PSU ground loop?

[AS74]

Howdy all,

Trying to fix hum from my pedalboard and suspecting a ground loop issue having something to do with my building efforts, I have found the hum still present even powering 'one' pedal..  Disconnecting chassis ground from the PSU is the only thing that stops the hum.
The PSU is a DC-9 Powerbank, 5 output regulated.

Any ideas/thoughts on a fix are much appreciated..

[digi2t]

Are you sure it's the power supply? Have you tried powering the pedals via battery?

[AS74]

Yeh,  I have tried every pedal with battery and all are quiet..
It hums through different amps, different wall outlets.  Also the hum is present with the PSU off.
Disconnecting chassis ground (AC Ground) stops the hum with PSU off.

[AS74]

Ground lift on the PSU a no no?

[R.G.]

Good test procedure - isolate things till you find out the one thing that causes the event to happen/not happen.

Also the hum is present with the PSU off.
Disconnecting chassis ground (AC Ground) stops the hum with PSU off.

Let's think for a minute.
- hum with only one pedal; then it can't be a loop!
- hum when the PSU is off
- disconnecting chassis ground stops the hum with PSU off

It sure sounds like there is some internal AC leakage in that PSU for some reason. Have you tried a different PSU. Possibly that PSU is defective in a subtle way on the AC part of it.

Ground lift on the PSU a no no?

It most definitively is a no-no. If the unit has a third wire ground, don't lift it.

There is a class of AC power wiring referred to as "double insulated" where the AC third wire ground is not present at all. That's because there is an extra layer of insulation to prevent shocks on any single insulation failure inside. This is very much a special case. Anything with a third wire ground must have the third wire ground connected to be safe(r) if a fault develops inside. So lifting the ground amounts to removing a layer of electrocution protection. Bad idea.

[AS74]

Thank you R.G

. So lifting the ground amounts to removing a layer of electrocution protection. Bad idea.

I instinctively knew this, but there is a sh*t load and more I don't know, hence my opening post.

It sure sounds like there is some internal AC leakage in that PSU for some reason.  Possibly that PSU is defective in a subtle way on the AC part of it.

There's not much in the PSU.
The bridge rectifier or 2200uF possible culprits?

Thanks again.

[Paul Marossy]

My bet is on a bad filter cap, which is not smoothing the rectified AC anymore.

[R.G.]

My bet is on a bad filter cap, which is not smoothing the rectified AC anymore.

I'd have gone with that, but he says it makes hum when the PSU is off.

[PRR]

> hum with only one pedal; then it can't be a loop!

One pedal AND one amplifier (I assume).

You have ground in the pedal-to-amp cord.

You have ground through two power cords.

Loop.

It's a tiresome problem in big unbalanced audio systems.

It isn't usually a problem in small systems like pedal+amp.

First step: pedal-power and amp on SAME wall outlet. (At my old place, source on stage and amp in booth injected 3V of lovely 60Hz between "grounds"; even in smaller venues two outlets may have a part-volt difference which matters on part-volt guitar levels.)

If amp has an AC outlet, feed the pedal from THERE. Keep the power and signal lines parallel about a foot apart. Wider increases ground-loop area, closer invites non-loop direct sneakage from power to signal. Unless something is awful odd, this should kill the loop.

Radical steps include audio transformer buffer. (Full-balance/float with a pound of iron cleaned my stage-to-booth ground potential, but that's rad for a stage rig.)

Many small power supplies do not have a 3rd-pin grounding wall plug. Unless they have giant RF caps inside, these should be loop-breakers and fully legal/safe.

I've owned ground cheaters but very rarely used one. Dead musicians mean a long night of paperwork and body-moving. I've even un-cheated the ground-cheat on a film-crew's audio so as "my" musicians weren't taking shocks. (Especially the one who signed my pay-voucher.)

[AS74]

I've measured around  0.6ish V AC from the regulators heatsink, being used as bracket to chassis, to AC ground. Normal?

No connection here removes the hum. So I believe insulating is in order.

[darron]

it's possible that heat sink is connecting your true earth from the wall/ground to your isolated negative output, which then gets reconnected back up with earth again through your amp.

don't zap yourself pulling things apart!

can you take the heatsink away from the chassis earth and put one of those aluminium independent heat sinks on there?



seems like a bit of a design flaw?

[AS74]

I've hopefully solved the prob by cutting the bracket/heat sink, leaving a 5mm gap between heat sink and now separate bracket thus hum is gone..

Do you s'pose the designers were relying on the chassis for heat sink too? 

[darron]

i'd imagine so.... i think current draw on most LM317t style regulators is > 100mA, but then with adequate heat sink it can go above 1A. that's why i suggested maybe putting on a small aluminium one unless you are only powering a few pedals.


i think the package allows heat sinking with earth. though the IC is design to be usable with audio, the manufacturer can't always tell how people will design around it... for a battery charger or something it should be awesome. like i said, seems a BIG design flaw by the PSU manufacturer...

[petemoore]

  Tap hammers get tried a lot, sometimes even create the desired effect.
   The big sledge is what smashes 'em all beyond rearing their ugly heads again mostly.
  :
  If PS doesn't compare to battery goodness, the battery takes over.
   When the battery is finished: the stepped AC from a floating secondary, filter/regulated/filtered which does compare takes over.
  This proved the Short sweet direct route to 'actually finished with PS/loop noise' ..every time so far.
   I tried some workarounds.
  Otherwise this = maybe that, but only when...+...
   Cut all the ground loops, float and filter the supply...cuts it to bare minimum it can be and still work.
 

[R.G.]

seems a BIG design flaw by the PSU manufacturer...

For a number of reasons, the best and brightest engineering talent tends not to be doing power supply design. PSU design flaws are correspondingly much more common than they should be.

[R.G.]

> hum with only one pedal; then it can't be a loop!
One pedal AND one amplifier (I assume).
You have ground in the pedal-to-amp cord.
You have ground through two power cords.
Loop.

Depending on whether the amp is grounded or not. Many are not, despite years of teaching people that third-wire ground is a good thing. And the amp may well have been ground lifted as well, despite years of telling people not to lift ground by anyone who thinks about electrical safety.

But yes, a loop is possible.

[Paul Marossy]

Huh I would have never guessed that heat sink thing was the problem - good call! I learn something new every time I come here!