What difference does the quality of a digital interconnect make?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
No, RCA is not 75 ohm and cannot be. My guess is that it is actually lower than that, as the centre pin is quite thick.

I suppose with careful EM modelling it might be possible to arrange tails or other compensating components (of a higher impedance) so that the combination of those plus RCA is a reasonable approximation to 75R provided the frequency is not too high. I doubt if anyone has bothered to do this as for SPDIF it is not really worth it. It is one of those cases where those who know how to do it know it is not worth doing.

you know for a fact that RCA can't be constructed to be 75 ohm, and perfectly match the whole transmission line impedance?
 
No, RCA is not 75 ohm and cannot be. My guess is that it is actually lower than that, as the centre pin is quite thick.

I suppose with careful EM modelling it might be possible to arrange tails or other compensating components (of a higher impedance) so that the combination of those plus RCA is a reasonable approximation to 75R provided the frequency is not too high. I doubt if anyone has bothered to do this as for SPDIF it is not really worth it. It is one of those cases where those who know how to do it know it is not worth doing.

This could be 75 Ohm (I'll try measure it)

Canare Corp.: RCA Crimp Plugs: 75 ohm RCA Crimp Plug(RCAP-C Series)
 
To put some of the characteristic impedances for digital designs into perspective:
Standard digital boards, no special manufacturing requirements for impedance (this will include digital audio designs) +/-20%
For more critical digital designs with DDR interfaces etc, you would purchase impedance controlled boards, where you will get +/-10-5% depending on your requirements (+/-10%) being more common.
When you get into the silly frequencies (GHz) you move to Rogers and other materials.

So for the SPDIF interface we are looking at anywhere between 60-90 omhs, you aint going to get any better with off the shelf commercial Audio equipment (any price range).
 
I suspect there is some confusion about the fact that the bits transmitted across a digital interlink are not actually “heard”.
Here is a good explanation of the SPDIF interface caracteristics: http://www.epanorama.net/documents/audio/spdif.html
Bit errors:
This usually causes very significant changes in the sound, often loud popping noises but occasionally less offensive effects. Any data loss or errors in either are a sign of a very broken link which is probably intermittently dropping out altogether. “
 
Extreme_Boky said:
you know for a fact that RCA can't be constructed to be 75 ohm, and perfectly match the whole transmission line impedance?
Yes. So does everybody else who knows anything about electromagnetism and transmission lines. Perhaps this ought to be included in the training for an audio salesman? Along with the circuit theory, Fourier analysis, and some basic calculus and trigonometry.

stijn001 said:
I suspect there is some confusion about the fact that the bits transmitted across a digital interlink are not actually “heard”.
Perhaps. In the case of SPDIF their timing is heard, after low pass filtering in the receiver PLL, but fortunately the remaining low frequency jitter is not affected by cable performance (except, possibly, if it is a severely microphonic expensive 'audiophile' cable) but only by source oscillator quality.
 
I suspect there is some confusion about the fact that the bits transmitted across a digital interlink are not actually “heard”.
Here is a good explanation of the SPDIF interface caracteristics: http://www.epanorama.net/documents/audio/spdif.html
Bit errors:
This usually causes very significant changes in the sound, often loud popping noises but occasionally less offensive effects. Any data loss or errors in either are a sign of a very broken link which is probably intermittently dropping out altogether. “

nope, trust me ive been here, its just a more general, wide ranging type of confusion and a bit of denial thrown in.
 
To put some of the characteristic impedances for digital designs into perspective:
Standard digital boards, no special manufacturing requirements for impedance (this will include digital audio designs) +/-20%
For more critical digital designs with DDR interfaces etc, you would purchase impedance controlled boards, where you will get +/-10-5% depending on your requirements (+/-10%) being more common.
When you get into the silly frequencies (GHz) you move to Rogers and other materials.

So for the SPDIF interface we are looking at anywhere between 60-90 omhs, you aint going to get any better with off the shelf commercial Audio equipment (any price range).

Yupz, and that's usually good enough (at least in the TV industry with DVB/QAM/ATV RF signals <1GHz and "high speed" databusses like parallel MPEG) especially when you can tweak the drive strength of the driver in FW. Digital signals are usually easier to get working right because you can employ tricks like series termination that won't work for (DVB/QAM/..) RF signals

If you are really interested in the subject I can recommend a course by "fast edges" (google it, I attended it and learned a lot), it's not cheap but very informative.. One thing I learned was to think of the signal as a wave ;-)
 
Last edited:
Series termination works just fine at RF, but it costs you 6dB of power.

Sometimes, like say when matching into a crystal filter or mixer from a video opamp, dds or such that is perfectly acceptable however.

Both the filters and the mixers are usually very termination sensitive, and behave far better if driven from a wideband terminated source, a high speed opamp with series resistor gets you that at the cost of a little noise performance.

The variable drive strength (And copious LVDS ports) on modern fpgas are quite handy, especially when backed by wicked fast serdes blocks, makes dealing with things like SDI-HD (1.55 Gb/s) suddenly very much less painful.

Regards, Dan.
 
Now that I have my brand new Cambridge Audio 740C I connected the Sound Blaster digital output to the 740C DAC input using the same audio wire that used before for analog audio. No problem whatsoever. It's normal cheap braided audio wire of about 9m lenght (27ft).

The 740C takes GOOD care of the received signal and/or there is a huge hype about the cable needed to cope with the digital stream because the sound I hear is as clear as my ears are able to tell.
 
Yes. So does everybody else who knows anything about electromagnetism and transmission lines. Perhaps this ought to be included in the training for an audio salesman? Along with the circuit theory, Fourier analysis, and some basic calculus and trigonometry.

No, you are wrong because you believe everything you read. With the right choice of dielectric it is possible to have RCA at 75 ohms + with the right manufacturing process all reflections due to unnecessary soldering / crimping and similar, can be omitted completely.
 
Last edited:
even if its possible (highly doubtful, perhaos its you that believes everything you read… as long as it has bling), given you can get 75Ω from a $5 BNC, its nothing but wank value.

then of course you have the sockets

i'd be very interested to see the 75Ω set up with that quite large flat wire
 
Last edited:
well you asked for it lolly boy

they don't think, they know because they can hear the difference.

they don't think
ahem.. :rolleyes:


The people, who pay the money for the cables I help construct, are High Enders. They have money and do not wish to think, they listen and if the cable delivers, they do not ask how much -> they have to have it irrelevant of the cost.

You can only dream to be able to get the opportunity to listen to such cable, let alone afford it

ahem

Extreme_Boky
 
even if its possible (highly doubtful, perhaos its you that believes everything you read… as long as it has bling), given you can get 75Ω from a $5 BNC, its nothing but wank value.

then of course you have the sockets

i'd be very interested to see the 75Ω set up with that quite large flat wire

it would be interesting, wouldn't it... but then, just one look would cost > $20,000
 
haha, you say that like it has some value … its not an answer to a technical question and it is not back up for a technical claim. its an avoidance technique plain and simple

the whole cable is of 75 ohm impedance; that includes the RCA connectors at both ends. My connector looks quite different from Canare's, though.

Have you done any research yourself with regard to this subject, or are you going to keep to entrain me with your presumptions and humorous observations?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.