You are here

Ground loop over wordclock cable; what to do?

All about the tools and techniques involved in capturing sound, in the studio or on location.

Ground loop over wordclock cable; what to do?

Postby Uncle Freddie » Mon Aug 27, 2018 11:41 am

Hi,

I recently connected two devices via wordclock; suddenly I have a ground loop. The buzz isn't extreme, but it is there.

I think I read somewhere that the wordclock signal spec is robust enough to handle the noise, but being a perfectionist, I want the hum gone.

My studio is in my 60 year old house with its 1958-era electrical wiring. The clock master and the slave are in different rooms, which means they are plugged into different electrical outlets. I know that the house's electrical wiring is the root of the problem, however I want to save any electrical remodeling for a full-on house remodel. (Ideally I would install 20 amp, isolated ground outlets everywhere.)

Unfortunately the house remodel isn't happening anytime soon. So my questions are:

1. Is it possible to isolate a wordclock signal? I found some interesting BNC hum eliminators and isolation transformers for the video industry; are any of these suitable for wordclock?

https://www.bhphotovideo.com/c/search?c ... 028759660&

2. If an isolator is good enough for SDI video, is it good enough for wordclock?

3. Is inserting an isolation device even advisable, or would it cause an entirely new bucket of problems?


Thanks,
Fred
Uncle Freddie
Poster
Posts: 44
Joined: Sat Apr 23, 2016 12:00 am
Location: Detroit, USA

Re: Ground loop over wordclock cable; what to do?

Postby Hugh Robjohns » Mon Aug 27, 2018 3:15 pm

Uncle Freddie wrote:I recently connected two devices via wordclock; suddenly I have a ground loop. The buzz isn't extreme, but it is there.

This is possible, but unusual. Most digital inputs and output are coupled via pulse transformers which make them floating ground-free outputs/inputs. And ground loops tend to result in instability within a digital system, rather than audible hums or buzzes (although that remains a possibility with analogue I/O in the digital system, of course).

1. Is it possible to isolate a wordclock signal? I found some interesting BNC hum eliminators and isolation transformers for the video industry; are any of these suitable for wordclock?

If, after measuring the BNC ground connections with a multi-meter to make sure that the wordclock cable really was bridging between the grounds of the separate equipment, I think I'd experiment with a 1:1 pulse transformer, such as this one from OEP:

https://cpc.farnell.com/oep-oxford-elec ... dp/TF01476

A video isolator such as the one you listed should remove the galvanic connection between the source and destination, and a base-rate word-clock signal will probably pass through with insignificant damage.

2. If an isolator is good enough for SDI video, is it good enough for wordclock?

In practical terms, yes, certainly for base sample rates. I wouldn't be so sure at double or quad rates, if that's relevant to your installation, though.

3. Is inserting an isolation device even advisable, or would it cause an entirely new bucket of problems?

I think I might also investigate whether the problematic ground loop is actually on the analogue side of things -- the audio outputs to your speakers etc. A dual-channel audio isolation transformer box like the ART DTI is a lot cheaper than that video isolator!

H
User avatar
Hugh Robjohns
Moderator
Posts: 24376
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Ground loop over wordclock cable; what to do?

Postby Uncle Freddie » Tue Aug 28, 2018 7:21 am

Thanks Hugh.

My setup in more detail: There is an Apogee Ensemble Thunderbolt in the main room, connected via wordclock to an Apogee Element 46 in the second room.

In the second room, the Element 46 doesn't actually have a grounded plug, nor does the Mac it connects to; but the monitor for that Mac IS grounded. Unplugging the HDMI cable makes the buzz go away.

To be specific about where I actually HEAR the buzz: the Apogee Ensemble's Master output shows the buzz as faint meter activity; I ultimately hear the buzz out of my main monitors.

I think my interim solution will be to leave the HDMI cable disconnected from the computer in the second room.




1. I clock everything at 96kHz; I should officially avoid video isolators, then?


2. You mentioned pulse transformers; assuming both Apogees have pulse transformers behind their wordclock connectors, can pulse transformers blow out?

If pulse transformers can blow, what tests could I perform to see if mine are blown?

I am asking because I swear the buzz was not apparent for the entire day, until suddenly it was, like a light had been switched on. There is no way I could have gone all day without noticing the buzz.




Finally, for what it's worth: I took a voltmeter and measured the voltage across the ground lugs of the electrical outlets in both rooms. It came out to 0.06-ish volts AC.


Thanks,
Fred
Uncle Freddie
Poster
Posts: 44
Joined: Sat Apr 23, 2016 12:00 am
Location: Detroit, USA

Re: Ground loop over wordclock cable; what to do?

Postby ef37a » Tue Aug 28, 2018 8:59 am

During my time in the network "bits" industry I came across "baluns" that were used to mate 75R BNC (might have been 50R, be aware) to a 110 Ohm UTP 4 pair network cable. These would also give isolation since network pairs are differential.

I shall look up the Old Firm and see if I can find them.

Dave.
ef37a
Jedi Poster
Posts: 10608
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk

Re: Ground loop over wordclock cable; what to do?

Postby ef37a » Tue Aug 28, 2018 9:04 am

Found these..https://cpc.farnell.com/pro-signal/psg0 ... DAQAvD_BwE

Two of those and a short patch cable. Job's a good 'un?

Dave.
ef37a
Jedi Poster
Posts: 10608
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk


Who is online

Users browsing this forum: Tim Gillett