TOPIC:

Something funny. I am using a 2 x 4 HD as a DAC and single ended to bal conversion.
Stereo signal into optical, outputs set as follows:
input 1 to output 1
input 1 to output 2, inverted 
(outputs 1 & 2 form the L output, balanced)
input 2 to output 3
input 2 to output 4, inverted 
(outputs 3 & 4 form the R output, balanced)

What I found is that there is no diff signal between outputs 3 and 4 in this setup. There is, if I un-invert output 4.
It seems that output 4 is already inverted even if not set to invert in DC.

Did I make a mental mistake, and if not, anybody else found this issue?

Edit: the measurement I made some time ago on building a bal output from input 1 on output 1 and 2 with output 2 inverted show a clear THD advantage - the 2nd harmonic spike on the se output is cancelled, as expected, by the bal operation.

Jan
 

This would be unexpected and obviously incorrect operation.
I will try on my 2x4HD unit.

That said, IIRC, you will see an absolute polarity flip on all outputs.

Dave.

Yes I would be interested to know what you find.
An absolute flip on all outputs wouldn't be a problem.
IF you find a stuck output I would think it would be software.
I know a thing or two about electronic design and I can't image a flip on one output by some wiring error unless it is a design error.

Jan

I don't see any issue with mine.  I gave it an input on the TOSLINK jack and routed Channel 1 to all four outputs.  They're all the same.

Dave.

Same here. All pretty much perfect, the AP sees -0.03deg from 0deg/180deg.
I am now going to look at the cable. A short in one of the cables would also explain it.

Edit: It's the cable ... It's a cable with a 3-pin male XLR on one side, and two RCAs on the other.
What I would have expected (and I asked specifically before ordering) is that the two hot RCAs would go to pin 2 and 3,
and the RCA screens (preferable only one to avoid ground loops) to pin 1.
But in these cables, the two RCA screens both go to pin 3 and 1, while both of the two hot RCAs go to pin 2!
It seemed to work as long as I didn't invert one of a pair, but of course the signal coming out of the XLR would be single ended.

And of course as soon as I inverted one of the pair, the result at pin 2 would be near zero.
I guess I am lucky that the 2 x4 HD survived, must have some series resistor on the output for just such a stupid occasion.
But, with another pair of homemade cables I had lying around it all works as expected.
miniDSP off the hook

Another mystery solved, thanks Dave for motivating me to dig in on a Sunday evening!

Jan

Good update!

Yes indeed, and if you remember, there was a long discussion on a DIYaudio.com thread regarding the appropriateness of grounding Pin 3 via an adaptor/cable on the output XLR of an audio device when doing a balanced to unbalanced conversion.
This is a tricky one because some components will not have any issue with this, but some will.  A capacitive load de-coupling resistor on active outputs will probably save the day should this happen.  But you can't count on it!

Some of the professional gear is actually designed to sense a Pin 1/3 short and automatically increase gain 6db on the Pin 2 side to compensate.

It's much more straightforward to adapt an unbalanced source to a balanced load than it is a balanced source to an unbalanced load.
Some vendor cables may not always be wired correctly to do what you want to do.

Cheers,

Dave.