Design procedure for effects loop voltage output

[Gaz]

I'm designing an effects loop on the bench here, and want to tune the output for both rack gear and pedals. The loop has a send control, so I'm just trying to determine the maximum voltage I should allow to leave the loop. I'm wondering if anyone has any advice about how how I should set my signal generator and amp settings.

I was thinking set the generator for maybe 500mv output and turning the gain all the way up... realizing I was pulling that methodology completely out of my ass, I thought I would inquire here.

How much voltage should I be aiming for? 1.2Vrms maximum with everything all the way up? This is where I'm stumped. I'm measuring the all the voltages with a true rms meter, btw.

Thanks!

[R.G.]

Partly this depends on what you're putting in the loop. If you're putting normal-ish guitar effects, you need to feed them a guitar-ish signal, which is about 100mV to maybe 200mV. If you're putting studio stuff, it probably wants line level, 1V peak (ish), about 775mV rms.

And the loop needs to send this, as well as reamplifying it back up to the internal level it started at.

It might even be good to put a guitar/line switch on it, with about a 10:1 level drop on the send and a 10:1 gain on the return when it's in "guitar".

[Gaz]

Partly this depends on what you're putting in the loop. If you're putting normal-ish guitar effects, you need to feed them a guitar-ish signal, which is about 100mV to maybe 200mV. If you're putting studio stuff, it probably wants line level, 1V peak (ish), about 775mV rms.

And the loop needs to send this, as well as reamplifying it back up to the internal level it started at.

It might even be good to put a guitar/line switch on it, with about a 10:1 level drop on the send and a 10:1 gain on the return when it's in "guitar".

Thanks, R.G., and for the voltage references. I would like to use the loop with both pedals and line level stuff. My question is still when tuning attenuation/make-up gain of the loop, do I set it for 775mV maximum? What I mean is that let's say I'm slamming power chords though a bridge pickup and have a Lexicon reverb unit in the loop, do I design it so 775mV is the most that can ever 'hit' the reverb, or do I design with a bit of voltage headroom? My concerns are not enough voltage for the line level gear, but also damaging something with over-voltage. Hope that makes sense.

[R.G.]

This is not my area of greatest expertise. I'm just casually aware of most of this.

Most line level stuff will take in 1V peak (about 775mV rms) with some headroom to spare. It depends a lot on the specific gear. Most line level stuff (and I wish someone with more studio experience would chime in here) can probably go to about 2V peak without distorting, I think, so at 1V peaks, 2V p-p they still have some input headroom left.

Guitar level pedals really expect about 1/10 of that, although the differences between single coils and "distortion humbuckers" make this foggy. But "guitar level" is usually considered to be about 100mV peak, about 1/10 the size.

This would be clearer if guitar amps followed the pro audio "line level" format, where there is a distinct preamp separated from a power amp by cables, and where the level on the cable is standardized (at!) by "line level". Guitar amps are usually designed as a unit, and the gain of the power amp from the input to the PI to the output of the amp may not be standardized to be at full power with odb line level at the power amp input.

I got out my old copy of one of the first printings of The Ultimate Tone, where there's a whole chapter on effects loops. It says that it's good to provide 2V p-p output capability for line-level loops, and attenuate by 10:1 for guitar level loops. This is either switchable or tweakable.

For the return, you need to provide either 2V p-p input for the line-level loop case, or provide a gain of about 10 if it's a guitar level loop - again, switchable or adjustable, since you may not know exactly what the stuff in the loop will do.

I think TUT recommends a follower driving the loop send, with a switchable attenuator, and then an adjustable return amplifier with variable gain.

I think providing a way to be sure you can get to 2Vp-p on the loop send is more important than limiting the voltage on the send jack. Unless you have really high levels coming out of the loop, you're not likely to damage line level stuff in general in the loop. Maybe the right way to do it is to make it have a clean output to well over 2V p-p, but limit at something well over that - maybe 8-10V with a pair of zeners.

[Gaz]

Thanks again, RG, that's helpful info.

...OK, this is really embarassing, and it made me feel like I'm losing my mind, but I almost asked this same question almost verbatim a year and a half ago:

http://music-electronics-forum.com/t29797/

[Gaz]

Oh my god, it gets worse. This is from the old thread:

"What I really should have titled the thread was: "Effects loop design procedure."

I don't even recall that thread. Oh dear.

[soundguruman]

The effects loop send should be 1V PP sine wave (same as output of guitar)
when the level control is about mid position.
AND this should be measured with an effects pedal plugged into "send."

Effects pedals are designed for a guitar level input...
which is about 1V PP maximum.

Other effects units will vary...

And loading the effects send with an effects pedal will drop the level down....
so, use a pedal connected to "send" when you determine that send level.

I find it works better if the proper level is "mid way" on the level control.

[Chuck H]

If you're going adjustable I'm going to second SGM here. I adjust for line level (.775Vrms) with the amp at a nominal setting and the send control halfway up. What SGM didn't iterate was that this works BECAUSE sometimes the amp is dialed down and you may want to increase the now lower send signal for better noise rejection. The problem with adjustable effects loops is that even if you attenuate the send and return analogous you still leave the user to mess up the gain structure with mis adjustment. I did some work with Dean Markley a few years back when he "re issued" his CD amp line. The only complaint from customers was that they were confused by the adjustable effects loop. Even experienced players had a lot of trouble with it.

For this reason I'm going to side with RG now and say 'shit can' any continuously variable adjustment and just go with a line/low switch. Assume the amp will be up loud and target .775Vrms for "line" and 200mV for "low". Switch/adust the recovery stage to re amplify to unity with these voltages through the loop. This way, as long as the switch is set correctly for whatever type of effect/s are in the loop, even if the voltages are a little skewed from the effects units achieving unity gain should be somewhat analogous. There will be a small, unavoidable price in added noise. It shouldn't be enough to worry about.

JM2C

[soundguruman]

Don't forget to plug in an effects pedal..
before you measure the send level.

The effects pedal loads the circuit and drops the level...

TOO LOW a level = lots of excess noise
TOO HIGH a level = distortion from overloading effects input.

1V PP guitar level = 353 mv RMS
If you are measuring RMS, which I hardly ever do, on a scope...on a scope, I am looking at peak to peak, or peak.
Then converting it to RMS if needed...

If that effects pedal is getting more than 353 MV, RMS it's going to overload.

[Mick Bailey]

I like both a send and return level control and line/guitar switch and agree with building the loop with the 'ideal' levels at mid-point. Some pedals give less than unity output and only having a send level can drive the input circuitry too hard and still not give enough output. Having variable gain on the return means that you can tweak the level to get it just right.

[Chuck H]

I like both a send and return level control and line/guitar switch and agree with building the loop with the 'ideal' levels at mid-point. Some pedals give less than unity output and only having a send level can drive the input circuitry too hard and still not give enough output. Having variable gain on the return means that you can tweak the level to get it just right.

That's the circuit that confuses most players. If it's for your personal amp it's great. The average geetar playa doesn't even know there's a difference between line and low level signals. Also, many effects processors have input and output level adjustments and/or switches. In the end the player ends up with a knob on the amp, a knob on the processor and possibly a switch just to set the input side. Then a knob (or digital adjustment) on the processor to set the output and another knob on the amp to adjust for that output. They ALMOST NEVER get it right and end up frustrated because their amp isn't doing what they think it should. Even with only the knobs on the amp it's rare.

[Gaz]

Thanks for the replies - all very helpful. Between reading this thread and going over the old one, I realized that the send control is pretty useless in that it has to be all the way up to achieve unity gain. I used to have the master volumes after the loop, so this could be compensated for somewhat, but it still wasn't ideal because sometimes I like to run the clean channel's master all the way up for max headroom.

The biggest problem with the masters after the loop was the big volume jump on long delay repeats when switching back to clean channel (because naturally the master volume is set higher with less preamp gain).

I wanted to move the masters before the loop to avoid that problem, but it further confuses things because the send level varies with the master volume settings. This is done is the Peavey 5150, Orange Rockerverb series, new Soldanos, and others, so I figured I could do it.

With the masters before the loop, the send control becomes more important because if the masters are turned up very high the send level through the loop goes up too. The send control can be used to throttle the signal to a safe level, but you still run into the issue of less than unity gain if the send level is turned down to anything below maximum, and now there's no masters on the other side to make up the gain.

So, I came up with this solution (which is still a compromise to some degree). When nothing is plugged into the loop, there is a 220K/220K divider going into the recovery stage. The recovery stage has enough gain to compensate so that near unity gain is achieved at the output. Now, when something is plugged into the loop, the top leg of the divider is disconnected, so there is an excess of recovery gain. This allows the Send control to get instrument level output at a mid-way setting (as suggested above!), and also has some headroom if the masters are set low. Conversely, if the masters are set high, the send level can be throttled back and there is still some extra gain on the other side. Also, I figure there is some extra voltage on tap for line level stuff, which usually has an output level that can bring things back to unity.

I'm really hoping this is not the stupidest idea ever

[Enzo]

Unity gain?

The point of loop level switches is to adjust the send and return for whatever is connected out there. If you bridge across the loop with a cord, the switch should be transparent. In other words in the guitar level, the send is knocked down to a low level, while the return has a boost, and then in line level, the send is not reduced and the return gets no boost. The net result is that the circuit takes whatever came from before and sends the same level on past. The changes only apply to the outside world.


Now you can cheat and have independent adjustments on both ends, so you can leave the send up AND use a return boost, or whatever other combination you like.

You say a send level is useless because it must always be maxed for unity gain. That may sound right until you want to feed the loop into some pedal that will be hopelessly overdriven by a line level signal.

[Gaz]

Unity gain?

The point of loop level switches is to adjust the send and return for whatever is connected out there. If you bridge across the loop with a cord, the switch should be transparent. In other words in the guitar level, the send is knocked down to a low level, while the return has a boost, and then in line level, the send is not reduced and the return gets no boost. The net result is that the circuit takes whatever came from before and sends the same level on past. The changes only apply to the outside world.


Now you can cheat and have independent adjustments on both ends, so you can leave the send up AND use a return boost, or whatever other combination you like.

You say a send level is useless because it must always be maxed for unity gain. That may sound right until you want to feed the loop into some pedal that will be hopelessly overdriven by a line level signal.

Enzo, I was probably misunderstood because the bit about the send level being useless was gleaned from something you said in the other thread I linked to. I agree with you about needing to knock back line level for some effects, but I am currently trying to design the loop for instrument level pedals so that the most it can spit out is 350mVrms.

If I have it set so that with nothing in the loop, and the recovery stage boosts the signal back to what it was before the loop, then what happens if I turn the send down? There's nothing to boost it back up to unity. That was the point you made to me in the other thread. But now with the masters before the loop, the send control makes more sense.... unless that's what you were saying! Sorry, I literally tend to get your points about a year and a half later... I wish I were joking about that.

Can anyone link to a schematic where a switch is utilized for different send and return levels? Something must be wrong with my idea/drawing above or my wiring because right now the nothing-in-the-loop signal is louder that plugged into the loop.

[Chuck H]

There is no panacea. A loop that's easy to manage, provides the lowest possible noise floor, works for both line and low level signals, doesn't require adjustment when the amp is adjusted, etc., etc. Not gonna happen! Here's what I did on my last build...

I set the loop for line. Send is at line level when the amp is cranked. The return amplifies this up to where the signal was before it was padded for the loop. No adjustments, switches or any other confusion (this was for a customer ) Just "send" and "return" no matter how the amp is adjusted, as long as the effect in the loop achieves unity, the result is unity. That is, I can put any in/out unity effect in the loop and the amps volume is the same if I remove that effect or turn it off. As I mentioned before, this does come with a small added noise consequence. If the amp is set up clean and low the drive voltage may only be a third of a volt going through a circuit designed to manage a full volt. So the noise floor is relatively high. But the thing is there really isn't much added noise. A fraction. This is the simplest and easiest to manage loop. I just tell the customer that guitar effects go in front of the amp (period). Actually there are a lot of pedals that will handle line level. Most EQ's, delays and chorus's. Most of what you might run in a loop actually. Unless you want to put distortion boxes and wha pedals in the loop it may not be a problem. Anyway... That's another 2C... (I'm in for almost a nickel by now!)