Crazy Power Supply Noise (One Spot)

Started by Fancy Lime, November 08, 2017, 05:26:19 AM

Previous topic - Next topic

Fancy Lime

Hi everyone,

when recording the sound sample for this:
http://www.diystompboxes.com/smfforum/index.php?topic=118966.msg1109620#msg1109620
I had a noise problem, which I do not fully understand. As long as the circuit (on breadboard) is connected to anything with proper ground, like an amp, there is no noise. But when it is directly connected to the computer audio interface (laptop, no connection to the house main ground), there is a crazy loud humm. Much like a ground-loop humm but there is no ground, much less a loop. When I touch the strings and something grounded like the radiator, the humm is gone. So for the sound sample I could avoid the problem by just routing the signal through the amp but that's hardly a satisfactory long-term solution, now, is it?

I had noticed the humm in the "direct to computer" setup before, but much less loud. I attributed it to the computer or computer power supply radiating electromagnetic fields into the breadboard because I never noticed it on boxed up circuits. However, this time I did some experimentation and it turns out, that the humm is independent of the computer. It also occurs when the computer is off and plugged out and a battery operated headphone amp is used.

Any ideas what this might be? It seems to me to have to do with the lack of an absolutely referenced ground potential. But why would that be worse on a unity gain circuit than on high-gain overdrives (where the problem was much less noticeable). And how would shielding of a box help (if it indeed does, which still needs to be systematically investigated), if that were the problem?

I am using a One Spot. Not the current version but an old one. It says "SN515 WK0227" on the side, which I assume to be the serial number and production week. I have not yet tested if the problem goes away with other power supplies but I doubt it.

Any help is greatly appreciated,
Andy
My dry, sweaty foot had become the source of one of the most disturbing cases of chemical-based crime within my home country.

A cider a day keeps the lobster away, bucko!

antonis

Andy, does the problem only appear when laptop operates via mains power supply or it also appears with battery operation..??
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Rob Strand

It's a common problem for ungrounded systems.
The problem can be made worse/better with different layouts.

Try placing the unit at the centre of  a sheet of aluminium foil say 300mm x 300mm and connect the foil to the circuit ground (ie. 0V).

The next step would be to wrap the foil around the unit.

Obviously you should be touching the sheet or the 0V to do the tests.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

R.G.

As noted, this is common in systems with no real ground connection, where all the interconnected units are two-wire only. It can even come from radiation from the lighting, which is often fluorescent.

The simplest solution is to apply a real ground wire to any one of the units, perhaps with a clip wire, or some such, for the time the rig is used for recording. This lets you get on with your recording, and spend the time between projects pondering the significance of what "ground" means in this setup.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

Fancy Lime

Hey guys,

thanks! So it's likely not a (major) design flaw then, is it? That was my main concern.

Let me see if I understood that correctly: One could make a tiny "ground referencing box". I'm thinking: input jack, output jack directly connected with nothing else going on other than a cable that plugs into a power outlet and references the rings of the jacks to ground (while of course leaving the phase and neutral wires disconnected and well isolated on both sides of the cable). That seems like a useful thing to have if that is a common problem, especially for mobile recording where we often have no real ground. BTW, I checked, connecting to any large electrical mass (a person is enough) kills most of the noise. Not quite as effective as the house ground but close, suggesting that the on-the-go-recording musician could probably just hammer a large nail into the moist ground and connect to that (which is pretty much what house ground is anyway, a sort of tent ground, if you will).

Another way, probably safer for people who are not comfortable messing with house mains power line connections would be to have a third jack instead that goes to a grounded amp and has only the ring connected to the other tho rings (leaving the tip to the amp disconnected or connected to its own ring via a 10MΩ resistor or so).

Thanks,
Andy
My dry, sweaty foot had become the source of one of the most disturbing cases of chemical-based crime within my home country.

A cider a day keeps the lobster away, bucko!