I'm probably being a heretic here but I typically float the connectors from the chassis so have to connect both ends of the shield. Using the chassis as a conductor is a fine way to create ground loops.
G²
It sounds like you've already got your mechanical layout done and are running the wires back and forth/front and back through the chassis, but there are ways to substantially shorten the signal path. Selector switches on the front could operate relays in the back instead of switching the signals directly. Volume control potentiometers can be in the back, operated by long shafts that come through the front panel (and of course rotary selector switches can be controlled this way too). I recall Heathkit ham radios (SB 300/400 series) doing this (perhaps more for reasons of high voltage isolation), though it relies on there being a direct line of space between the front panel knob and the control in the back.
Yes. Definitely have the input connector ground insulated from the chassis and have its own wire that goes directly to the ground of the circuit the signal goes to. If there's chassis or "ground loop" current (mains current capacitively coupled through power transformer windings) through the signal cable's ground, this should minimize its effect.
I isolate connectors from the chassis as well, but I don't understand how this forces the 'grounding' of the shield at both ends.
By using the signal ground conductor and the shield both as conductors, aren't we inviting a ground loop?
Just curious. Grounding gets complicated........
Yeah but I didn't think about that beforehand. Again, the runs aren't that long, but I do have some other boards that are far away and long runs cannot be avoided. Perhaps I could have done it another way but between the headphone amp and a sub equalizer board and the actual amp boards, the signal would be all over the place anyway.
My case is wooden so it's pretty hard to accidentally ground it to the case. =D Also means I have crappy EMI protection but oh well.
Good point though.
Last edited:
We use 9451 at work. The only difference from 8451 is the foil is bonded to the jacket on 9451 so it's a little easier to strip the jacket.
G²
No. That will possibly create a ground loop, which would actually conduct RF into an amp. Sheild should be connected at only one end. Beldfoil supplies 100% sheilding around the inner conductors. What little RF might get in would show up as common to both inner conductors and should cancel capacitively between the conductors and sheild.
Doc
Doc
I gave in. I couldn't sleep at night with the horrors of running unshielded 24AWG right past the mains transformers and all that junk...and then the same stuff for wiring up the potentiometers...it was plauging my soul...so...
I wound up buying 10ft of Belden 8451. A few feat left over for some other stuff down the road maybe.
And then I bought 4ft of Belden 7787a..which is 3 wire coax supposed to be used for component video (I believe), however, I'm going to use it for my potentiometer wiring and just leave the braided sleeve as "shielding". $3 a foot...better be good. Any ideas on how to optimize the use of this would be appreciated.
I wound up buying 10ft of Belden 8451. A few feat left over for some other stuff down the road maybe.
And then I bought 4ft of Belden 7787a..which is 3 wire coax supposed to be used for component video (I believe), however, I'm going to use it for my potentiometer wiring and just leave the braided sleeve as "shielding". $3 a foot...better be good. Any ideas on how to optimize the use of this would be appreciated.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
I noticed on the Belden 7787a spec sheet that the minimum bend radius is 4.5", so you probably want to get rid of the outer sheath as a start (to try to get a bit more flexibility).
Also, be careful when soldering the centre conductor, as some of the foam plastic insulators between the conductor & shield are not too heat resistant. A lot of the coax nowadays is intended for crimp fittings, not soldering.
A bit late to mention it now, but....
A few times, I've 'made up' a shielded cable by using copper braid shield stripped from cable and threading the (twisted) conductor bundle into the sheath. With a bit of care it's possible to make up a 'wiring harness' with conductors exiting the shield at different points. Use heat shrink to tidy up the ends.
The % coverage isn't enough for critical RF stuff, I suppose, but it works OK for most audio projects where I tried it. I do have the advantage of a few hundred feet of RF coax lying in the junk box.
To connect to the shield(s) the neatest method is to solder a solid conductor to the twisted braid shield, and use that to make your connection.
An older ARRL Radio Amateur handbook is a good source on dealing with coax.
Yes, only connect one end of sheild. Traditionally the source end, but things being equal either will do. Best answer is sheild should be connected to end which offers best groundpath for RF.
The 8451 will actually offer better (100%) sheilding than your fancy new multi coax, Whups, I take it back. I see the coax has a foil wrap.
Doc
Last edited:
ahh, also called 'pseudo balanced'
I would suggest to try all shields connected to preamp
yeah, if all wires have own shield, split the cable, as suggested
and you can disregard the foil
just cut it so it doesn't short anything
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.