|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
68.222.64.24
In Reply to: Jon, thanks and here are some things I'm sorting through posted by bartc on November 30, 2006 at 10:41:20:
Actually, you have the basics down, except for #1.Coax is not necessarily the best geometry, it is just one of the most conveninet, and available in a wider range of god materials.
Twisted pair with an equivalent amount of shielding, WELL SPACED AWAY FROM THE PAIR, will outperform a coax, but the problem with most commercial twisted pairs is that they are so close to one another (thin insulation, which raises the capacitance and dielectric involvement), and have very little spacing of the shield away from the twisted pair.
That is why I tell folks to make a twisted pair from coax cores, the insulation is thicker, and you can then space the shield away enough to avoid the worst of the shield distortion of the EM fields.
Jon Risch
Follow Ups:
Like your cross connect for a digital IC?Or a twisted pair of Cat 5 teflon wires and surround them with a teflon spacer and add the shields?
What about silver vs. copper?
I actually have on hand silver wire in 29 AWG and teflon tubing to match. Conceivably, I could make a 30" twisted pair of that, surround it with spacer (cotton or teflon), shield it. Would that be of any greater value than the coax already out there on the market for this specific digital application (DAC to CDP)? Zanash recommended something similar over on Digital.
For an RCA plug digital IC, I would stick with the proper 75 ohm impedance cables.Now some folks have used the twisted pair with shield as a digital IC, but I do not personally recommend it, and have always recommended the Belden 1695A with the added shield as the prefered digital IC for 75 ohm SP/DIF use (RCA plug digital IC's).
The twisted pair recipe might work really well if paired with an AES interface, which looks for a 100 ohm impedance, and uses XLR connectors. Here, my recommended analog construction, using the pin 1 ground connect wire wrapped OUTSIDE the shield ofthe twisted pair, would come very close to the mandated AES 110 ohms characteristic Z, and provide many of the benefits needed for clean digital signal transmission.
As for silver vs. copper for a digital IC, to me, this is a much more subtle difference than the insulation or shielding differences.
Jon Risch
That clears up a lot.I'm looking at Blue Jeans Cable for best price and already built up 1695A with Canare RCAPs, but they don't add the extra shield. I can do that myself if necessary.
Was intrigued by the Oyaide FTVS-510, silver core, extra dielectrics, etc., but that's a rather expensive cable.
Outside of that, have you ever experimented with resistively terminating your digital cables? Once you go this route, you'll understand how important VSWR / loading characteristics are at the RF frequencies that digital signals are transmitted at. I would be curious to hear your results if you have experimented with such things. Sean
>
Yes, see:
DIY Cables and RFI/IGP
http://www.audioasylum.com/audio/cables/messages/80108.htmlCable RF Termination and Cable Loading
http://www.audioasylum.com/audio/cables/messages/80110.htmlI think that those will cover most of your concerns.
Sean, I haven't done digital cables. Only analog. That's why I'm asking these questions of Jon. The above is only what I"m gleaning from reading many posts here and elsewhere and I'm asking him if he thinks that's the right paradigm.Meantime, I'm not at all sure what you mean specifically by resistively terminating the digitals. If you mean adding a resistor somewhere somehow to try to match the impedance, I'd be very interested in how that's done. Do you know?
I have experimented under Al Sekela's tutelage with R-Cs on cable terminations of PCs and speaker cabling. But that's for RFI control, not necessarily impedance matching per se, though as I recall it is done because of impedance mismatches.
I'll do what i can to answer your questions Bart : )Resistively terminating a cable can be done two different ways. A common method is to place a resistor across the hot and ground connections at the load end of the cable. The resistor value chosen is typicaly the same as the nominal impedance of the cable being terminated. What one ends up with is that resistor tied in parallel across the input impedance of the device that the cable is terminated into. In most cases, this can help to reduce VSWR and signal based reflections from occuring. In some cases, this approach can load down the output stage of the source component, making things sound worse.
The other method involves much the same procedure, but uses a specific value resistor that is chosen based on not only the cable's nominal impedance, but also that of the input impedance of the terminating component. This approach can net even better results if the source component has a robust output stage, but in many cases, may result in worse sound. That is,in comparison to using either the cable alone or the resistive termination method mentioned above.
Stereophile had a brief article on this a few years back. There was supposed to be a follow-up article, but i never saw one. These approaches can work quite well at the component level, but should not be used on an amplifier / speaker interphase or power cord. This requires something different, which is probably what Al was helping you with. Sean
>
This post is made possible by the generous support of people like you and our sponsors: