Patchbay Rewire : Balancing/Unbalancing Hum : Normalisation

[nohatnoswim]

Just moved place and did a rewire.
Power is 3 phase, star system from 1 single socket.
Implemented significant normalization on patchbays for first time
All my balanced gear is whisper quiet.
Same gear & signal paths pre move were all solid.

Problem:
16 outputs of synths going into a Vintage Balancing Rack and then into SSL interface.
Outputs are all humming at various levels ( some of the synths aren't even powered on).
Bypassing normalisation with patch cables doesn't seem to change anything on the few channels it was tested on
Visual of typical path for the Juno 6 as an example in attached diagrams.

No specific dedicated grounding scheme in place - cables into patchbays, and out - grounding on patchbays is only within the plugs and the gear.

Is this
a) The normalisation,
b) The wiring path independent of the normalisation
c) Both
d) Other

Thanks,

NHNS

 

[Bo Deadly]

Not sure I follow everything you're saying. Either way, I would say divide and conquer. Pick one synth that is the noisiest and some sort of amp with a suitable input for the synth but that is also known to be very clean (maybe a USB audio interface). Then unplug everything from the mains socket and plug in just the synth and the clean amp and see if the synth is still noisy. If it is, then now you have greatly narrowed down the problem. If it's not, insert a bay and gradually add more gear until you hear a problem. Better yet, get your USB audio interface going with some software to look at the noise spectrum. Verify that the noise floor is 80dB+ (depending on expectations of gear) and that you don't have any big mains Hz.

It could be that you moved to some place where there's a lot of EMI. For example, if you're in an apartment building and there are power cables in the walls for HVAC or elevators and such, those can emit a lot of EMI.

Similarly if you put a lot of stuff on one socket (yeah you have a star system which is fine but) it could be that it's drawing so much current that it's emitting more EMI than you want. In practice, you are going to need to use different mains sockets and distribute things a little. If gear on different circuits is interconnected using only balanced connections then it won't matter anyway. So if you have some kind of mixer or audio interface, then create a station for that connected to one mains socket. Then have different satellite stations on their own mains sockets with balanced interconnects to the mixer station.

 

[abbey road d enfer]

16 outputs of synths going into a Vintage Balancing Rack and then into SSL interface.
Outputs are all humming at various levels ( some of the synths aren't even powered on).

I strongly suspect a ground loop. Are the jack sockets in said "Vintage Balancing Rack" isolated from the chassis?
Does it hum if you take it off the rack?

 

[nohatnoswim]

I strongly suspect a ground loop. Are the jack sockets in said "Vintage Balancing Rack" isolated from the chassis?
Does it hum if you take it off the rack?

I'm not 100% sure - it's a custom unit with all I/O's on 3 large EDAC's, so it's always been a bit of a challenge to trace back and I've not done it as I didn't need to. It also came pre-wired to TT plugs so I never changed the plugs on the EDAC to TT as it didn't need to be done.

Up to 4 weeks ago it worked fine in my old apartment with a much weaker electrical situation than where I am now, though I believe there are only two differences with the audio cable
1) The normalisation on the patchbay - previously it was managed by patch cables.
2) The plug on the synth end is now balanced into unbalanced on the patch / previously that was reversed.

It weighs about 50 lbs / 20 kg, so it's a pain in the @ss to work with pulling in and out of the rack, plus it's slightly too wide and doesn't go in and out smooth - not impossible, but I've tried to avoid it !

Same with the patchbays - I've 4 TT patchbays in the rack and it's tight. But looks like I@m going to need to remove them to work on them, so be it !

 

[nohatnoswim]

Not sure I follow everything you're saying. Either way, I would say divide and conquer. Pick one synth that is the noisiest and some sort of amp with a suitable input for the synth but that is also known to be very clean (maybe a USB audio interface). Then unplug everything from the mains socket and plug in just the synth and the clean amp and see if the synth is still noisy. If it is, then now you have greatly narrowed down the problem. If it's not, insert a bay and gradually add more gear until you hear a problem. Better yet, get your USB audio interface going with some software to look at the noise spectrum. Verify that the noise floor is 80dB+ (depending on expectations of gear) and that you don't have any big mains Hz.

It could be that you moved to some place where there's a lot of EMI. For example, if you're in an apartment building and there are power cables in the walls for HVAC or elevators and such, those can emit a lot of EMI.

Similarly if you put a lot of stuff on one socket (yeah you have a star system which is fine but) it could be that it's drawing so much current that it's emitting more EMI than you want. In practice, you are going to need to use different mains sockets and distribute things a little. If gear on different circuits is interconnected using only balanced connections then it won't matter anyway. So if you have some kind of mixer or audio interface, then create a station for that connected to one mains socket. Then have different satellite stations on their own mains sockets with balanced interconnects to the mixer station.

I've no USB interfaces, only MADI SSL's with 56 in / 40 out over DB 25's, but agree divide and conquer is the way forward.

My star scheme is pretty clean on that single electrical point, and due to my furman power distribution setup ( I've got 6 furman units all connected to a central sequential boot up furman unit which is in turn connected to a single socket ). I'm not moving things to other sockets cause I don't believe I need to. This setup has worked up to 4 weeks ago in my old apartment and has worked for me for over 15 years as a practice in previous setups, so I don't want to start introducing additional breakers and socket points into the equation if not necessary.

Interestingly, any unbalanced outboard I have (comp's / eq/s) - all super quiet in new setup.
Balanced synths going direct into interface with no DI/Amp, all super quiet.

Seems the problem is the DI box (and it is a super sweet box btw).
Just need to figure out where the hum is introduced as all the lines are exactly the same, so it's somewhat positive that they all behave the same, as opposed to one of them.

And as above post, 4*96 TT patchbays, which are not the easiest to work with for quick changes, especially with the normalisation patches in there, but I've left at least some room to get them out of the rack and to play with them if needed.

I'll be flat on the floor with a soldering iron for the day, will report back on results of experiments !

 

[Audio1Man]

When doing a DEBUG on a system you should have some basic tools. Some of my tools include a 1:1 (600 Ohm), 1:1 (10k Ohm) Isolation audio transformer, AC MAINS isolation transformer along with extra patch cables and my favorite is my HUMMER (12 VAC half wave Shockey rectifier and 120 ohm limiting resistor that give a nasty hum noise) this gives me a 100ma current that I used to drive the ground XLR pin1 (or other connectors types) to search for the noise.

Many time you don’t have a single grounding noise problem.

Good hunting Duke

 

[calaverasgrande]

I've had the same problem for years with many of my synths. Especially the ones I consider 'prosumer' since they only have a two prong power cord or a wall wart.
I've noticed that it often boils down to two machines fighting each other. One device wants to dump noise in a way that the other box is only too happy to pick up.
I ended up getting an isolated power supply designed for guitar pedals that also works for some of my synths. That took care of many of them. Preventing noise leaking from one to another. The last couple of problem children that were not easily powered in this way got placed on seperate Tripplite Isobar power strips. They aren't a panacea, but I've had good experiences with them.
Somebody needs to make an isolated power brick like the CIOKS but just for synths and drum machines.

 

[abbey road d enfer]

Seems the problem is the DI box (and it is a super sweet box btw).
Just need to figure out where the hum is introduced as all the lines are exactly the same, so it's somewhat positive that they all behave the same, as opposed to one of them.

And as above post, 4*96 TT patchbays, which are not the easiest to work with for quick changes, especially with the normalisation patches in there, but I've left at least some room to get them out of the rack and to play with them if needed.

Just to make sure, is this patchbay an addition to your previous installation?
Quite often, for simplicity of wiring, the sleeves are bussed together and sent to the "star ground", which is OK for balanced connections, but not for unbalanced ones.

 

[Newmarket]

I've had the same problem for years with many of my synths. Especially the ones I consider 'prosumer' since they only have a two prong power cord or a wall wart.
I've noticed that it often boils down to two machines fighting each other. One device wants to dump noise in a way that the other box is only too happy to pick up.
I ended up getting an isolated power supply designed for guitar pedals that also works for some of my synths. That took care of many of them. Preventing noise leaking from one to another. The last couple of problem children that were not easily powered in this way got placed on seperate Tripplite Isobar power strips. They aren't a panacea, but I've had good experiences with them.
Somebody needs to make an isolated power brick like the CIOKS but just for synths and drum machines.

Keyboard being unearthed shouldn't be a problem in itself. Arguably an advantage if they have unbalanced outputs (and I guess they do) as it avoids a "Ground Loop" scenario. Although I'm always reluctant to use that term since it covers a lot of different situations. So let's call it an "Earth Loop" here. But the power supply can indeed be a problem if it is 'daisy chained' and introducing 'Ground / 0V Loops' or if there is excessive capacitive leakage to effectively do the same. 'Digital' devices can be a particular problem here - often experienced with particular fx pedals - and your keyboards will be 'Digital' with accompanying clocks / oscillators etc.
And yes - the answer is usually a dedicated and well isolated psu.

 

[Bo Deadly]

But the power supply can indeed be a problem if it is 'daisy chained' and introducing 'Ground / 0V Loops' or if there is excessive capacitive leakage to effectively do the same.

Yup. Daisy chaining things is a great way to introduce ground loops. And by that I mean just connecting one piece of gear to another using any kind of cable like a regular 1/4" instrument cable or maybe even a midi cable (I know midi signals are isolated with an optoisolator but do shields connect grounds together?).

 

[Matt Syson]

Without having read all the posts in detail you are saying your gear was fine in a different place but now it isn't and I am presuming mostly HUM (60Hz) is the issue?
Running from a single wall socket source SHOULD be good but have you considered the environment so 'loops' of power wiring for the lights, Aircon/water heaters etc which maybe radiating a 60Hz magnetic field. Anywhere where a 'live' and neutral return wire are physically separated will radiate a hum field. Sloppy/lazy wiring of 2 way switching for your room lights for example. If live travels one physical route but the return current takes a different path that is a good way to create a problem.
You say you have 3 phases, what are each of the 3 powering and do you have a mixture of 120/230? apliances maybe not even in your room? Is it possible you have unbalanced neutral current that is not following the same physical route as the lives?
Matt S

 

[nohatnoswim]

Just to make sure, is this patchbay an addition to your previous installation?
Quite often, for simplicity of wiring, the sleeves are bussed together and sent to the "star ground", which is OK for balanced connections, but not for unbalanced ones.

No, I've had 5 Bantam Patchbays in my setup for quite a while.
This is just a rewire as I moved place, and in the process I implemented normalisation so I wouldn't have to look at patch cables hanging out all over the place.

Majority of it is fine, with the exception of everything going through this DI.

 

[nohatnoswim]

Without having read all the posts in detail you are saying your gear was fine in a different place but now it isn't and I am presuming mostly HUM (60Hz) is the issue?
Running from a single wall socket source SHOULD be good but have you considered the environment so 'loops' of power wiring for the lights, Aircon/water heaters etc which maybe radiating a 60Hz magnetic field. Anywhere where a 'live' and neutral return wire are physically separated will radiate a hum field. Sloppy/lazy wiring of 2 way switching for your room lights for example. If live travels one physical route but the return current takes a different path that is a good way to create a problem.
You say you have 3 phases, what are each of the 3 powering and do you have a mixture of 120/230? apliances maybe not even in your room? Is it possible you have unbalanced neutral current that is not following the same physical route as the lives?
Matt S

Not quite that simple Matt.
Gear was fine on old config, but because I've implemented Patchbay upgrade at the same time as move, one is likely complicating the solution to the real problem which is the other.
85% of gear is whisper quiet in new setup with new patchbays, only synths running through my vintage amp rack are problematic.

Something interesting I did notice was for example on my Juno 6, only the Left is humming.

When I say humming, I mean it's so loud it's clipping on some channels - it's a massive noise.

I'm going to play with my Mopho first and see if I can get a clean signal as it's on the end of one of the patchbays and easiest to access.

I'm in a 230v country, though I do have 1 110v Furman powering a few 110v items, also most of which are whisper quiet.

The wiring in this place is likely solid, given there aren't many countries with more regulations about doing absolutely anything than this one !

 

[Bo Deadly]

Something interesting I did notice was for example on my Juno 6, only the Left is humming.

When I say humming, I mean it's so loud it's clipping on some channels - it's a massive noise.

That sounds like a straight up continuity problem.

Ya know TT contacts can get oxidation on them and start to loose connectivity? There is a special bantam / tt burnisher tool that's used for cleaning contacts. But you can also just squirt a little Deoxit on a TT plug (cradled in a rag to keep it from spraying everywhere) and then run it in and out a bunch of times, wiping off the plug and then in and out and so on.

Otherwise, break out the meter and measure the resistance between the tip and ring of the TRS and the corresponding pins on the DB25 or whatever it is going into the converter. It should be basically at or near the shorting resistance of your meter probes.

 

[RBBlackstone]

Is there any chance things are physically configured differently?
This reminds of an issue I had with rack mounted piece of gear that was right over circuitry that picked up the magnetic field. Is there anything in close range to the balancing PCBs?

Your drawing also looks like the hot(red) is on the ring of the schematic.

 

[nohatnoswim]

Think I've solved it, or at least I've stopped the hum. And yes it's connected with what is different in the new setup - the patchbay normalisation.

My normalisation on the patchbay had jumpered the Hot & Cold from top to bottom where required (just like the plugs are designed to do),

However, the EDAC from the Balancing rack to the patchbay feeds [ hot -> hot ], and [ cold + ground -> ground ] on the bantam socket- the cold on the tt socket is unused - so I've had a normalising bridge connecting nothing to nothing on the colds. Time well spent, though at the time I considered it harmless and easier to remove than install later if needed.

Anyhooo, bypassing the normalising with a patch cable appeared to solve the problem, so I added a bridge to the grounds on each individual set with a jumper cable top to bottom, and voila - it's working without hum !

 

[mody1]

Just moved place and did a rewire.
Power is 3 phase, star system from 1 single socket.
Implemented significant normalization on patchbays for first time
All my balanced gear is whisper quiet.
Same gear & signal paths pre move were all solid.

Problem:
16 outputs of synths going into a Vintage Balancing Rack and then into SSL interface.
Outputs are all humming at various levels ( some of the synths aren't even powered on).
Bypassing normalisation with patch cables doesn't seem to change anything on the few channels it was tested on
Visual of typical path for the Juno 6 as an example in attached diagrams.

No specific dedicated grounding scheme in place - cables into patchbays, and out - grounding on patchbays is only within the plugs and the gear.

Is this
a) The normalisation,
b) The wiring path independent of the normalisation
c) Both
d) Other

Thanks,

NHNS

maybe the SSL balanced inputs are connected floating so the you end up without shield for the cables on that side

 

[nohatnoswim]

It's possible - I didn;t actually go that far down the chain and only focused on the Synth Output / Amp Input stage. Though the convertor inputs were all alive in the chain as the Amp Outputs are normalised into the convertor inputs, but focusing on the first stage seems to have done it.

For whatever reason (and I haven't fully thought it through as I just want to make some music after about 6 weeks down time) these particular connections need the ground connected on the patchbay. That might be a consequence of some other cabling along the path, but it worked so I'm not digging unless I really really need to go down that route.