RF Attenuators = Jitter Reducers

Do you have a SPDIF transformer in your Digital Device

  • Yes

    Votes: 40 71.4%

  • No

    Votes: 16 28.6%
  • Total voters
    56
Status
Not open for further replies.
Jump to Latest
#462
You're welcome! No, not going to SYs. He's about 1800 kilometers from here. =)

Just as a preview, here are 2 FFTs of my built in RealTek card feeding the analog inputs of an M-Audio USB Fastrack Pro. One is the signal with built in jitter, the other without jitter.
You can see there is a lot of mess down low. But I left the markers in the same place so that you can see the difference. It looks much, much cleaner using the same card for in & out.
Will post more complete stuff later.
 

Attachments

  • realtek-maudio loop44K.png
    realtek-maudio loop44K.png
    8.3 KB · Views: 191
  • realtek-maudio jitter loop44k.png
    realtek-maudio jitter loop44k.png
    8.1 KB · Views: 181
#463
Not quiet the perfect equipment for analogue waveform analysis as the loopback test would suggest, probably? Would you have the same for loopback graphs? We have a high noise floor & spuriae reaching up to -60dB - hmmm. So this would be the base level. You don't happen to have the cards that SY is intending to use, do you?
 
#465
Here is the loop back signal with built in jitter in the same card. (M-Audio). Much cleaner.
Don't know how much of the noise above is the Realtek output and how much is the clock mismatch.

I'll look int the rms values. I do not have the same cards as SY.
 

Attachments

  • m-audio jitter loop.png
    m-audio jitter loop.png
    7 KB · Views: 177
#466
panomaniac said:
Here is the loop back signal with built in jitter in the same card. (M-Audio). Much cleaner.
Click to expand...
Indeed it is
Don't know how much of the noise above is the Realtek output and how much is the clock mismatch.
Click to expand...
Yes it could be from either but the important point is it's showing the full system noise & resolution i.e if this was your measurement system you would have some knowledge & appreciation of what can be revealed by it? Maybe Joseph K, T or indeed SY have something more to say about this.

I'll look int the rms values. I do not have the same cards as SY.
Click to expand...
OK, thanks
 
#467
spdif signals from Hiface, one with 6dB attenuator, one without attenuators. I don't see a difference in the rise time, and the ringing looks about the same as well. 1M of RG59 to 72R load. 10x probe.
 

Attachments

  • hiface no atten.jpg
    hiface no atten.jpg
    37.7 KB · Views: 146
  • hiface atten.jpg
    hiface atten.jpg
    39.7 KB · Views: 144
#470
Why not try it instead of assuming - it can't harm your DAC. I advise you to try this!

Eh, what did you state already - remind me. What scope/speed? etc
Why did you use a 72R termination - you know the SPDIF standard is 175ohm, don't you? You need to give some detail about this set-up for your credibility.

Here's the previous 20ns scope shot with attenuators
An externally hosted image should be here but it was not working when we last tested it.


Compare & contrast with SY's
 
Last edited:
#473
Jitter is flank time differences. Not rise times or overshoot. I would argue that the signal could look like **** as long as the criteria for flank detection is fulfilled in a stable manner between every pulse. But please correct me!

This is of course only interesting if one extracts the clock form the SPDIF link. If a system don't, the only bad that can happen from a poor link is actual data loss as such a regime would be absolutely indifferent to SPDIF jitter.

(Then someone argued that it has a DC component i.e. trailing ones but SPDIF is coded NRZ and bi-phase mark so it hasn't)

/
 
Status
Not open for further replies.
Top Bottom