More than one person has inquired as to how long we can make our U-Byte cables. So, here is the scoop, posted here so that all concerned can read.
It needs to be made perfectly clear that this thread is specifically about our U-Byte cables. Not SPDIF cables in general. If you want to argue about how long they should or should not be and why, blah, blah, blah, this is not the place. This is being posted in the Commercial Ads section for a reason. It only pertains to our products.
So, in a sense, it is self-serving BS. So, don't bother to send us any hate mail about that fact. We acknowledge it up front.
Ok, I first measured a "stock" U-Byte cable, i.e. 16' long. It had a rise time of 4 nSec, which comes out to around a 88 MHz BW, assuming a -3 dB point for a single-pole response. A longer version of 84' (yep, 5 times as long) gave a rise time of 8 nSec, and a 44 MHz BW. So, the BW is only cut in half by increasing the length by 5 times.
The additional attenuation was negligible.
Ok, so what does this mean to someone who might be interested in one, but does not have EE degree?
Well, let's compare these numbers to the stock output of a Slim Devices Duet. Using a 3' (1 metre) U-Byte cable, it has a rise time of 6.4 nSec and a BW of 55 MHz. So, when using a 84' U-Byte cable on a "better" output, you add less than 2 nSec of rise time.
At this point, I would hope that someone would wonder how much of the rise time comes from the cable and how much from the interface itself. That would help to see how much the cable slows things down.
The rise time of the unit that I used to make these measurements is roughly 3.6 nSec, when using a 3' cable. Measured directly, it is around 3.4 nSec. Going to 16' only slows it down to 4 nSec. Getting resolution on my scope at these rise times is not 100% accurate. But let's assume for the sake of argument that each foot of cable adds 0.05 nSec in rise time to a SPDIF signal.
At 84'.............we get 4.2 nSec........add to 3.4 nSec........we get 7.8 n Sec. Hey, close enough for my sake.
So, in answer to all of your questions, you now have some numbers to help you in your decision. OK, but how slow is too slow??
Good question. I would hope that you would get at least the 7th, and hopefully the 9th harmonic in order to get a decent looking square wave. So, that would mean roughly 25 MHz of BW. That comes out to a rise time of 14 nSec. Which means we would want to limit the cable to around 10 nSec contribution of that number. Or 200'
C'mon.......no one here is ever going to run a 200' cable, are they? No, probably not. So, in a self-serving manner, I will proclaim that a U-Byte cable of under 100' should be fine. Why you would want to use one that long escapes me, but strictly going by the numbers, it ought to work. There should be plenty of BW, minimal pulse dispersion and no attenuation of any concern.
I hope that answers most of your questions, or at least gives you something useful that you can hang your hat on.
Going back the lab and work on a not-so-secret product. Almost ready for beta testing. Which reminds me........I will need 4 guinea pigs. One confirmed, one picked out, although he doesn't know it yet. Leaves room for 2 brave souls.
Pat
