| dkxdn |
Hi
I have a CD player - Teac VRDS 10 with a digital coax output - impedance = 75 ohm.
Then I want to connected it to a pro gear Behringer DCX2496 that has an input impedance = 110 ohm
Could that be a problem - I use a 75 ohm cable ?
Should it be terminated and how ? |
|
|
| UrSv |
| Seems you have consumer S/PDIF 75 Ohm and the Behringer input you refer to is AES/EBU 110 Ohm. They do not mix AFAIK. The unit also has S/PDIF in but optical. Maybe that's an option from your TEAC? |
|
|
| UrSv |
| ...but if I'm not completely lost you may be able to get help from Guido Tent here on the forum... |
|
|
| peranders |
The consumer signal is much weaker, that's that big problem and if you have short cables the impedance difference isn't a major problem (reflexions).
Why don't you test? If it works, it works, very "digital". |
|
|
| dkxdn |
| I have try it and it works but the DCX2496 has a level indicator that very offen goes up in the red so I thougth there could be something wrong - Termination ??? |
|
|
| Guido Tent |
| quote: | Originally posted by dkxdn
I have try it and it works but the DCX2496 has a level indicator that very offen goes up in the red so I thougth there could be something wrong - Termination ??? |
Hi,
Termination will attenuate the signal even further. You'll need to actually convert from SPDIF to AES. This may help:
http://www.epanorama.net/links/audiodigital.html#spdif
regards |
|
|
| Jocko Homo |
| quote: | | ......and if you have short cables the impedance difference isn't a major problem (reflexions). |
You don't want to get me started on this, P-A.
Will it work?
Yes.
Will it sound good?
Depends on your expectations.
Jocko |
|
|
| peranders |
| quote: | Originally posted by dkxdn
I have try it and it works but the DCX2496 has a level indicator that very offen goes up in the red so I thougth there could be something wrong - Termination ??? |
The normal signal level from a consumer signal is much lower than a professional.
1 Vpp vs several volts.
As Jocko says, if you are satisfied, don't do anything. It sounds like **** => get a converter (transformer). |
|
|
| Elso Kwak |
| quote: | Originally posted by peranders
The consumer signal is much weaker, that's that big problem and if you have short cables the impedance difference isn't a major problem (reflexions).
Why don't you test? If it works, it works, very "digital". |
Sorry Per-Anders,
But I get the giggles from your post. The cable termination issue is explained in Horowitz page 879 a.f. in my edition: Transmission Lines, Chapter 13.09, stating that a transmission line should be properly terminated at both ends.
Looking forward to your 3000th post. I really envy you.
This forum should have a box obligatory to have checked before posting "Have you searched Horowitz?" instead of "have you searched?" for technical issues.:clown:
Postscript: I am a bloody chemist but you are an EE. To my ears your post sounds like pouring water into concentrated sulfuric acid or trying to extinguish a petrol fire with water.:att'n: Warning don't try to repeat these silly experiments!!! :att'n: |
|
|
| peranders |
I say that digital signal are the same as for instance analog video signals. If you have reflexions which don't are near the logic threshold, the problems aren't huge. Some ringing or extra steps (due to reflexions) won't be a big problem.
I think we should ask dkxdn what he thinks about the sound coming out of his box. Does it sound horrible or does it sound nice? If he thinks it sounds not too bad I'll guess the problem isn't so big after all. But it gets more important the longer the cable is. In this case we have only missmatch in one end of the cable. |
|
|
| Jocko Homo |
Maybe he doesn't know what jitter sounds like.
Maybe his D/A box has so much that a little more won't hurt.
I doubt it, though.
Jocko |
|
|
| dkxdn |
| No I dont no how Jitter sounds - please explain |
|
|
| peranders |
| quote: | Originally posted by Jocko Homo
Maybe he doesn't know what jitter sounds like.
Maybe his D/A box has so much that a little more won't hurt.
I doubt it, though.
Jocko |
Are you saying that a mismatch introduces jitter? |
|
|
| mrfeedback |
| quote: | Originally posted by peranders
Are you saying that a mismatch introduces jitter? |
Hi peranders,
Yup, Jocko has been saying this for a looong time.
Eric. |
|
|
| mrfeedback |
| quote: | Originally posted by dkxdn
I have try it and it works but the DCX2496 has a level indicator that very often goes up in the red so I thougth there could be something wrong - Termination ??? |
Which indicator goes into red ?.
There are input gains that may be set too high.
Eric. |
|
|
| Fred Dieckmann |
"Are you saying that a mismatch introduces jitter?"
I believe he might be......... I believe that it does and that seems to be why so much attention is paid to this issue. In addition to jitter increased reflections introduce RFI and high frequency garbage that gets into the analog sections of the DAC and following analog stages of the system. Short digital cables sound bad. They provide a low propagation delay which puts the primary reflection even closer to the logic transition point. I think you might benefit from reading some from the www.sigcon.com website and the book:
High-Speed Digital Design: A Handbook of Black Magic by Howard Johnson
Do some searches on this forum because this subject has been thoroughly discussed.
What happened to the good natured PA that was always telling me to relax? |
|
|
| peranders |
| quote: | Originally posted by mrfeedback
Hi peranders,
Yup, Jocko has been saying this for a looong time. |
Oki, doki, but can anyone present any numbers? How much jitter do you get, just because of the mismatch? |
|
|
| peranders |
When I mean mismatch I mean so little that any reflexions won't be lower than the transistion points for logic. I'm not _that_ dumb. The assumption is that the main flank is triggering the logic.
Also duty cucle within reasonable values is not either any important, or? |
|
|
| Fred Dieckmann |
The answer is 42. What exactly was the question?
"I'm not _that_ dumb."
I'm sorry......... exactly how dumb are you? |
|
|
| peranders |
| quote: | Originally posted by dkxdn
I have try it and it works but the DCX2496 has a level indicator that very offen goes up in the red so I thougth there could be something wrong - Termination ??? |
I think we can make this conclusion:
75 ohms out + 110 ohms cable(or do you have a 75 ohms?) + XLR-contact + 110 ohms input => sounds quite alright but you get a red light. Have you checked in the manual when this light is suppose to be lit? I suspect a weak signal because consumer signal is much weaker than a professional.
If it sounds great, why bother further? You could test to have the right connections and see if it makes any difference.
75 ohms out <-> 75 ohms cable <-> transformer 75 unbal -110 bal <-> 110 ohms cable <-> 110 ohms input. |
|
|
| fmak |
[QUOTE]Originally posted by Fred Dieckmann
[B]"Are you saying that a mismatch introduces jitter?"
---------------------------------------------------------------------
It has been suggested to me that the best sounding connection for dCS upsampling 2 cable gear are two shorts runs of IEEE1394 cable. This seems to makes sense as the dCss are connected inside with 6 in or so of twited pair cables, terminating in 2 XLRs.
Anyone with ideas of charcteristic impedance of 4 wire and 6 wire IEEE 1394 cable and how to terminate in XLR?
The other other Fred |
|
|
| peranders |
| quote: | Originally posted by fmak
and how to terminate in XLR?
|
Isn't your input already terminated? You need just a cable, yes? |
|
|
| Elso Kwak |
| quote: | Originally posted by peranders
I think we can make this conclusion:
75 ohms out + 110 ohms cable(or do you have a 75 ohms?) + XLR-contact + 110 ohms input => sounds quite alright but you get a red light. Have you checked in the manual when this light is suppose to be lit? I suspect a weak signal because consumer signal is much weaker than a professional.
If it sounds great, why bother further? You could test to have the right connections and see if it makes any difference.
75 ohms out <-> 75 ohms cable <-> transformer 75 unbal -110 bal <-> 110 ohms cable <-> 110 ohms input. |
Hi Per-Anders,
I got a red light! I will not ride to death!
Congratulations, you won!!! You are the owner of the cosmic truth :rolleyes:
Up to 5000++posts! Hurry, hurry, no time to loose!
:bigeyes: :bawling: :clown: :D |
|
|
| fmak |
| quote: | Originally posted by peranders
Isn't your input already terminated? You need just a cable, yes? |
-----------------------------------
No the questions is impedance of IEEE1394 4 wire and 6 wire cable and how best to terminate to an XLR. |
|
|
| Lars Clausen |
Based on measurements i have found the surprising conclusion that changing a coax cable from 50 cm to 150 cm did not change jitter performance at all. (Non terminated).
Proper termination of the cable also did not have influence on the jitter at the output of the cable.
It was measured on a steady clock (with 2pS jitter level 3-sigma) , the result might be different if you were to use a bi-phase modulated SP/DIF signal though. |
|
|
| peranders |
Interesting results Lars.
It must be wrong cause I said so also without any measurements.
Anyway, I think we are talking about different things here. Fred's book is about (I'll guess) digital communication with high speed and in such cases it's important to have control over speed in cable, losses, terminations, etc. The question is when these parameters start to matter.
Lars it's also interesting that you got 2 ps from a standard oscillator and Guido gets 6 ps from a special developed one (unkowned what the special thing was).
fmak, it seems very hard to get facts about FW cables. It's not so hard to measure the impedance if you have at least 10 m cable and a squarewave generator with fast rise time. The generator doesn't need to have same impedance as the cable. You just change the termination in the other end. It's clearly visible on an oscilloscope. |
|
|
| Fred Dieckmann |
"the result might be different if you were to use a bi-phase modulated SP/DIF signal though."
Yes it might................:smash:
I would back terminate clock lines even for
runs of a few inches. but then again I know what signal integrity is and actually measure and listen to the results of such efforts. |
|
|
| Lars Clausen |
Peranders: As far as i know Guido's clcok is spec'ed at 2.5 pS RMS jitter (3-sigma :) ).
Also i didn't use a standard clock for this measurement, but two different low jitter clocks, one from Madrigal Audio (Very good!). |
|
|
| peranders |
Oki, doki, wrong of me. Maybe it was that he could hear 6 ps?
Anyway, Lars have you tested a regular 2 USD oscillator also? |
|
|
| Peter Daniel |
| quote: | Originally posted by Fred Dieckmann
I would back terminate clock lines even for
runs of a few inches |
I'm supplying clock signals with 4" and 6" runs. What's the best cable to use for that (shilded, unshielded, twisted pair?)
Also, should the ground (for clock signal) come from clock board (with ea signal), or is it OK just to provide signal alone and connect the ground from main PS? |
|
|
| Lars Clausen |
I tested one of those, and it actually measured quite good. ;)
This type: http://www.interquip.com/productInfo/pdf/7/146.pdf
Just a few dB from the limitation of my test equipment, but i think i can modify it and make a more accurate comparison. I will post here again if i can make some progress. |
|
|
| Jocko Homo |
We are talking about bi-phase modulated signals, and cable length does matter in recovered jitter. As does the the quality of the termination.
Don't argue with me about this one.......
Jocko |
|
|
| mrfeedback |
AFAIK the only red signal lights on the front panel of the DCX2496 are input and output level overload indicators.
Presumably the input gain controls are after the AES/EBU/SPDIF input stage.
The DCX2496 controller software is worth a look (565k download) http://www.behringer.com/DCX2496/DCX2496_v1-15.zip .
Eric. |
|
|
|
