How to calibrate an FX loop

[Gaz]

I've built a cathode follower buffered FX loop, but am unsure about how to calibrate the send level. The loop does have a send level control, but I want to adjust it so the max output is sensible.

Is it correct that I should set up the loop so that with MAX signal the send spits out 0dB? If so, can I simply measure the voltage out with a true RMS meter and convert to dB?

Thanks a lot.

-Alex

[Steve Conner]

I believe the usual way is to turn the send knob up until your effects units start to clip, distort or otherwise show signs of discontent, and then back it off a bit.

Stompbox level is nominally -10dBV, that is to say about 300mV RMS, or a few volts peak-peak: roughly the output of a Les Paul played with vigour.

But nothing is set in stone, the actual headroom available inside effects units varies widely, which is why most amps include the send level control, so that the player can optimise the level for his particular FX chain.

It wouldn't hurt to mark the -10 and +4 positions on your send level knob so you know where you are. +4 is 1.2V RMS, if I remember right.

I've owned rackmount effects units with input level meters and knobs of their own.

[Enzo]

Calibrate is not a word I would usually associate with levels in a guitar amp. Close enough is close enough.

If you have made a buffered send with level control, then have you also made a variable recovery stage also with level control? Whatever you start with, if you dial it down for a loop, you need something to dial it back up to whatever it was before your buffer so it can continue through the amp.

[Gaz]

Ok, thanks. To be more specific: With the preamp and master volume (which is before the loop) and send control set to max, I'm getting about 3.5Vrms at the output of the send. Since I can pre-set that max voltage output to any level I'm just wondering if I should lower it to 1Vrms dB (0db). I just don't know if having to over 1Vrms is necessary or if it may even be harmful if I plug it into something that is designed for a much lower input.

[TransLucid]

3.5vrms seems reasonable to me. Now if it was 100 volts (not difficult to obtain), that could cause problems.

[Enzo]

There are a couple kinds of FX units. SOme are designed to plug a guitar into. Others are line level things. If you send that very hot 3.5v signal into some floor pedal, you may have a hard time not overdriving it to the point of uselessness. Or not. The point I tried to make above was, what do you do when you send your signal to some low level device that cannot return the signal at that level? If your circuit has 3.5v of signal at that point and you are NOT using the loop, then the next stage expectsw the 3.5v. But if your floor toy can only put out a volt or so, then your amp will only see that volt or so continuing on.

[Gaz]

Enzo, the loop has a send control to turn down the 3.5vrms, and there's a recovery stage to boost the signal back up to the proper voltage as if the loop wasn't there at all.

What I can't seem to find an answer to is how to properly design the loop. I know I can just turn down the send control, but I'd just like to know how they do it in commercial gear. Do they set it up so that the max voltage out under max signal conditions is 0db? Or is there some 'headroom' for when the input signal is lower? The latter would be something like I have now.

[Chuck H]

Level adjust effects loops built into amps is sort of passe anymore. It was sort of a late 80's early 90's gadget. It changed about the time rack mount effect processors became available. And all of those have input and output level controls built in. So all you really needed to worry about was a nominal level to drive pedals. Most pedals are happy to a bit over line level. With the exception of distortion pedals, which have no business in an effects loop anyway. Compressor and noise gate pedals are hit or miss at line level. Otherwise it seems to always work out IMHE. Delay, chorus, EQ, etc. all seem happy at line level. So... Most of what you see now is the plain Jane series effects loop. Most are designed to put out line level. Some processors and pedals can take more than line level but seem to work fine at (and are probably designed for) line level. If you are going to have an adjustable output I would think that about 2V max is fine. That would preserve the loops ability to properly drive an effect even when your using a whimpier clean tone and such. And to reitterate what Enzo has been pointing out, If you design the amp, and it sounds good with the loop NOT in service with , say, 6V on the grid following the loop, then you want your loop recovery stage to be adjustable such that any likely signal from the send, probably from 150mV up to 1.5V, can then be recovered and adjusted to 6V. And to reitterate again... If, for example, you get that perfect 6V with the loop sending 1V and you then plug in an effect that can only take 300mV, without an adjustable recovery stage you will have no control over the voltage reduction on the following stages.

If by calibrating you mean that you want one knob that adjusts both the send and return levels to be optimum for a given effect, that could be a little tricky to design. But it should work with 99% of the effects out there and has the added benefit of the user not being able to screw up the amps performance by having the send and return levels improperly adjusted. Which was a bit of a problem with many such loops in the late 80's and early 90's.

[Gaz]

Chuck, thanks a lot for the lengthy response - it's very helpful.

If, for example, you get that perfect 6V with the loop sending 1V and you then plug in an effect that can only take 300mV, without an adjustable recovery stage you will have no control over the voltage reduction on the following stages.

Yes, that's a good point, and I hadn't gotten that far yet. So with that said, do you think it'd be a wiser choice to move the master volumes after the loop? Right now I have it set up like this:

[J M Fahey]

I'd basically try to be around 300mV RMS, as Steve said, which most floor pedals can handle.
+4dB/1200mV RMS are usually meant for rack mount studio processors, which rich Guitar players could afford.
Just look at the rack units favored by them; but it's not usually the case for us mere mortals.
Floor pedals expect a guitar straight into them, so don't go way over what a guitar can provide.
Examples:

This is what a regular guitar player can use, most are standard floor pedals and will be happy with, say, 300mV:



This is what Keith Richards uses live: *most* of them (although not all) Studio level stuff, which could be happy with +4dB/1200mV levels.
Funny thing is, Keith's Tweed Twin amps do NOT have effects loops , so some interfacing is made (look at the homemade "mystery" rack units) to allow that.
Clearly the Guitar sound, processed or not, is fed straight into the Twin's inputs.

[Gregg]

Since there is a possibility to overdrive an effect with line level signal I would "calibrate" the loop max output level at -10dB to be on the safe side and have the recovery stage built accordingly. If you have a Send level pot then you can "calibrate" the loop to 0dB and adjust level depending on the situation.
You'll get the max level from your preamp when all pots are at max (where most influence on level will have the mid control, then treble) but in real life nobody is playing like that so this also is an arbitrary decision.

[pdf64]

The 'evil twin' had quite a nice fx loop with a switch to maintain unity gain at 3 signal level settings, see http://support.fender.com/schematics..._schematic.pdf

[Gaz]

Thanks for the input Gregg, and the schem pdf. Haha, I knew you geeks (I use that term lovingly) would bust my balls on the "calibrate" thing, but that's why I bring certain design questions here.

What I really should have titled the thread was: "Effects loop design procedure." Regardless, this thread has helped tremendously.

After a bit more experimenting and re-reading the posts, I came to a few conclusions.

Having the Master volumes -> CF Send -> Send Level was a bit redundant and made adjusting levels too finicky. I replaced both masters with a fixed voltage divider dropping the level down to approx. 350mVrms with max preamp signal and Send Level turned all the way up, then put the masters after the recovery stage. This way gave more control over the recovery stage

I was using a 12AU7, but after dropping the Send level down to pedal-friendly levels, the recovery stage did not have enough gain to get back up to the level that was present before the loop was installed. A 12AX7, however, did the trick! The only downside is that I really wanted to use a 12AU7

I'm gonna keep playing around with it...

Thanks all.

[pdf64]

If the loop can be bypassed, I don't see what the rationale is for using tubes (rather than silicon devices) for the output buffer and loop recovery (unless it's intended for all tube fx(!)).
Pete.

[Chuck H]

If the loop can be bypassed, I don't see what the rationale is for using tubes (rather than silicon devices) for the output buffer and loop recovery (unless it's intended for all tube fx(!)).
Pete.

You need to preserve the purity of the all tube effects