Master clock cables and SQ


Is there any difference in SQ between 50 ohm and 75 ohm cables when using a master clock that can be configured for either impedance? I am thinking specifically of the SOtM sCLK-OCX10 master clock. Good 50 ohm cables are very expensive and few are made. Good 75 ohm cables are made by every decent manufacturer and at a variety of price points. But is there a difference in SQ? Thanks!
Ag insider logo xs@2xkingbarbuda
Two different master clock - clock cable - streamer systems. (Think SOtM master clock and SOtM sms200ultra Neo). All equipment is the same. All cables are from the same manufacturer, the same termination and the same length. All impendances from the in and out connections and cables match.

The *only* difference is in the first system, all the impendances are 50 ohms. In the second system all the impendances are 75 ohms.

Question: which system sounds better? The first system (all 50 ohms)? The second system (all 75 ohms)? Is there no difference between the two? Why?
I can’t think of any reason why 50 ohms would be preferable to 75 ohms at the lengths that are involved in a home audio system, given that everything else is equal, as you indicated. So my instinct would be to go with 75 ohms, since as you indicated most audiophile-oriented unbalanced digital cables are 75 ohms.

Regarding Teo’s comments just above, his cables are of course very unique and I have no knowledge of how or if conventional transmission line theory may apply to them. But in the case of conventional (non-liquid) digital cables, regarding ...

50 ohm vs 75 ohm is a transmission line technology concept... improperly applied to these very, VERY short lengths of cable.

... a generally accepted rule of thumb is that a cable should be treated as a "transmission line" if the length is greater than 1/10 of the wavelength of the signal, and beyond this length reflections on the line resulting from impedance mismatches can become important enough to affect performance. And I should emphasize that this rule of thumb is not specifically related to audio, where extremely tiny effects such as timing jitter in the sub-nanosecond area may be audibly significant.

The OP’s master clock generator is unusual in that it generates the clock in the form of a 10 MHz sine wave, rather than as some approximation of a square wave. As a rough ballpark approximation let’s assume that a conventionally designed digital cable he might choose has a propagation velocity of 75% of the speed of light in a vacuum. That would be about 225 million meters per second. The wavelength of a 10 MHz sine wave in that cable would be about 225 million meters per second divided by 10 million cycles per second (i.e., 10 MHz), which is 22.5 meters. One tenth of that is 2.25 meters, or about 7 feet.

So in general, i.e., for most non-audio applications, maintaining a good impedance match at that signal frequency would be considered essential for cable lengths exceeding about 7 feet. In high quality digital audio applications I would expect that threshold to be considerably tighter, which is to say that I would expect impedance matching to be critical for considerably shorter lengths than that. And in the much more usual case of clock signals that are approximations of square waves, rather than the sine wave that is generated by the OP’s particular equipment, the signals involved would have significant frequency components associated with their risetimes and falltimes that are **much** higher than 10 MHz, and so in those more usual cases I would expect that threshold to be considerably tighter still, likely in the area of just 1 or 2 feet. And very possibly even less, depending on the risetimes and falltimes of the signal provided by the specific source. (For those who may not be familiar with this terminology "risetime" and "falltime" refer respectively to the amount of time a digital signal requires to change from its lower voltage state to its higher voltage state and vice versa. Or more specifically, those terms are generally defined as the amount of time required for the signal to change from the 10% point between its two voltage states to the 90% point, and vice versa, since what happens when the signal is close to those voltages is usually not important).

The bottom line: Keep impedances matched, if you are not using one of Teo’s cables!

Regards,
-- Al


50 ohm vs 75 ohm is a transmission line technology concept... improperly applied to these very, VERY short lengths of cable.

Which means they generally do not conform (in situ) to the theory that is applied to them.

in such scenarios as this, the transmission is one of a square wave, and thus the leading and lagging edges of the square wave IS the 'information' in the system.

And that... is going to be subjected to more or less distortion, mostly due to cable construction over that of characteristic impedance (in termination), in some important ways.

So a given 50 ohm cable might perform better as digital audio cable vs that of a 75 ohm cable, in a stated 75 ohm termination on both pieces of gear. might, being the operative word.

Eg, our liquid metal digital audio cables have no conformation to any transmission line standard, yet they perform in the top echelon of RCA terminated digital audio cables. Different technology, different result.
Right. But I will clarify and sharpen the question further:

Two different master clock - clock cable - streamer systems. (Think SOtM master clock and SOtM sms200ultra Neo). All equipment is the same. All cables are from the same manufacturer, the same termination and the same length. All impendances from the in and out connections and cables match. 

The *only* difference is in the first system, all the impendances are 50 ohms. In the second system all the impendances are 75 ohms. 

Question: which system sounds better? The first system (all 50 ohms)? The second system (all 75 ohms)? Is there no difference between the two? Why? 

Why am I asking this question? Because 50 ohm systems are a pain, because there is limited availability of good 50 ohm master clock cables at a reasonable cost. However, good 75 ohm digital cables are plentiful with many choices and price points. Yes. All impendances in the system connections and cable must match. The entire question for me is there any reason to have a 50 ohm system if you can have a 75 ohm system?

I hope that helps. Thanks everyone. 
What @almarg said -- the deciding factor is what the impedance is on the device you are connecting the clock too -- I've never seen a device with clock inputs that has both 75 and 50 -- it's either one or the other and that's what you need to drive.

By the way a poorly performing clock interface can be very odd to track down -- sometimes it just fails to work which is clear but other times problems can manifest as distortion on transients and you can spend hours seeking other causes of the problem in your system before realizing it's just that the clock has gotten mis-synched. That's all to say don't mix 75 and 50 or you can cause yourself a world of pain!
If there is a choice between all three impedances (clock generator output impedance, cable impedance, destination component input impedance) being 50 ohms or all three impedances being 75 ohms, I wouldn’t want to speculate as to which would be better. It might be dependent on the design of the specific components, or on the specific cable that is chosen, or it might not make any difference at all.

Good luck. Regards,
-- Al

Right. But assuming you have a choice of outputs on the clock and you get cables with matching impedance, is 50 ohms and better than 75 ohms or vice versa? Thanks. 
The output impedance of the clock generator and the impedance of the cable should be the same as the input impedance of the component the clock is being provided to. If the input impedance of the component the clock is provided to is 75 ohms but one or both of the other impedances is/are 50 ohms distortion of the clock waveform will result.

Regards,
-- Al