because the tapers don't line up perfectly - but it's close enough.  I'm sure I'm not the first to think of it but I'm super proud of it since this is my first foray into amp modification.

I'll test the loop with the amp at higher volumes either tonight or tomorrow and see if I need to tweak anything else.  I will probably add a 100K pulldown resistor to the output, to help reduce any pops when I flip the switch.  Don't need one on the input since the switch diverts the input to ground when off.  I'll post the final schematic here when I'm done.

Thanks to all the folks who helped me: ashcat_lt, MrStab, PRR, GibsonGM, merlinb, mth5044, and R.G. for some primer stuff in another thread.  You guys really helped me get a basic understanding of how this circuit works and how to modify it to do what I wanted.  This has been a great learning experience, and I'm hoping to apply some of what I've learned to my next project.

[Sage]

Well, I tested the effects loop on the clean channel of the amp with the volume dimed, and even with nearly 16db of attenuation on the loop I get distortion.  Took the pedal out of the equation and ran a patch cable directly from the send to the return, and I'm still hearing it.

I suspect that the signal going into the board is too hot for the TL072 to handle, at least at 18V.  I probably need to move the attenuator in front of the send buffer instead of after it.  I'll rewire it tomorrow and see if that fixes the issue.  I'm worried that I'll still get distortion from the return boost, but we'll see.

[Sage]

Putting the attenuator in front of Cx appears to have fixed the distortion, but when the loop is switched on, there's a significant volume drop.  Maybe it needs to be inside the capacitor?  Or could this be an effect of using a 5k potentiometer at the input?  Would a higher resistance resolve this?

[Sage]

Sorry, I should have elaborated:
If you put the pot at the input to IC2, you probably want 10k or even 100k, because we don't know exactly what the amp circuit is like. It might not be able to drive a 5k pot happily.

So it's 5k pot at the output, OR 100k pot at the input. Your choice.

Well using 100k at the input seems to have resolved my volume drop problem. Can someone explain why 100k would be easier for the amp to drive than 5k?  I would have expected the opposite.

[lead2203]

There are a bunch of loop circuits if you check Google. Look at some of the Crate and Ampeg schematics...simple loop that works good ...like the Ampeg 502(I think) Lee Jackson stuff. or even the Marshall loop....Study these and ...you will "see the picture clearer" ...plus.. you will end up with something like  these circuits anyway....  With any loop ..you should always attenuate the signal before the buffer .......even a tube one. A 220k(can try 1M also) before a 10k to 47k to ground works good...I end up at 220k/22k or 47k most times.....or 220k/10k to ground if switchable masters are after the loop....letting the buffer do its work .....with nothing in the way.. is best. I would put the return before the op-amp...although after works good too.  There is no need for a dual potentiometer unless you have a wet/dry mix type setup ...and with a op-amps that is an easy setup to do.

[Sage]

Yeah, I've been studying other loop schematics, but it seems like they're all so different from mine.  Here's what I ended up with after a bit of trial-and-error:



I did end up dropping the dual potentiometer.  Even though it was a cool gimmick, the adjustability wound up being unnecessary.

It works great; I can turn the clean channel up all the way without getting distortion from my delay pedal.  I have the loop itself on a true bypass switch.

The only problem that still remains is a sort of squeak when I power the amp off.  It doesn't happen when the loop is bypassed, and it doesn't happen if I power down from standby, but if I flip the power from on to off while the loop is on, after a moment you hear a short squeak that fades out and increases in pitch as the amp powers off.  If my delay pedal is plugged into the loop, you can hear the squeak echo through the delay, so whatever it is gets sent out through the loop.  I don't know what's causing this, exactly, or how to address it.  Any ideas?

[lead2203]

I don't see a big difference...Is its transparent? how much does it color the tone..if it does?.... Check the loop in the Koch amps ...close.. but with a mix set up. The 100k does not help the amp drive the loop. Its just setting up proper impedance and level ..I think 100k/ 10k would be the right arrangement. The Boss delays are the worst ones ...they distort the delay when hit with too much signal..and you could hit the rails of the op amp still here also if driven by a tube amp.
As far as the sqweek ....try a .47 or close to that ...filtering the op amp power ...put these right at the op-amp......and put a 39pf to 100pf over the 100k in the return side....check your audio and power grounds... keep them separate.

[Sage]

I don't see a big difference...Is its transparent? how much does it color the tone..if it does?....

I increased the value of the input caps in order to lower the filter cutoff, moved the attenuator in front of the input, dropped the potentiometer in favor of fixed resistance for the attenuator and boost, and increased the value of those resistors by a factor of 20.

It's totally transparent to my ears.  I've got it wired to a switch for true bypass and I don't hear any difference whether it's on or off.

The 100k does not help the amp drive the loop. Its just setting up proper impedance and level

So this is just a matter of making sure that a low impedance signal goes into a higher-impedance input?  Like at the front of an amp?

The Boss delays are the worst ones ...they distort the delay when hit with too much signal..and you could hit the rails of the op amp still here also if driven by a tube amp.

Mine's a TC Electronic Flashback x4.  It's got a pristine digital delay that I'm using to test.  As long as it can handle the signal without distorting, I figure I've attenuated the signal enough.

As far as the sqweek ....try a .47 or close to that ...filtering the op amp power ...put these right at the op-amp......and put a 39pf to 100pf over the 100k in the return side....

What exactly would this accomplish?  I'd like to get a better understanding of the problem before I try to fix it.

check your audio and power grounds... keep them separate.

... There's more than one ground?

[merlinb]

The only problem that still remains is a sort of squeak when I power the amp off.  whatever it is gets sent out through the loop.  

Presumably this is due to something oscillating as the rail voltages collapse at unknown rates, briefly pushing something into an unfavourable operating condition. It's unclear whether this is IC1b or an earlier stage of the amp. One way to find out might be to replace R10 with a diode (pointing towards C8). This would allow the FX loop to remain on for longer after power down, since the caps can't discharge back through the diode. If it still squeaks, it is most likely an earlier stage.

[Sage]

Any recommendation on what kind of diode?  I've got a 1N5239B Zener lying around.

[merlinb]

Any recommendation on what kind of diode?  I've got a 1N5239B Zener lying around.

No I mean a regular diode. 1n4148, 1n4001, whatever. Any will do. Literally just remove R10 and put the diode in its place.

[Sage]

Finally got around to replacing R10 with a 1N4148 and sure enough, that solved the problem.  Thank you again for all the help.  Here's my final schematic, for anyone who's interested:

[jotajota0003]

anyone can create a pcb this circuit for me?

[Transmogrifox]

Quote from: lead2203 on July 25, 2014, 04:03:43 PM

    As far as the sqweek ....try a .47 or close to that ...filtering the op amp power ...put these right at the op-amp......and put a 39pf to 100pf over the 100k in the return side....

What exactly would this accomplish?  I'd like to get a better understanding of the problem before I try to fix it.

A capacitor looks like a low impedance to changing voltages.  How much of an impedance it appears to be depends on the RC time constant.

When you use a large value with RC time constant less than your useful audio bandwidth (<20 Hz, or probably <60Hz guitar with 7th low string), you increase the amount of DC voltage fluctuation that can be transferred to the input of your op amp.

For example, with a 1uF and 1M resistor, a power up or power down ramp on a prior stage will pretty nicely couple a 1-second or faster ramp to the input of your op amp.

If this ramp causes the bias to shift to or near the rail it can make the op amp go unstable in some circumstances.  At the very least it can create an audible "thump" on power up or power down.

Generally it's good to design RC cut-off frequencies to about 1/10th of the lowest frequency you expect.  That way cascading multiple stages has a negligible effect on the frequencies of interest.  Remember the cutoff is already down 3 dB, so you want it low enough that the attenuation in the pass band is less than a fraction of a dB.  If you have several such stages coupled throughout your signal chain, then 1/2 dB loss cascaded 5 or 6 times has an audible effect on your low-end response.

In this case, usually guitar is about 80 Hz lowest unless you detune or have a 7th string low B.  To design for something reasonably balanced between the two constraints:
1M, 22nF ==> 7.23 Hz cutoff.

Probably anything between 10n and 47n would be ok in this instance.  Clearly 1uF is ok if you are liking it now.

In the end if your circuit with the 1u doesn't cause any more problems then just leave it alone.  It doesn't hurt anything if you aren't hearing strange sounds any more on power up and power down.  Just thought I'd answer this question since it didn't appear it was answered.