unshielded SPDIF digital cable?


Hi, yesterday I tried a home made unshielded SPDIF cable and I must say that it sounded surprisingly good. I compared this home made cable (consisting of two twisted solid core gold/silver wires with teflon insulation) with a trusty Nirvana AES/EBU. The results: the unshielded SPDIF cable sounded more "natural" and warm, emphasizing the midrange. The bass is a bit diffuse and the treble a bit subdued, but there is comparable amount of detail and PRaT. Do you have experiences with unshielded digital cables? What are the sonic characteristics of these cables in comparison with their shielded counterparts? Of course most experts would not recommend using unshielded digital cables and they won't recommend the type of cable I'm using either because the characteristic impedance of the cable, which is very likely wouldn't be the standard 75 ohms. But there are some experts (more enfant terribles) also who don't care about shielding or 75 ohms characteristic impedance, like Kimura San of 47 Labs and Kondo San of Audio Note Japan. Could it be that the sound I hear from the unshielded SPDIF is in fact the sound of jitter, which adds a bit of warmth to the sound? Your comments please.

Chris
dazzdax
Hi Chris,

Several questions: What are you connecting? Does the dac have a jitter reduction function, such as the Benchmark's do? How long are the cables? AES/EBU is normally 110 ohms, and uses either BNC or XLR connectors. Is the SPDIF cable connected between different connectors than the Nirvana cable, or the same connectors, and are they BNC, XLR, or RCA?

Twisted pair cable typically has an impedance in the area of 110 to 120 ohms, not very different than the AES/EBU standard. Lack of shielding, I suppose, could in some situations result in a degree of isolation (at high frquencies) between the chassis of the two components that would reduce coupling of digital noise from transport to dac.

Regards,
-- Al
Hi Al, I have a JMF Audio DXC2.2 D/A-converter with jitter reduction feature. My transport is MBL 1621. If the DIY has an impedance of 110 ohms, then it is a gross deparment from the recommended 75 ohms! That would be the same as screwing up digital data transmission. The other possibility is that I like the sound of jitter + phase anomalies!
P.S.: The connectors are WBT locking RCA's.

Chris
Wow! Nice equipment!

It's very conceivable to me that the impedance mismatch would not be significant, given that the transport appears to be redbook only (16/44.1, as opposed to 24/96 or 24/192), and assuming that your cable is approximately the 1.5 meter length that is considered optimal for transport-to-dac connections, and given the fact that the dac has a FIFO buffer to deal with jitter.

A FIFO (first-in first-out) buffer memory clocks the incoming data into a memory based on the clock that is recovered from the incoming signal, but clocks the data out of the memory with the dac's own internal clock. So jitter is totally eliminated unless the incoming waveform is so severely distorted, or jitter is so severe, that the phase-locked loop clock recovery circuit sees extraneous or missing clock edges as a result of the distortion.

Basically, the impedance mismatch results in reflection of a fraction of the incoming energy back to the source, where a fraction of the reflection will be re-reflected back to the load. Most of the re-reflection (the fraction that is not re-reflected once again) will sum together with the original incoming waveform, resulting in distortion. That distortion will only be significant if its timing is such that it arrives coincident with an edge, or if it is EXTREMELY severe such that it causes a high-level (logic 1) or a low-level (logic 0) to be seen as an edge.

A 1.5 meter length (or more) will delay the re-reflection past the original incident edge, and so it most likely won't become a problem until it is delayed to the next edge. That won't occur, in the case of redbook 16/44.1, until the cable length becomes greater than around 10 meters. See my post dated 3-23-09 in this thread:

http://forum.audiogon.com/cgi-bin/fr.pl?ddgtl&1237611683&openflup&13&4#13

I'm still not clear, though, on what you have been comparing. The transport and dac both appear to have AES/EBU and SPDIF interfaces, and xlr as well as rca connectors (as well as bnc's on the transport). Isn't your Nirvana cable a balanced AES/EBU cable, while your diy cable is unbalanced, with rca connectors, and if so doesn't your comparison encompass several variables other than the cable itself and its lack of shielding (including balanced vs. unbalanced signal paths within the two units, different impedances, different voltages, etc.)?

Regards,
-- Al
Hi Al, I compared the DIY with Nirvana AES/EBU and Atlas Opus SPDIF. Both the Nirvana and Atlas sound quite similar. The sound is clear and spacious, but rather "hifi-ish", with some emphasis on the highs. High pitched violin sounds a bit thin through these cables. The DIY brings a bit more warmth into the sound and while it is not as good in the bass department, it sounds more "musical/natural" to me, which is a remarkable thing. I've heard that some fellow audiophiles who own some $$$$ audio rig prefer the very simple 47-Labs OTA for their digital than the expensive boutique cables. Apparently the reason was not because they haven't the money to buy those expensive cables...

Chris