Getting rid of console hum via the PSU?

[living sounds]

My console (heavily modified 200B) still shows too much hum, even after upgrading part of the PSU, using bigger and better caps on the power rails everywhere, doing the grounding mod etc. Now my thinking is that the problem can be solved by just making sure the PSU delivers even cleaner DC. I've had success with simple diode bridge/RC filters before the PSU section on my GSSLs. The console's PSU should be capable of handling the current requirements (I've upgraded the op amps, but they're not power hungry types and don't consume more than the PSU can take).

Does this make sense? I don't see how I can do any more changes to the grounding inside the console. Or would just buying a better PSU be the way to go?

Thanks!

 

[radiance]

How much AC ripple do you measure on your DC? Have you taken out all channels?

 

[JohnRoberts]

I can't offer a simple answer, but that won't stop me from replying.

It seems logical that reducing ripple/hum from the power supply should reduce it in the output, and this is the brute force approach.

I would suggest dividing this down into a two part problem. #1 is reducing the hum source, #2 improving the hum rejection of the circuitry.

For example, you mention increasing the value of PS caps everywhere. A casual analysis suggests that this should stiffen up the power supplies and reduce hum, but that ASSumes ground is some infinite current sink. In reality grounds have impedance too, so using bigger caps everywhere means more ripple current flowing into and through the sundry ground paths. This could increase hum voltage on some internal ground. Note: this is a hypothetical example, and I am not suggesting this is your actual hum issue, but I offer this as an example of how this needs to be thought about, chess not checkers. it is worthwhile to dig deeper and try to understand exactly where and how the hum is getting into your audio. 

How is the hum level in a channel direct out, compared to a bus output? How is that same bus output with 1 channel assigned vs. tens of channels assigned (muted vs. un-muted)?  etc.

The circuit topology is probably impractical to make major adjustments to, so reducing the source of the hum may be the easier path to pursue from this point, but IMO it is worth trying to understand where the hum is actually getting in.

JR 

 

[living sounds]

Thanks alot!

I've just made some measurements, but couldn't (reliably) detect any ripple with my modest test equipment (multimeter and basic scope). The hum gets successively worse with more channels connected, but the worst offenders by far are the aux sends located next to the master bus, even though they are connected via the same star ground scheme joining all channels and busses.


The upgrades to the power rail caps have been done in accordance with Jim Williams' suggestions (who routinely mods these consoles).

As far as the spectrum is concerned 50 Hz is the most prominent peak, but 15o Hz is almost as much, 250 Hz is the third highest peak. 100 Hz is a lot lower than even higher frequency harmonics.

I haven't checked single outs, since there are only unbalanced inserts. But from tests I think the hum originates mostly from grounding between channels and busses.

Time to work on the PSU?

 

[pucho812]

I would start off with the PSU and just the center section, then add a channel  at a time and see when it really starts to be noticeable.

 

[living sounds]

I would start off with the PSU and just the center section, then add a channel  at a time and see when it really starts to be noticeable.

There's hum above the noisefloor present even with only the master bus connected.

 

[pucho812]

has the center section been recapped? does it need a recap?

 

[living sounds]

No, I've upgraded, recapped, modded the whole console as well as the PSU. It works and sounds great - minus the hum.

 

[pucho812]

ground loop?

 

[living sounds]

ground loop?

I guess so. But apart from the grounding modification I've already implemented described here

http://www.tangible-technology.com/articles/200b.html

I don't see what else I can do without major changes to the whole console structure. That's why I'm looking at the brute force approach of modifying the PSU some more (or replacing it).

 

[JohnRoberts]

Thanks alot!

I've just made some measurements, but couldn't (reliably) detect any ripple with my modest test equipment (multimeter and basic scope). The hum gets successively worse with more channels connected, but the worst offenders by far are the aux sends located next to the master bus, even though they are connected via the same star ground scheme joining all channels and busses.

There may be useful data in that...  Does the hum increase linearly with the number of channels assigned, or by power law (square root of sum of squares)? i.e are 8 channels 8x the noise of one, or <3x. If noise build up is linear, the problem seems referenced to the input of the sum amp (ground difference between master section ground and all the channels). If the build up is power law, that suggests the noise is incoherent in the channels. 

This test is a little unreliable with hum, since it will generally not be random phase as it generally comes from only one source and therefore will sum linearly due to a constant phase relationship.  This makes hum a little (a lot) harder to localize than hiss.

The upgrades to the power rail caps have been done in accordance with Jim Williams' suggestions (who routinely mods these consoles).

Successful hands-on experience is a good thing. That said there may be limits to what is possible from the basic design.

As far as the spectrum is concerned 50 Hz is the most prominent peak, but 15o Hz is almost as much, 250 Hz is the third highest peak. 100 Hz is a lot lower than even higher frequency harmonics.

Not unusual for PS rectification related noise.

I haven't checked single outs, since there are only unbalanced inserts. But from tests I think the hum originates mostly from grounding between channels and busses.

Time to work on the PSU?

or the master section ground.. 

If that master section ground was the algebraic sum of all the channel grounds there should be a first order cancellation of individual ground differences.  So instead of brute force, generate a new ground that cancels out the differences. If for example this synthesized ground was the sum of 24x 4.7 ohm resistors the actual impedance would only be a couple of ohms so not a thermal noise concern.   

This is just shooting from the hip and based on your premise that the hum is in ground difference...  If Jim Williams has tweaked these out, he may have measurements of what to expect and what is possible.

JR

 

[ruffrecords]

ground loop?

I guess so. But apart from the grounding modification I've already implemented described here

http://www.tangible-technology.com/articles/200b.html

I don't see what else I can do without major changes to the whole console structure. That's why I'm looking at the brute force approach of modifying the PSU some more (or replacing it).

I have read that article twice now and I still do not understand how the ribbons are supposed to be cut.

Cheers

Ian

 

[nielsk]

Is there a patch-bay interfaced to the console?
From the posted link it looks like they have tied the chassis ground and 0v grounds together at each module, this may be a problem.

 

[living sounds]

Thanks again! The console is currently in use, but I'll do a thorough test soon.

Everything is indeed connected via patchbays. I'll test if disconnecting it completely will make a difference as well.

Now the PSU is a little more complicated than the usual bipolar one, with the transformer using seperate windings each for positive and negative voltage. On the negative rail there is a residual +13V after rectification, according to the schematics (attached here). So I guess an additional diode-bridge-CRCRC filter before the diode bridge would have to look different to the one used for the GSSL.

 

[JohnRoberts]

That PS is a little odd... They are making the minus supply with another positive regulator, that's why it has the weird unregulated supply...   

JR

 

[ruffrecords]

ground loop?

I guess so. But apart from the grounding modification I've already implemented described here

http://www.tangible-technology.com/articles/200b.html

I don't see what else I can do without major changes to the whole console structure. That's why I'm looking at the brute force approach of modifying the PSU some more (or replacing it).

I your hum is due to a hum loop then no amount of PSU improvement will help it.

Cheers

Ian

 

[JohnRoberts]

ground loop?

I guess so. But apart from the grounding modification I've already implemented described here

http://www.tangible-technology.com/articles/200b.html

I don't see what else I can do without major changes to the whole console structure. That's why I'm looking at the brute force approach of modifying the PSU some more (or replacing it).

I your hum is due to a hum loop then no amount of PSU improvement will help it.

Cheers

Ian

Unless you reduce the magnetic field by say substituting a torroid for an EI transformer. I suspect magnetic loops are overstated as causal, more likely contamination of common ground paths. with non signal currents.

JR

 

[gar381]

Check your mains voltage for any sag.  Soundcrafts of that age hated Mains sag and they would hum
like hell.  It was not unusual to have to use the 100V setting on the PSU
in the US due to low mains voltage.  Some Models of Soundcrafts
hated anything below 115V.

Best of Luck

GARY

 

[nielsk]

I see nothing wrong with the power supply design. If there is ripple on the rails you should be able to see it with a scope, and you can measure it with a VOM.
It always helps to reduce the variables, get it to where the only thing interfaced to the console is the speakers, ideally with a transformer balanced interface (possibly even temporarily wire in a 1:1 transformer to isolate the speaker's amp from the console if the console has active outs, as I'm pretty sure it does). If you still have the hum, and you can't see it on the rails, there is most likely a ground loop somewhere in the console.
I'm still suspicious of tying the chassis and P/S ground at each module...

 

[plumsolly]

Where is the power supply in relation to the console? I was a able to reduce the hum in my Auditronics console by moving the power supply further away from it physically...


Ben