Ground lift for 3 regular channels and 1 mixed output?

[MrStab]

Hi guys,

i'm in the planning stages for an EQ pedal i've been asked to make with 3 separate channels, each for a different pickup coming out of one guitar. i have no idea where the outputs will end up, and i'd like some degree of flexibility, so i'd like to put ground lift switches or isolation transformers in there. The grounds would be connected inside the guitar (which i don't imagine would be a massive problem if the EQ is early in the chain), but who knows where else they could merge with different house setups and gear.

Simple enough for the 3 channels on their own, but here's the snag: there's to be a 4th output, with an (inverting op-amp) active mixer to combine all 3 channels into one. What now? Would isolation transformers just before each of the 4 outputs do the job, after all active circuitry? Should i just link the grounds together for the mixer then re-isolate at the output?  i assume yes, but thought i'd ask and see what you guys think. At this point, a simple SPST to cut the output ground is out of the question, right?

p.s. the Geofex ABY box makes me think "yes" too, but that only has one input.
p.p.s. all pedals will share one (regulated/isolated) power supply. Non-issue if all signal outputs & PSU are isolated? Any safety risk?

cheers!

-g

[PRR]

You want 4 straight outputs and 4 transformer outputs.

Any or all destinations may be on different grounds. If not, the direct output is marginally cleaner. If not, the transformer output breaks the buzz. Let the user and sound-tech figure what works best in each situation.

[MrStab]

so both, essentially, via. switches? that seems more watertight to me.

thanks!

[MrStab]

Sorry for the bump: the "clean" part has just hit me. Could you point me to any resources about this non-linearity with transformers/inductors in audio? i know absolutely nothing about it, having apparently lived in some fantasy where transformers are ideal components until now.

I've been asked for True-Bypass as well (damnit!!), so signal impedance is probably something i should be looking up here, too. I'll try to read about passive DI boxes in this regard, as i seem to be able to find more stuff i understand about that on Google than the first thing.

cheers!

[PRR]

Water tight?? If you put it in water, it will grow barnacles as well as bypass and other stuff.

You need to draw this out. You've got a pedal, a guitar, an EQ, three (or four?) outputs, each with both straight and transformer options.... I picture more a rack box than a pedal.

[MrStab]

Barnacles add more low end, don'tcha know?

It's a big 190x190mm enclosure with multiple 3-band EQs in it. I think a pedal was requested as there's not a lot of other stuff to justify a rack unit. Everything fits well with ample stomp room. I'll try and get the time to draw something up once I'm off this train.

Sent from my Vodafone Smart ultra 6 using Tapatalk

[MrStab]

I've convinced the requestee to ditch True Bypass. It's simplified, but this is essentially my dilemma:



Each channel is on its own PCB, so the grounds coming from each op-amp symbol represent the board grounds from those. I need to find a way to optionally (or permanently?) isolate GND 1-4, as the signals all share GND-GTR (as well as possible rendezvouses on downstream equipment).

i realise now the lifts for channels 1-3 will need to be made before the signal enters the mixer and subsequently GND-4.

cheers!

[PRR]

I can't believe a mixer without mix-balance pots, so I added them. (Actually I have done that, with three professional actors who self-balanced super well; I don't trust pickups and strings to balance.)

Every place you go out of the box, add a transformer and an insulated jack. See R.G.'s splitter. Start with the direct outputs. If buzz happens, start switching to the isolated outputs.



I do not understand your DC bias arrangements; I assume this detail is TBD.

[MrStab]

thanks for taking the time to draw that, Paul. i think i'm on the right page.

just to clear up what i omitted:

• There's coupling between every stage, with each section using a dedicated divider network (w/ buffer) for biasing. The bottom resistor & decoupling cap grounds from that network in, say, EQ2 are dumped to GND2.
• While i will use an inverting mixer, that part of my diagram's not at all meant to be literal - i just didn't want to make all the channels seem shorted together, so i broke convention.
• As it stands, the "master" volume control from each EQ stage will determine both the output level for that channel and what goes into the mixer.

So it wouldn't matter that the inputs to the mixer are taken from before the isolation transformers? i can't see any issue so long as external devices only see the isolated output, now i think about it more.

R.G's Splitter seems to have a zobel network on the output, but you've left them out - would you say that's a non-issue in this situation? I'm probably opening a deep can of worms with that one. If it makes a difference, each channel has a max bandwidth of 20KHz, with a total possible gain of 17.8 on chosen centre frequencies (inc. recovery gain).

cheers!

[PRR]

Why would the inboard mixer need to take isolated sources?

Review R.G.'s comments on the transformers. There might perhaps be ~~100r feeding the transformers, and maybe 1K feeding the direct outputs, for short-resistance and capacitive loading. The high-end may want an R-C to tame, though maybe not necessary; you can work this out as the build beta-tests.

I don't disapprove of 20KHz response, but for guitar it usually is not necessary; maybe not wanted. (Electric guitar gets really nasty with >6KHz response.) Just because young people "can" hear 20KHz does not mean we want to hear 20KHz from mid-pitch instruments. (Cymbals and a few others the exception.)

[MrStab]

In hindsight, i have no idea why i'd assumed the mixer needed isolation! i didn't think so when the thread began, but the Spanner of Uncertainty somehow found its way into the works.

I'm not quite sure which part you mean on R.G's article but i think i get the gist of what you're saying regardless. I'm most-likely just being an idiot and overlooking something.

Occasionally acoustic players ask me for EQs (such as in this case) and as they're usually foregoing guitar amps in favour of a P.A, desk or interface, unfortunately i have to cover the audible range. In this case there are both piezos and pickups involved. I wish i could kill off everything above ~5KHz!!

cheers!