Why is my snake humming?

[Pbassred]

I tried posing this on the pro audio forum but they don't have the right knowledge to fix the problem. They would just say "buy a better snake" They might be correct but I wouldn't learn anything by doing that.
Here goes:

For live audio I run a "snake" Composed of balanced pairs with a single screen around the lot. An XLR stage box at one end and a break out box to 12 XLR plugs at the other. It works fine in passive mode. cross talk is no problem.

Using phantom power from my Yamaha desk (48VDC down both conductors and back along the screen) it produces mains hum.
Clues:-
*I have another desk and a vocal compressor that also produce phantom power, but they don't make the snake hum.
*The Yamaha doesn't make other leads hum.
*I can't find any shorts or opens between pairs.
*It doesn't make any difference whether I have a Mic conected or leave the line plugged in.
*It does it even with only one lead connected to the desk.
*Lead#4 is worse than the others.

[Frank Berry]

Do you get the same hum when using a single microphone cable rather than the snake?
If so, the yamaha phantom supply may have too much ripple.

[rcavictim]

You are misusing your snake shield. It is there to provide electrostatic EM shielding. It should be an open circuit at one end of the cable, preferably the output end (always ground at the source). You should not ask it to handle current carrying duties. Even if your phantom power supply is hum free having it in series with the shield closes the shield into a loop antenna which can couple stray AC field hum into the signal conductors from magnetic induction.

[peace brainerd]

Wasn't that a b-side from an old ZZtop 45? I forget the year.

[Pbassred]

Frank,
No a single balanced XLR wont create hum. Only that particular snake with that mixer ( so far).

rcavictim,
I kinda follow what you are saying. I was thinking along the lines of an antenna. However if I do understand you, there are 2 things that don't make sence; 1/ The hum happens even when there are no mics connected ( therefore the screen IS open circuit at the stage end. 2/ My understanding of Phantom power is exactly that. Shield return.

blue beard, Its ok to forget.

[Enzo]

Phantom power IS exactly that, shield return. SHields are broken to stop ground loops. Can't have it both ways - if a shield causes a ground loop, you can't cut it and still use phantom. ANd if you use phantom, you can't cut the shield. There is minimal current in phantom, and if the channel input has remotely any common mode rejection, then there should be no fundamental issues. After all, mix desks have been sending phantom down snakes for a long long time.

I am not sure, but from your description, the pairs are not shielded, just the whole cable. SO there is one common shield for all the pairs. That by itself might not be enough, but I am betting the shields are also bonded to the case. That is, the Yamaha may isolate the mic in shields from the chassis, and your snake is hooking them back together causing the ground loop in the desk. Does that make sense? In other words the phantom supply was meant to float, not be grounded. MAybe?

Look in the snake heads. Are the pins 1 wired over to the jack body lugs? If so, that might be it.

Try this. With the snake removed, take a clip lead or something and connect a pin 1 of a channel in to the chassis. If that makes hum, we have a clue. Use a small nail, screw, or piece of wire to fit in the pin 1 socket.

Looking for the bad relation, what happens if you connect only one cable from the snake to the desk? Does it still do it, or does it take more than one? If so, is there one channel in particular - perhaps channel 4 - that seems to be necesary? That is, if you leave ch4 unconnected, does it help? Possibly then one - or more - of the channel strips has an issue with the grounding at its input. OOps, you said it does it with just one. I feel more likely the pin 1 is tied to jack body theory works.

[Pbassred]

The wierdness continues.....
Actually I had thought of the grounded body issue before and started cutting the links. No difference though

You want odd? here is odd. There are 2 unused pairs - completely unconnected. When you touch them Hummmmmmmmmmmmmm . Ok, perhaps not that odd. Induced signal. I tied them down to earth, but that didn't fix the cable overall.

OK then. Proper oddness. I was looking for bad channels. so I hooked them all up and set the channels identicaly. Some channels are worse than others. Then I tried switching on pairs of channels and found that some combinations partially cancel the noise. Switching all of the channels on was bad. but switching some chanels and some combinations off was sometimes better.

Last night I hauled the snake down to a band rehersal and plugged it in to a Makie 6 chan desk. Quiet as a mouse

yamaha phantom supply may have too much ripple

I'm inclined to agree. There is something resonant going on here. The transformer is remote on these and gives 18-0-18 AC (19.6-0-19.6v open circuit). I don't know how it gets 44V DC (actual) from that but it might be a clue. I'm inclined to go in and throw a big cap across it, but I don't know what that might upset.

[Pbassred]

Hang on. 18VAC fron the PSU isn't peak to peak. My multimeter is taking an average. So if it is rectified, the meter would read 18VDC RMS! but the peak must be higher. I seem to recall from collage that adding a cap to the output had the effect of increasing the RMS voltage.

So 18+18 =36V
Less 0.6V x 2 through the bridge rectifier diodes = 34.8V
34.8 / 0.707 (Sine wave) = 49V

Someone like to check my numbers please?

[Enzo]

Which Yamaha mixer do you have?

[Enzo]

Forest for the trees department. I have been focusing on the snake, but what if there is a burnt open ground in the Yamaha? Check from the pins 1 to chassis on the jacks. Then check the phantom suppy to make sure its ground refs are intact.