[S Clark]

Dave,
I asked this over in the lab circle, but no one came back with an answer, and I thought you might know.  You see digital cables advertised as 75 ohms. Why? What makes that the magic number for resistance?

Scott

[wywires]

Has to do with Characteristic Impedance in a transmission line of finite length and is calculated as follows:

Z=the square root of (L/C)

L=Inductance per unit of length
C=Capacitance per unit of length
Z=Characteristic impedance in ohms

Not really a big deal at the lengths typical in audio systems. Can change radically based on frequency as L and C do.

[S Clark]

Thanks for the reply, Alex. I understand that it is an impedance, but why is 75 the magic #?  I assume that it is an attempt to match a signal going through a connector to reduce wave reflections.  I also assume that it has to do with the frequency of the digital signal. 

[Uptown Audio]

It is the spec of the transmitter and receiver to be 75 ohms. The cable has to also be 75 ohms to get an exact impedance match and avoid loss and interference. Coax is good about rejecting interference but it isn't impervious at the connectors. If the line has a poor interface, it will induce jitter and that can effect sound quality. Reflections in the coax cable cause the timing error and mismatched impedance causes reflections.
-Bill

[wywires]

Thanks for the reply, Alex. I understand that it is an impedance, but why is 75 the magic #?  I assume that it is an attempt to match a signal going through a connector to reduce wave reflections.  I also assume that it has to do with the frequency of the digital signal.

75 Ohm is a hypothetical standard for S/PDIF to try to match input impedance of the destination to the output impedance of the source. Again, this is important for very long lengths but a few ohms of CI here or there won't mean much in a home audio system. Again, L and C do vary with frequency.

[dBe]

Scott, what they said.

It has been my experoence that the termination (connector) is the real culprit in transmission of data.  The reflections within the cable from improper terminations wreak havoc with the interface.  With an SPDIF cable I prefer a 1.5m length, too.  All of this is empirical on my end.  Does anyone have a preferred length for optimum transmission?

Dave

[Uptown Audio]

Scott, what they said.

It has been my experoence that the termination (connector) is the real culprit in transmission of data.  The reflections within the cable from improper terminations wreak havoc with the interface.  With an SPDIF cable I prefer a 1.5m length, too.  All of this is empirical on my end.  Does anyone have a preferred length for optimum transmission?

Dave

Both the cable and the termination combine to create the overall 75 ohm interface. It is common however for some people to have a good 75 ohm cable and improperly terminate it and so change its impedance. I find 2m to be a good length. Anything over 3 feet should really be a requirement as lengths below that tend to cause DACs to suffer the most from the shorter reflections. This reflection issue is even true with optical cable, although the 75 ohm impedance is not relevant. You could probably use 100 feet of proper 75 ohm coax without data loss. If I recall correctly, Wadia used to supply a 16' or so length of ST glass fiber cable for connection of their audio systems.
-Bill 

[geezer]

I don't think 75 is magical. There had to be some standard defined to avoid confusion, and it works and is convenient. The standard was defined many years ago for other purposes.

[S Clark]

Interesting comments about optimum length. I'd have gone with the old "shorter is better" maxim. I've also read that RCA connections are not as good as BNC.  Have any of you found this to be true?

[HAL]

Scott,
I agree with Dave, the biggest issue is termination.

There are 50 Ohm and 75 Ohm BNC connectors.  The company has to choose the 75 Ohm version to use with RG59 or similar cable to give a 75Ohm system.

From what I have read, RCA's were developed for 50 Ohm applications.  Canare is one of the MFG's that makes a 75Ohm version of an RCA connector for use in S/PDIF applications.     

[S Clark]

It has been my experoence that the termination (connector) is the real culprit in transmission of data.  The reflections within the cable from improper terminations wreak havoc with the interface.
Dave

If it isn't revealing a trade secret, how do you like to terminate? If you took the digital output RCA out of your dac, and wired the cable direct, would you improve transmission significantly. Of course, your dac would have a tail at that point.

[S Clark]

Scott,
I agree with Dave, the biggest issue is termination.

There are 50 Ohm and 75 Ohm BNC connectors.  The company has to choose the 75 Ohm version to use with RG59 or similar cable to give a 75Ohm system.

From what I have read, RCA's were developed for 50 Ohm applications.  Canare is one of the MFG's that makes a 75Ohm version of an RCA connector for use in S/PDIF applications.   

Oh gee    Here we go again.  I've bought power cords from Electra and Wywires, IC's from Sonny, Electra, and KCI, and now we have Digital cables.  Blue Jeans sells a Canare 75 ohm with your choice of cables for the cost of a mexican dinner... and Wywires sells one endorsed by members of the House of Saud (it's a joke, Alex).  They are both 75 ohms.  Guess which one gets glowing reviews from people that I trust.  As per usual, I guess I'll start cheap, and end up spending more than the wife really knows.

[rollo]

  Well not only Wadia likes a 16 footer. Ray Kimber as well. If you can try the same 75 ohm cable in a one mtr. length and a 5 mtr. length let us know the result.
   The biggest improvement to our DAC was adding a pulse trannie to insure a 75ohm connection. The other big deal was using a PI Digital powercord. Go figure.


charles
SMA

[TomS]

Steve Nugent does a good job of explaining that, while cable length definitely matters, you really need to consider end to end connection - cable, connectors, wiring, circuit traces, transmitter and receiver chips http://www.positive-feedback.com/Issue14/spdif.htm. BNC connectors, for example, typically use a proper crimp, not solder. Fast rising pulses can do a lot of strange things since that means lots of higher frequencies are present above that fundamental.

[Æ]

Impedance and Resistance are not quite the same thing.
If you would like to know a little bit more about "characteristic impedance" read this very short article.
http://en.wikipedia.org/wiki/Characteristic_impedance