MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Do I have an earth loop problem or something more specific?
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Author Do I have an earth loop problem or something more specific?
StutzJr
Hi folks, I've done some cursory reading on this page, so I understand that this topic is probably well covered already. I will be reading further, but still want to ask if my specific issue is different.
I'll try to get straight to it-
I've had a (euro) semi modular "Pittsburgh Lifeforms SV-1 Blackbox" for some time, which I now want to rack. I've removed the SV-1 from its box and mounted it inside a "new" 19" metal case, below a mostly filled 84HP rail powered by a "Tiptop uZeus" power supply with flying ribbon cables. As all the ribbon connectors on the uZeus were already taken, I opted to keep the existing "blackbox" DC power board attached to the back of the Pittsburgh SV-1 so that it could use it's own wall adapter and not need to draw power from the uZeus.

During a test patching session everything seemed to be working fine on all modules, but when I brushed my finger on the edge of the SV-1 panel I noticed a tingling sensation and stopped straight away and cut the power.
(Some pictures of this setup are attached below.)

I am not yet convinced this is an "earthing" issue, as both the uZeus and the SV-1 are supplied with double-insulated switchmode supplies, which both seem to be correctly isolated from the mains side according to my quick continuity tests. I am not sure if the SV-1's adapter board was designed to be retained when rack-mounting, but I still don't see why it should not work since every euro patch cable creates a common ground anyway?
Have I missed something here? What steps can I take to identify or narrow in on the real issue, intermittent or otherwise?


Pelsea
I see plastic washers on the rack mounting bolts. That plus painted rails could mean there is no common connection between the 0 volt reference of your two PSUs, at least when there is no patch up. The ideal fix would be to connect the PSU common or 0 volt lines together with a substantial wire. However, removing the washers may give you a quick fix.
Rex Coil 7
Pelsea wrote:
I see plastic washers on the rack mounting bolts. That plus painted rails could mean there is no common connection between the 0 volt reference of your two PSUs, at least when there is no patch up. The ideal fix would be to connect the PSU common or 0 volt lines together with a substantial wire. However, removing the washers may give you a quick fix.
Write the bloody PDF Pete. d'oh!





... write it ...


thumbs up
Pelsea
Working on it
Rex Coil 7
Pelsea wrote:
Working on it
You'll be a hero around here for your efforts. I'd do it myself, but I'm afraid I am considered less credible due to being a High School diploma holder. Besides, you're better spoken, and most certainly far more experienced with teaching folks than I am. My teaching experience is limited to tactical training of folks tasked with security of nuclear weapons systems.

lol lol Sniper

I didn't mean to come off as terse or brash with my comment (previous post above) ... it is more of a statement addressing how needed your experience and wisdom are. My frustration stems from a place of feeling centered around the dire need to provide education on the subject of grounding vs zero volt notions and proper configuration of power distribution systems. At this point the subject is so corrupted by misinformation combined with an increasing proliferation of more (and more) complex modular systems and modules that the need for someone such as yourself to "correct the record" (so to speak) has become urgent. Perhaps beyond urgent.

Big giant huge respect sent your way, good sir.

Brian of Earth.
StutzJr
Pelsea wrote:
I see plastic washers on the rack mounting bolts. That plus painted rails could mean there is no common connection between the 0 volt reference of your two PSUs, at least when there is no patch up.


That is a great pickup and an excellent point. I didn't have any metal washers of the right size when I put this together last week and totally didn't consider all the consequences. I will amend this and do some more thorough testing with my dmm and will report back tonight my time.

If that resolves the issue, which I suspect it might - I'll be happy for you to crop a copy of my photo from above to use as example of what not to do in your pdf guide!
Graham Hinton
StutzJr wrote:
During a test patching session everything seemed to be working fine on all modules, but when I brushed my finger on the edge of the SV-1 panel I noticed a tingling sensation and stopped straight away and cut the power.


You do not have an earthing problem, you have a no earth problem.

What you are experiencing is leakage current from your PSUs coming through parasitic (stray) capacitance through you to ground. It feels like a tingle if you brush your fingers on the panel lightly, but stops if you press firmly, right? This is quite common on low end PSUs.

It is not lethally dangerous, you are more likely to do some other damage recoiling in surprise, but it is indicative of a potentially serious problem: you have no protection against your system becoming live due to a mains fault in it or something it is connected to.

Connecting together various pieces of metal that are not connected to anything else won't help, you may as well sacrifice a chicken.
Llouwelyn
[quote="Graham Hinton"]
StutzJr wrote:
you may as well sacrifice a chicken.


lol
MarcelP
[quote="Llouwelyn"]
Graham Hinton wrote:
StutzJr wrote:
you may as well sacrifice a chicken.


lol


It’s blood must be liberally spattered on the earth. Or is that ground? hmmm.....
StutzJr
Graham Hinton wrote:
You do not have an earthing problem, you have a no earth problem.

What you are experiencing is leakage current from your PSUs coming through parasitic (stray) capacitance through you to ground.


And this is the correct answer! Thanks Graham, looks like chicken is back on the menu! - though I am currently considering options to wire an earth lead the case until I invest in a better power supply.

I have now replaced all the fibreglass washers with zinc plated ones. The issue I had earlier was a side distraction as after I acquired this case I discovered that imperial 10-32 or 3/16 UNF screws are hard to come by locally so I had to scavenge the workshop and forgot to get some suitable washers on the way home.

Anyway I now have some evidence of the issue Graham has described. Sorry for the shaky photos, if Pelsea wants some better pictures I'd be happy to retake them. In each case the scope is measuring potential between my hand and a socket ring on the module. Note the ~50Hz frequency of the "floating earth"...


Power off:


SV-1 powered only:


Tiptop uZeus powered only:
Graham Hinton
StutzJr wrote:
I am currently considering options to wire an earth lead the case until I invest in a better power supply.


The part that needs earthing is the 0V of each system and that is difficult as neither has a connection point for doing that nor connecting 0Vs together. Such PSUs are alright to get you started, but once you expand beyond it it is time to rethink. It's like keeping a kid's bike with stabiliser wheels when you need a grown up's bike.

Quote:

I have now replaced all the fibreglass washers with zinc plated ones.


The only function of those washers is to protect the angle being scored by the screw head. If chassis bonding is required you need terminals making contact with the metal beneath, not hope that you might get through paint or anodising. Again you also need an earth.

Again following rules blindly is still voodoo. If you want to know whether something is properly earthed (once you get one) a Personal Appliance Tester should be used.

Quote:

In each case the scope is measuring potential between my hand and a socket ring on the module. Note the ~50Hz frequency of the "floating earth"...


It is normal to get mains frequency waveforms by touching a probe tip due to the high impedance of the scope inputs and the fact that is difficult not to be standing in a mains magnetic field. It is a standard way of identifying a probe.
StutzJr
Graham Hinton wrote:
The part that needs earthing is the 0V of each system and that is difficult as neither has a connection point for doing that nor connecting 0Vs together.


Hence my original question about using the "blackbox" power adapter on the back of the SV-1 in my original post. *My assumption* is that the design of these systems relies on the rail/frame as a signal ground in addition to the ribbon connector. So my approach is to verify whether this is valid. When I still had the incorrect fibreglass washers installed, I did notice poor continuity between socket rings on the SV-1 and modules in the top row with my multimeter. After I replaced the washers with metal ones I now have good continuity between socket rings on the SV-1 and modules in the top row - unpatched and with no power connected - therefore only common through the rails and frame. Sure I've only tested this with a shoddy handheld dmm but I actually do have access to an appliance tester and I will be doing a proper safety test before operating the system again.

Quote:

The only function of those washers is to protect the angle being scored by the screw head. If chassis bonding is required you need terminals making contact with the metal beneath, not hope that you might get through paint or anodising. Again you also need an earth.


Yeah, I'm not too worried about scratching those rack ears, I only used the washers to make sure i'm getting strong even pressure on both sides of the gaps on those tiptop ears when I tighten the screws rather than splaying those gaps open and digging the narrow screw head into that gap and getting stuck.
Initial observations seem to suggest continuity to the frame is already significantly improved.

Quote:

It is normal to get mains frequency waveforms by touching a probe tip due to the high impedance of the scope inputs and the fact that is difficult not to be standing in a mains magnetic field. It is a standard way of identifying a probe.


I don't disagree with that, but I will point out that I did not change any settings on the scope between taking those 3 images. Note the lack of any similar magnitude of AC pickup in the first image with the power off.


I won't be able to get back to this unit until later in the week, but will report back if I'm any closer to resolving the issue then.
Pelsea
I am interested in where that parasitic voltage is coming from. My bet is one of the transformer boxes is close to or sitting on that metal rack.

I know nothing about Australian power practices, but I would provide a safety ground for that rack with a metal enclosed power strip (with 3 wire plug, of course) screwed to the rack. Mount it on the back or side so the power warts are well separated from the electronics. Scrape off enough paint to ensure a good connection.
djs
Pelsea wrote:
but I would provide a safety ground for that rack with a metal enclosed power strip (with 3 wire plug, of course)


This is the real answer here.


You have two problems:


1. Your 0v reference is not equal between the two sections of the system (Tip top uzeus and Pittsburgh). This can be fixed in several ways.


2. Your 0v reference (and the case itself) is not connected to the ground (earth) from your wall outlet. This is what Graham Hinton is getting at. Without this connection, you will have sporadic mysterious problems as your modular grows. Also, it's a safety issue if there's ever a mains electrical issue that can come in contact with you (via front panel, patch cord, etc).
Pelsea
This is my go to power strip

https://www.tripplite.com/4-outlet-power-strip-6-ft-cord-5-15p-12-in~P S120406

They come in various lengths, and the top part of the extruded case is connected to the third wire ground (the bottom half of the case is not). It mounts with metal straps.The best ground connection is via the case screw at the end.

If you want to go first class, check out a rack mount PDU-- it will ground the rack rails.

https://www.amazon.com/Tripp-Lite-Network-Grade-Wide-Spaced-RS1215-RA/ dp/B00006B834/ref=asc_df_B00006B834/?tag=hyprod-20&linkCode=df0&hvadid =194024095585&hvpos=1o2&hvnetw=g&hvrand=14885859704713744138&hvpone=&h vptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9032137&hvtargid=pl a-309420854336&psc=1


And then there's this:
https://www.amazon.com/Earthing-Grounding-Improves-Inflammation-exchan ge/dp/B07B2CVMCR
Pelsea
So I wrote the pdf. I put it in its own thread so folks can comment on it and I can make revisons before it gets stuck.
ranix
StutzJr wrote:

Hence my original question about using the "blackbox" power adapter on the back of the SV-1 in my original post. *My assumption* is that the design of these systems relies on the rail/frame as a signal ground


Danger, Will Robinson! You said "signal ground" but I believe you meant "return path".

When Graham says there's no ground, he probably means there's no ground at all. As in, there is no ground pin on the adapter you plug into the wall, and therefore no ground in this cabinet. I bet your power supply is a wall wart with a 2 conductor DC barrel connector on the end.

Since there's no ground connection, there's no path for stray current caused by RF interference in your circuits to go to and the whole cabinet will have a fluctuating reference voltage (probably fluctuating at 50-60hz). It's hard to measure this.

Also, you are vulnerable to a failure condition called a "ground fault", during which the chassis can go live. Normally, the chassis is coupled to the ground pin that plugs into your wall and stray current that hits the chassis would be discharged through the ground pin. Since you don't have a ground pin, stray current hitting the chassis will be discharged through you (or, if a module is patched to another cabinet WITH a ground connection, likely through the other cabinet and causing damage).

If both a cabinet with a ground return and chassis ground and a cabinet with no ground return and no chassis ground are connected in the same physical enclosure (like a 19" rack) sometimes it will work to couple the two chassis together with a ground cable. Also, as mentioned above, sometimes this connection will be made through the rack screws in lieu of the ground cable, but often you need to grind paint away for this to be a low impedance connection. It's not reliable so if you do that you should use a real cable to connect the two grounds.

In my 19" rack I do have a TipTop uZeus as well as two 5U cabinets which all have no ground connection and are powered via wall wart adapters. In this rack I also have a 19" power strip and some other powered equipment with ground cables and chassis grounding. All these devices do establish a common ground through the rack screws and I don't have noise issues. This is not very safe and I am likely to one day suffer some equipment damage from this, but the DC I'm dealing with is only 12V so I'm not worried much about injury. I power down all my equipment before I sleep or leave the house by flipping off the power strips and I keep a fire extinguisher in the music room.
Rex Coil 7
Pelsea wrote:
So I wrote the pdf. I put it in its own thread so folks can comment on it and I can make revisons before it gets stuck.
applause Trampoline spinning Dono-Kun Dance w00t!! thumbs up
StutzJr
Many thanks for your efforts Pelsea, I will be spending some time studying your paper this afternoon before proceeding any further with my issue.

In the meantime, thanks also for everyone reminding me to use the correct terminology, I do appreciate it.

ranix wrote:
Danger, Will Robinson! You said "signal ground" but I believe you meant "return path".


My apologies, that would be a much more correct term to use.

Quote:

...there is no ground pin on the adapter you plug into the wall, and therefore no ground in this cabinet. I bet your power supply is a wall wart with a 2 conductor DC barrel connector on the end.


Yes, in my original post I did state that the wall adapters were double insulated switch mode supplies. I was unaware that alone did not already imply that they were Class II "" 2-pin non-grounded adapters. The main point of this thread was/is to determine if this is a problem, and if so, how to deal with it. I will be certainly spending some time studying Pelsea's paper before proceeding.
ranix
StutzJr wrote:
I was unaware that alone did not already imply that they were Class II "⧈" 2-pin non-grounded adapters.


It probably does imply that, but I'm not familiar with that terminology and I was being very explicit because I don't want to imply anything incorrect when discussing power supplies. Also, if I do make an incorrect statement, if I'm very specific there's a good chance someone will correct me. That's important to me because understanding power supply connections and grounding is necessary for personal and equipment safety.
Rex Coil 7
ranix wrote:
StutzJr wrote:
I was unaware that alone did not already imply that they were Class II "⧈" 2-pin non-grounded adapters.


It probably does imply that, but I'm not familiar with that terminology and I was being very explicit because I don't want to imply anything incorrect when discussing power supplies.
Yuh, cuz if you do, this gets thrown at ya! .....



ranix wrote:
... Also, if I do make an incorrect statement, if I'm very specific there's a good chance someone will correct me. That's important to me because understanding power supply connections and grounding is necessary for personal and equipment safety.
I hear ya ... same here ...

... because .....





Yeeeeupp!
cookie?!?
StutzJr
So for the short term I have investigated and verified a suitable means of grounding the chassis of my case.
In my observations, the AC waveform that I had depicted above has been all but eliminated from the scope, there is no noticeable noise, and most importantly there were no issues identified in the setup when checked with an electrical appliance tester. For now that achieves an outcome that is good enough for me. It is often said that "a temporary fix that works often becomes a permanent solution", but I have not forgotten Rex Coil's advice to me earlier in the year.

I do not wish to detract further from the good advice I have been given here, and will consider suitable options to more adequately power this system. I was fortunate to inherit this case, and it would be a shame to underutilise its potential.
Dcramer
Good to hear, but don’t leave us dangling, which end of the chicken did you have to probe? Chicken
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Page 1 of 1
Powered by phpBB © phpBB Group