Ground lift (cable shield) on Dry/Wet fx loop - should I lift send or return?

Started by bensaddiction, October 19, 2010, 05:19:01 PM

Previous topic - Next topic

bensaddiction

Hey guys,

I'm designing a ground control box for my younger brother. It's a bit in depth but the heart of it consists of 2 true bypass loopers with buffered dry/wet to switch to channels of studio rack fx in with his guitar setup (midiverbs, tc delays, more to come hopefully) The box is also going to have a tuner/mute (which really is an A/B box when you think about it) and 2 switches to control channel & reverb on his Fender amp. Also one of the FX loops has a switch and pot for feedback on the FX send/return as well as a nifty little 'Freeze' button i've dreamt of for the feedback circuit. All this back through a multicore from the foot controller to amp/rack fx

Anyway, I'm a little worried that ground loops could raise there noisy heads so I'm going to put in a few ground switches where appropriate.
I'm going to have a switch to lift the tuner out (if we want to try A/B switching amps during a recording rather than as a tuner out)
I also think it would be wise to be able to lift the FX sends in case some hum originates from there.


So:

Is it normal to put in a switch to lift the send or return or both (I'd rather just do 1 or the other but if anyone here has experience and recommends both then I'll do that). To be clear here I am discussing lifting the shied ground from the ingoing/out going cable - not power ground.

I'm not going to consider transformer isolation at this stage, mainly because the size, weight and cost elements, also transformers colour the sound in some fashion especially the affordable range (even expensive lundahls do a little bit)

In order to ground lift effectively - Im going to have to isolate the jacks from the chassis yeah? So that the switch actually blocks the ground to the cable's shield. I figure I'd just use some rubber washers on the inside and outside to do that like I see on some other pedals. Is that the correct thinking?


Also, on a related grounding issue. I'll try to make sure I follow a star grounding pattern for the active and other elements of the circuit through to the bottom of the chassis around a bold in the base. Is it also recommended to use rubber washes on things like pots and switches so that the grounding happens in the one spot as well, or is that going too far/incorrect? (I only ask because this is what I see on my Crowther Double Hotcake and it seems like its the right way to really cut down on potential loops yeah?)

any answers to any of those questions would be great guys. Cheers

defaced

Typical to ground to the low impedance (send/incoming) side, so lift the return/outgoing side.  This is common in amps where shielded wire going to pots can cause ground loops with the chassis. 
-Mike

bensaddiction

Thanks defaced!


Only I'm a little confused by that sentence...  you said: ground the- "(send/incoming) side, so lift the return/outgoing side"...
Because:
the send is actually the outgoing (Dry signal going from controller to FX loop) and the incoming is actually the return (the FX loop coming back into the controller box). Unless I'm using the terms from a different perspective.

Or did you mean to use the terms in this fashion: ground incoming as input (like from the guitar), send (as in to the FX loop). And then lift the outgoing (to the Amp) and return (from the FX loop)


If its the second interpretation then it sounds like solid reasoning to me, and I have noticed that with grounding to inputs on other designs so thanks.

defaced

I'm an amp builder, so I think of it from the amp's frame of reference.  The send jack on the amp is the input jack to the effect, that's the side to ground.  So yes, ground inputs.  That's probably a better way to think of it because it's more true to the "ground the low impedance side" reference I've read many different places. 
-Mike

bensaddiction

Ah! thanks, that what I was thinking but yeah terms like this need a POV don't they.

cool.

Also - since your an amp man, how does this sound:

The foot controller is going to be big in the end with two rows of footswitches and some pots & switches above that as well. As a result its too big & different to a hammond box and I'm just going to use an aluminium faceplate and build an angled wooden enclosure.

Normally grounding and shielding is fine in the stomp boxes I've made as I've had no troubles with not sticking strongly to star grounding patterns and not really had to worry about shielding my wiring as they are inside the all metal case and don't need to travel far.

So do you think using plastic washers to isolate the top controls (pots, jacks, toggles and footswitches) so that I can ground in just one place on the chassis is wise, or a waste of time. Some of the wiring involves sending the signal back across the length of the box so I've decided to try shielding at-least all the main wires that I can, particularly the wires to the footswitches which will always be in the circuit even when bypassed. I've bought some cheap RCA cables as I read a post saying that they are a good way to get cheap and usable shielded wire to cut up. I've also bought some double conductive aluminium tape to try to turn the inside of the wood and the faceplate into as much as a perfect faraday cage as I can. (I'm not going to even allow it to run on batteries so that the cage doesn't accidentally stop functioning during a battery switch.)

Is there anything else I could or should be doing. Also, when shielding wiring between components does it work to just leave the shield in the middle and only attach with the conductor to where it needs to go, or should I be grounding one end of the shield aswell as if they were actually guitar cables? I'm really trying to make a routing box that is really powerful and with no hum & I'm trying to see if that can be done without having to somehow find an all metal case. Considering you build amps which no doubt can be more susceptible to grounding and shielding issues - am I trying too hard or not hard enough?

defaced

Ugh, ground schemes.  Everyone has one, and they're all amazing.  To follow suite, mine is, of course, the best there will ever be and you should force your project to conform to it.  

Not.

But seriously, some of my approach is because it's easy to think about what's happening, and if I can visualize it, I can build/fix it later.  Efficient/inexpensive/whatever is great until you have to fix it.  Then it's usually a grade A pain in the ass.  

Here are my rules:

  • ALL grounds that carry signal are grounded via a wire.  Nothing goes through jacks for pot casings. (In other words, don't do what most effects guys do and use the box as your signal ground return)
  • Isolate all jacks from the chassis using fiber washers (check Mouser, Keystone it the MFG of these.  They're in the catalog too)
  • The the chassis ground connection happens at the input, and aside from the safety ground in AC mains wiring, it's the only place to ground to the chassis. Use a wire to make the connection between the input jack and the chassis.
  • If the chassis is painted, pot casings get a bus ground across the back and that gets grounded.  If the chassis is bare, the pot casings ground through the nut and life goes on.  This is just to shield the pot, so it doesn't make a big difference.  You can also use star washers to bite through paint. 
  • All decoupling happens locally (preamp filter caps namely, but that's not a worry of yours)
  • I prefer to use a bus wire when possible.  In an amp there are currents to manage and there's a linear path so it works out well.  In your case, that might not be a good idea.  I have also used a modified bus/star arrangement where a bus picks up stars along the way.  Same idea, slightly different execution.  Still probably not a big concern. 

Comments:

Foiling the box is good.  Make sure the panels of tape conduct to one another.  I've never played with the stuff, but I have read the adhesive isn't conductive.  

If is no active circuitry in your box, you probably won't need shielded wire.  Shielded wire is to prevent pickup from noisy sources.  So if your chassis is grounded, you are in effect in a Faraday cage, and it should be clean from outside noise.  So you're left with noise made inside the box, which if it's just re-routed signals, and if they're not noise emitters, then shielded wire shouldn't be necessary.  To that end, Faraday cages are passive.  As long as there is a ground (Earth) connection to the cage, they're working; I'm not sure where the battery thing came in at.  

eBay (or Apex JR) is another good source of shielded cable.  I just got 25' of teflon coax for like 12 bucks shipped on eBay.  RCA cable can cause you nightmares if the inner conductor shorts against the shield.  This is because the isolation is low melting (like PVC insulated wire) and can happen during soldering.  This is why I've standardized all of my wire, as much as I hate stripping it, to teflon coated.  600v rating and doesn't melt?  Perfect for amp building. 

QuoteAlso, when shielding wiring between components does it work to just leave the shield in the middle and only attach with the conductor to where it needs to go, or should I be grounding one end of the shield aswell as if they were actually guitar cables?
Depends.  In a guitar cable, the shield carries the ground signal and acts as a shields, so it gets connected at both ends because the signal circuit must be complete.  In amps, the shield is just for noise reduction, the ground buss/star/whatever carries the ground signal.  In pure noise rejection applications, one end of the shield is grounded, the other is left open.  Any noise on the shield will be shunted to ground by this instead of dumped into the signal.  

I think I hit everything.  As always, find what works for you and use it.  This is what works for me.  
-Mike

bensaddiction

Cheers mate,

I hadn't thought about the likelihood of the rca cable melting through like that  while soldering so I'll look into coax or other options like you said (or just regular mic/guitar cable). RCA just appealed to me because its thin & easier to manoeuvre than mic cable.


The battery comment was about not also allowing the unit to also power from a battery, making a 9VDC plug the only way. That way once I finish foil taping and making sure all sides of the inside of the box are connected like a cage I can seal the box & never open it. If the guitarist was opening and closing it all the time to get at batteries it might make the foil not touch somewhere.

The tape is advertised to have conductive glue but I'm still going to fold the edges over to be sure.

There are some minor active circuits (dual op amps for each of the two fx loops dry/wet/buffered outs, phase reverse circuits and a discrete buffer for the final output, as well as LEDs to indicate basically everything. But yeah they aren't really making alot of electrical noise. Also the box will also contain a 9vpower filter circuit from a Beavis Audio schematic. Its to filter a 9VDC wall wart to supply to circuitry in this box with an output so it can daisy chain and power his other pedals. So its not a lot of real noisy stuff (from what I can gather), but I'm trying to stay on the safe side.


One last question though - your first dot point about pot casings and jacks. I get the point about not using the box as the signal ground return, but what does this then mean in relation to my question about if I should use fiber washers for the pots, or did that statement mean there not needed and I should just focus on working on the ground at the input jack to chassis?

Anyway thanks for all that, you've set me straight. Now I just need to build this darned thing and then I'll be able to hear if there are any problems.

defaced

Ah, that makes sense about the battery now.  I didn't think about it from a lifecycle point of view, just a theoretical point of view.  Good planning on that one. 

If a pot is connected to the chassis through it's mounting nut/washer, the only thing that's grounded is the pot casing which is just a shield.  If the pot comes loose, the shielding of that pot might get a little sloppy, but because every signal ground is returned using a wire (meaning no bent pot legs up to the pot casing), the true grounding for the signal that pot is working with is always there unless the wire breaks.  So isolating pots from the chassis doesn't really impact your ground scheme, it just impacts how you shield the pot.  
-Mike


defaced

No problem.  Post up pictures of this when it gets going.  It sounds like a cool project. 
-Mike