SPDIF ground loop issue

Thread Starter

pgo1

Joined Nov 7, 2012
67
I'm running two pieces of audio equipment connected by SPDIF and I'm getting a ground loop. I could buy an isolation transformer but it's an extra expense and I don't want the hassle. I read elsewhere that a DC blocking cap can solve the ground loop problem, eliminating the DC path, but the sleeve on the coax is connected to ground

can I cut the sleeve on the cable, or disconnect the connector from ground?

otherwise, is there anything I can try?
 

crutschow

Joined Mar 14, 2008
34,285
Do either of the units have a TOSLINK optical input or output.

If you truly have a ground loop, then you could use a converter such as this, that goes from coax SPDIF to TOSLINK optical and then back to coax SPDIF at the other end.
That will block any loop.
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
@pgo1 What is your definition of a "ground loop"?
It is customarily recognized as a difference in ground potential between two or more points that connected to an earth ground reference.
The common method to eliminate is to bond these points to a common earth ground point.
 

Thread Starter

pgo1

Joined Nov 7, 2012
67
What are the pieces of equipment and how do you know you have a ground loop?
a) it's an audio usb interface and a SPDIF audio interface b) when i plug the SPDIF cable from the master to the slave, I get ground hum in my speakers

Do either of the units have a TOSLINK optical input or output.
although the (master) audio interface *has* optical link, I am plugged in to SPDIF coaxial connection as the slave only accepts coax

SPDIF is transformer coupled. How are you managing to get a ground loop?
I checked the audio interface (master) and it does seem to be transformer coupled (there is a small transformer adjacent to both the SPDIF input and output). I'm not all too familiar with SPDIF but from some very hasty measurement, the master out sleeve is still connected to ground (measures less than 1 ohm to a nearby ground point) whereas the master in seems completely isolated. The slave in / out are not transformer coupled and the sleeves of both coax connectors are connected to ground. I only get a ground loop when I plug a cable from master out to slave in, and *not* from slave out to master in. Like I said, I don't know a lot about SPDIF but my money is on there being some engineering error on the slave (being made by a more 'boutique' manufacturer) than the master

@pgo1 What is your definition of a "ground loop"?
It is customarily recognized as a difference in ground potential between two or more points that connected to an earth ground reference.
The common method to eliminate is to bond these points to a common earth ground point.
my definition for this situation is I get 50Hz hum when I plug the cable in. Connecting both to a common ground is not impossible but a little impractical so I would like to find out if I can do this another way - if i need to graft an isolation transformer in somewhere I could do that, but I would like to understand exactly what's going on first
 

Thread Starter

pgo1

Joined Nov 7, 2012
67
The XLR version is always isolated and uses 110Ω twisted pair cable. I think that the 75Ω coax might not be.
Get an isolating transformer
https://uk.rs-online.com/web/p/audio-transformers/1068577
and make up a lead.
A bit of heatshrink around the transformer, and it might look like the snake that's swallowed an orange, but it will break the ground loop
love this solution, I can even go slightly better and graft the transformer *inside* the receiving equipment. However just one question, do you know what kind of configuration the output SPDIF might be using, that it's transformer is connected to ground? Seems to defeat the point and I'm wondering if it might be some defect in the interface.. from what little I can see of the copper pour around the transformer the ground is not connected to either of the transformer pads which makes me a little suspicious that it might be a connection somewhere else, or maybe I'm reading too much into it
 

Ian0

Joined Aug 7, 2020
9,677
love this solution, I can even go slightly better and graft the transformer *inside* the receiving equipment. However just one question, do you know what kind of configuration the output SPDIF might be using, that it's transformer is connected to ground? Seems to defeat the point and I'm wondering if it might be some defect in the interface.. from what little I can see of the copper pour around the transformer the ground is not connected to either of the transformer pads which makes me a little suspicious that it might be a connection somewhere else, or maybe I'm reading too much into it
I wonder if it was meant to be isolated, but the metal parts of the socket managed to connect it to the case.
 

Thread Starter

pgo1

Joined Nov 7, 2012
67
I wonder if it was meant to be isolated, but the metal parts of the socket managed to connect it to the case.
doesn't look like it, the case and socket are high quality construction, but I will investigate further when I have the chance to take it apart again
 

Thread Starter

pgo1

Joined Nov 7, 2012
67
ah, got a message back from the support -
The SPDIF output is hard grounded because current regulations require this. SPDIF coax is an unbalanced cable, and the shield is not allowed to radiate high frequency signals.

If you are into soldering then put a small capacitor between ground and shield on one side of the cable, thus breaking the ground connection for lower frequencies, and plug that side into the receiving unit.
so ill solder that cap
 
Top