Optical cable jitter measurements

[Maxhawk]

I'm not stating that one cable sounds better or worse than the other. This is simply an objective measurement for informational purposes.

Today at work I did a jitter measurement on three different optical TOSLINK cables: (1) Dayton Audio 3', (2) Sound Professionals SP-TOC-99HQ 1m, and (3) Wireworld Supernova 5+ 1m.

I used an Audio Precision SYS-2722 audio analyzer which was also my optical source. I did a jitter measurement from 0 to 100 KHz. You can ignore the big spikes since they appear in all the plots. Judging my their frequencies, they probably have something to do with the sampling frequencies in the source.

[ojg]

Great experiment, thanks!

Now that you have revealed to us that you have access to such equipment, I would of course like to see plots for the following:

1. 75ohm coax with BNC
2. 75ohm coax with RCA
3. 110ohm balanced XLR



On the other hand, I believe the jitter in TOSLINK is much more dependent on how well the Tx and Rx devices are implemented wrt power supply decoupling etc.

[Jocko Homo]

I think that a more thorough test would be to measure the jitter on the recovered clock, using TOSLINK and coax.

I've done some of those test wrt different coax types.

Jocko

[A 8]

Interesting,

It seems very low compared to whats been stated and sometimes measured before.

Here is another thread also discussing this issue.Toslink jitter

[Maxhawk]

Originally posted by A 8
Interesting,

It seems very low compared to whats been stated and sometimes measured before.

I read through many of the messages in the post you reference, and I think the differences are that (1) the oscilloscope measurements are not band-limited and also include the HF jitter, and (2) they're using transmitters and receivers of unknown quality. Since the Audio Precision is a very expensive piece of test gear, I assume that its optical transmitter and receiver are "near perfect".

The company I work for makes digital settop boxes, and our optical SPDIF transmitter must meet a jitter spec of 4ns, which is about twice the typical measurement to allow for margin. The product uses a standard off-the-shelf Toshiba transmitter which is an industry standard for CD and DVD players.

Perhaps when I return to work on Tuesday I'll repeat the measurements at a higher frequency, and also with a coax connection for comparison.

[A 8]

I believe toslink performance is very dependant on how the transmitter and receiver have been implemented and think a comparison to coax using the same gear would be very interesting to many of us.

Thx/Michael

[Maxhawk]

Took some more measurements with 75 ohm coax, generic RCA, and balanced XLR.

This is the same jitter test as before, but out to a bandwidth of 1.5 MHz instead of 100K.


This is a comparison between the lowest-jitter optical cable (Wireworld) and a Canare LV-61S 75 ohm BNC:


This is the same Canare BNC cable compared with a balanced XLR connection:


And this compares the Canare 75 ohm BNC to a generic 10' RCA cable with RCA-BNC adapters on each end. It's very surprising to see the jitter noise floor to be lower in the RCA, but the peaks are obviously much higher:



Draw your own conclusions, and let me know if there are any other measurements you want me to do.

[fmak]

Draw your own conclusions, and let me know if there are any other measurements you want me to do. [/B][/QUOTE]
------------------------------------------------------------------------------------

Excellent. Why have you decided on 1.5 MHz bandwith and not higher. Instrument limit?

[Coolin]

It might be stupid but will a neutral density filter (or any light filter) in the optical cable make any difference?
And how about series resistance in the others to dampen out any reflections?
What kind of insulation does the Canare LV-61S have? Is it all copper?
On their web site it only mentions a PVC outer layer.

[tiroth]

Nice work! Nice to quantify how bad the "professional" XLR is.