• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Cables and jitter - an interesting phenomenon

ppataki

Major Contributor
Joined
Aug 7, 2019
Messages
1,624
Likes
1,893
Location
Budapest
I thought to share my experience with the community; I hope to get some explanation to it (since I cannot figure out the root cause)

The system in my living room currently looks like this:

Source (PC) connected via USB to a miniDSP UDIO-8 multichannel DDC which is connected via its factory provided DB25 SPDIF (digital coax) breakout cable to three Sabaj A30A FDAs
For the SPDIF inputs there is a jitter correction control on the Sabaj amps, from 1 to 8 - value 8 providing max correction
I can set the optimal value by starting with value 8 and then start dialing it down until audio dropouts appear let's say at value 2; so then I shall set it to value 3 (just as an example)

What I found out is the following: the optimal amount changes depending on what USB cable I use between my PC and the UDIO-8....
To me this seems really weird but that is the case

I have a Supra USB (A-B) cable, with that I have to set the correction value to 3 to avoid any dropouts. I tried a Cable Matters Thunderbolt 5 (C) cable and with that I can dial the correction down to 1 and have zero dropouts

I would love to understand what is happening here, just out of curiosity....
Thank you
 
How do you determine what the optimal "correction" value is? BER measurement at the SPDIF receiver? Eye diagram? If you get drop-outs, then signal transmission is obviously buggered, but that doesn't mean one notch next to that setting is optimal.
 
If you're using these 2 cables I'm assuming you're using different USB ports (USB-A and USB-C) on your computer, in which case it might not be the cables causing jitter but the ports (and since they're a different USB standard I'm assuming this is possible).
 
I don't have a better way of determining it
The point that rather interests me is how is it possible that the 'optimal' value is different when using different cables
 
If you're using these 2 cables I'm assuming you're using different USB ports (USB-A and USB-C) on your computer, in which case it might not be the cables causing jitter but the ports (and since they're a different USB standard I'm assuming this is possible).
Yes, that is a very good point indeed
I guess we might have found the answer
 
This smells of common-mode issue to me, basically a ground loop - coax S/P-DIF often is not properly galvanically isolated, and it is single-ended so signal integrity can be affected. I wonder what would happen if you were to use a USB isolator...

Now that being said, if you can leave jitter correction (which I assume is ESS DAC DPLL control) at 8 and this works fine, you might as well be leaving it like that... the only reason you can dial that down at all is avoiding the periodic skipping that can occur with some types of sources like LG and Samsung TVs due to DPLL bandwidth being too small to accurately follow their output (so the DPLL periodically unlocks and resyncs). It's an issue that became quite prominent after a bunch of DAC vendors had to switch from AKM to more rustic, wider-bandwidth Cirrus S/P-DIF receivers post AKM factory fire.
 
This smells of common-mode issue to me, basically a ground loop - coax S/P-DIF often is not properly galvanically isolated, and it is single-ended so signal integrity can be affected. I wonder what would happen if you were to use a USB isolator...

Now that being said, if you can leave jitter correction (which I assume is ESS DAC DPLL control) at 8 and this works fine, you might as well be leaving it like that... the only reason you can dial that down at all is avoiding the periodic skipping that can occur with some types of sources like LG and Samsung TVs due to DPLL bandwidth being too small to accurately follow their output (so the DPLL periodically unlocks and resyncs). It's an issue that became quite prominent after a bunch of DAC vendors had to switch from AKM to more rustic, wider-bandwidth Cirrus S/P-DIF receivers post AKM factory fire.

The UDIO-8 shall be isolated:

1725737510271.png


Nevertheless I have lifted the ground of my PC to see if I had some ground loop issues (I know it is a no-no but for testing purposes it is fine) and there was no difference so I guess I have no ground loop
I am positive that the phenomenon is due to the USB port being different as pointed out by PyramidElectric above
 
Back
Top Bottom