Jump to content
IGNORED

Minimum length for 75Ω coax cable (Revisited)


Recommended Posts

There was some discussion recently concerning the conroversy over the best minimum length for a 75Ω digital coax cable to connect a USB-S/PDIF converter to a DAC. Sorry, but I did search recent threads, and could not find the particular discussion before deciding to recreate this new one.

 

In a nutshell, the traditionalists say, quite simply, the shorter the cable the better, while some folks insist, through testing that the ideal length, to prevent internal bounce-back reflections from corrupting the signal is 1.5 meters. I read a few "white papers" and comments in other audiophile forums on the subject, but personally found the notion counter-intuitive and silly.

That is, until I ran my own tests.

 

I have two identical Canare LV-77S cables to connect my Musical Fidelity VLINK192 to my Peachtree Dac•iT, purchased here:

Canare LV-77S Pro Coaxial Digital Audio SPDIF Interconnect Cable 0.3m items in Cable Solutions store on eBay!

 

One is .5 meters in length, the other, identical cable is 1.5 meters long. I've been using the shorter one, until:

 

My dealer loaned me a Rogue Sphinx hybrid integrated amp to try out. This unit has a 2x 12AU7 tube-powered preamp, a class D power section, and includes a phono stage. With the source knob set to "phono" and the volume up full, I could hear my line-source music "bleeding through" the circuit, and faintly coming from the speakers (along with a lot of hiss from the not-so-great phono stage in this amp).

When I did the same thing using the longer cable, the "ghost" music disappeared! If there is "data reflection" in the short cable that is corrupting something down the line of the signal path, this might be evidence of it. I can't explain it , but I've now switched to using the 1.5 meter cable.

Link to comment

I never discount anything that Michael Ritter, founder of Pacific Microsonics and Berkeley Audio Design, has to say on the issue. PM was good enough to be acquired by Microsoft. They are engineer's engineers and their works over many years speaks for itself. They recommend 1.5m or longer. Good enough for me. I see no reason to second guess them on this. Not sure this is the effect you are experiencing, but still good to know.

 

Lead Dress

It is good practice to keep various types of signal and power cables physically separated from each other. This is especially important for the USB cable to the computer. It should be routed away from the chassis of the Alpha USB and any other component or cable that is part of the audio system. 1.5 meters is a good default length for USB, SPDIF and AES cables.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment

+1 on the USB cable routing.

Good improvements from trying a few different paths & getting it away from other cables.

Appreciation of audio is a completely subjective human experience. Measurements can provide a measure of insight, but are no substitute for human judgment. Why are we looking to reduce a subjective experience to objective criteria anyway? The subtleties of music and audio reproduction are for those who appreciate it. Differentiation by numbers is for those who do not." — Nelson Pass

Link to comment

Shortest cable is good for analog domain because you worry about interference/noise pick up. For digital, you need to worry more about reflections than anything else and a longer cable is almost mandatory - or at least helps mitigating the reflections (the matching of impedances is never perfect).

 

This is because the SPDIF receivers have 'gated' inputs. When they 'lock' on signal they ignore any other signal that arrives past the small window of time when the input 'gate' is open. The 'gate' is opened periodically to re-sample the incoming signal.

If the reflected signal comes back quick enough following the real signal (short cable), it will find the 'gate' open, it will over-impose over the original signal and affect the decoding process.

The average receiver has a 8ns 'window'. Light travels 8x0.3m=2.4m in vacuum in that time. Over a 75 ohm cable, that has slower speed (0.8 of light in vacuum), it can result in 1.9m.

reflection being a 2-way travel, cable can be 1/2 of that length - 1m would be minimum needed. Recommended 1.5m would be safe 'enough'.

I personally stay at 6ft/2m lenghts, just because I don't know exactly the parameters of my cables or my SPDIF receiver - that 8ns 'average' might be actually 16ns on my particular receiver chip.

 

LE: If you use a too long of a cable (like 5m), the second reflection might come in the 'gated' time to affect the receiver. But the level of second reflection is so low (each reflection attenuates the signal) that it might not be detectable.

Link to comment

The above 'SoNic67' post is just a bunch of random facts and assumptions, taken way out of context and then mis-applied to SPDIF.

 

At the end of the day, SPDIF is a very robust, very low frequency system. For the short length cables that most of us have, any cable in your house with the correct connectors will work just fine. If however you need cables about 10 meters (33 feet) or longer, then it should be a 75 Ohm co-ax.

Link to comment

.......................

My dealer loaned me a Rogue Sphinx hybrid integrated amp to try out. This unit has a 2x 12AU7 tube-powered preamp, a class D power section, and includes a phono stage. With the source knob set to "phono" and the volume up full, I could hear my line-source music "bleeding through" the circuit, and faintly coming from the speakers (along with a lot of hiss from the not-so-great phono stage in this amp).

When I did the same thing using the longer cable, the "ghost" music disappeared! If there is "data reflection" in the short cable that is corrupting something down the line of the signal path, this might be evidence of it.

....................

 

Well, you can't get analog cross-talk or bleed through into a digital input, so I would look for a different cause than cable length.

Link to comment

Actually, his post uses some nice back-of-the-envelope math to estimate the reflection delay.

 

If you don't buy into the digital reflection thing, then don't and run short cables if you want. That's your business. Personally, I'd stick with 1.5 - 2.0M

 

More info on this is presented by Steve Nugent of Empirical Audio here:

 

spdif

 

If the rise-time is 25 nanoseconds and the cable length is 3 feet, then the propagation time is about 6 nanoseconds. Once the transition has arrived at the receiver, the reflection propagates back to the driver (6 nanoseconds) and then the driver reflects this back to the receiver (6 nanoseconds) = 12 nanoseconds. So, as seen at the receiver, 12 nanoseconds after the 25 nanosecond transition started, we have a reflection superimposing on the transition. This is right about the time that the receiver will try to sample the transition, right around 0 volts DC. Not good. Now if the cable had been 1.5 meters, the reflection would have arrived 18 nanoseconds after the 25 nanosecond transition started at the receiver. This is much better because the receiver has likely already sampled the transition by this time.

 

The above 'SoNic67' post is just a bunch of random facts and assumptions, taken way out of context and then mis-applied to SPDIF.

 

But I do agree that what he is describing sounds like something completely different. Phono, maxed-out, tube, bleeding-through...

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment
Well, you can't get analog cross-talk or bleed through into a digital input, so I would look for a different cause than cable length.

 

I don't claim to know what caused this phenomenon, but I repeated the test several times, changing *only* the .5 meter coax vs. the 1.5 meter, and the effect was the same.

Link to comment
Well, you can't get analog cross-talk or bleed through into a digital input, so I would look for a different cause than cable length.

 

This has being discussed in other threads. There is a coaxial cable manufacturer that suggest (from oscilloscope test) an 8' (2.40 meters) coaxial cable length, this is due to "reflections". Of course "termination" and the construction of the cable is very important also, as he stated.

 

Roch

Link to comment

So I guess the question is do you want to trust Lavry or Empirical Audio and Berkeley Audio on this one.

 

In my opinion, one doesn't really have to make any compromises to go with a >1.5-2.0M cable run... Other than maybe having to buy that extra 0.5M over a 1.0M. If you don't think it is a possible concern then buy something shorter... It's not really even worth the typing. The cases have been made numerous times. It's just where do you want to fall on the issue.

 

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment
So, in all those years you didn't hear about signal reflections on transmission lines? Let me doubt that you are really EE.

Refresher: http://www.ddpp.com/DDPP4student/Supplementary_sections/Zo.pdf

 

Good transmission lines don't reflect! That could be a description of a good transmission line. From the sender end of a good transmission line, it's impossible to tell how lone the line is. It could be infinitely long or just 1 meter long.

Link to comment

From your linked paper:

 

However, suppose that we have a 5-foot-long pair of conductors, and we place a resistance equal to

Z0 across the far end. If we instantaneously place a voltage source across the pair

as shown in (b), the same current flows forever as in the infinite-length case.

Thus, when a line is terminated by its characteristic impedance, we needn’t

consider transmission-line effects any further.

Link to comment

How can you say there are absolutely no impedance mismatches from Tx chip to Rx chip though the cable-connector interface, to through the circuit board and through to the silicon? And to say this with enough confidence to neglect the small and simple accommodation of just using a 1.5m wire? Surely, not all SPDIF sources and receivers are built to a perfect ideal.

 

Good transmission lines don't reflect! That could be a description of a good transmission line. From the sender end of a good transmission line, it's impossible to tell how lone the line is. It could be infinitely long or just 1 meter long.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment
From your linked paper:

However, suppose that we have a 5-foot-long pair of conductors, and we place a resistance equal to

Z0 across the far end. If we instantaneously place a voltage source across the pair

as shown in (b), the same current flows forever as in the infinite-length case.

Thus, when a line is terminated by its characteristic impedance, we needn’t

consider transmission-line effects any further.

Why did you stop there? Next phrase was very interesting:

The situation is different for a transmission line of finite length that is not

terminated in its characteristic impedance.

Yes, that's a big IF. Real life situations are not ideal. Line is not infinite. Matching impedances is not perfect. Also, the model of real lines is a distributed parameter, not the usual concentrated parameter approximation.

Cable impedance needs to match what was considered at design stage in order to minimize any tolerance influences.

 

Zeel%20Fig7.gif

Link to comment
There is no RCA 75ohm. Only BNC.

True. But you can still use a 75 ohm cable to it. The induced difference is minimal (especially when board components are adapted properly), but yes, it is. That's why some of the higher end equipment have BNC connectors.

Link to comment
My "2 cents" opinion? I've used as S/PDIF cable one of this:

[ATTACH=CONFIG]7381[/ATTACH]

 

It works. Very well :)

 

 

That is a short cable.

 

2012 Mac Mini, i5 - 2.5 GHz, 16 GB RAM. SSD,  PM/PV software, Focusrite Clarett 4Pre 4 channel interface. Daysequerra M4.0X Broadcast monitor., My_Ref Evolution rev a , Klipsch La Scala II, Blue Sky Sub 12

Clarett used as ADC for vinyl rips.

Corning Optical Thunderbolt cable used to connect computer to 4Pre. Dac fed by iFi iPower and Noise Trapper isolation transformer. 

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...