EMRFD Message Archive 5594

Message Date From Subject
5594 2010-12-28 09:46:22 ae5ew Speaking of Audio Cabling
I will admit that current audio cabling is not ideal when using a PC. Take the standard 3.5mm stereo audio jack/plug. The commons (signal returns) are both tied together at each end of the cable. That shouldn't be a problem with a single, non-spliced, cable. But, let's say we splice in an extra ten feet. The signal return for each signal must be kept separate from each other and must follow the signal it is paired with. If the right channel signal return is swapped with the left channel signal return, expect problems. Worse, or at least, just as bad, is tying both signal returns together at each end of the splice. Granted, these types of cable runs were designed for ONE end to end cable run with NO splices.
Better, was the RCA phono connectors. They keep the signal returns separated throughout the length of the cable irrespective of cable 'splices'. Assuming they are wired correctly, of course. Yep, those cable connectors you find on your antique VCR, record player, tape player, and even modern DVD players.
This also applies to almost any cable runs with signals having individual (paired signal and signal return (twisted pair)) signal returns. Data lines are extremely important to keep the paired signal return with the signal it started with. Switching it with another data signal return is/was a recipe for disaster. Control lines are not any where near as sensitive to return lines. Unless the control line switches very rapidly.
Charles AE5EW
5603 2010-12-28 22:12:34 Andy Re: Speaking of Audio Cabling
> I will admit that current audio cabling is not ideal when using a PC.
> Take the standard 3.5mm stereo audio jack/plug. The commons
> (signal returns) are both tied together at each end of the cable.

More likely, there is only one return "wire", which is shared by both
channels. I think you'll find they are all two conductors inside one
shield.

Even if it were two return wires, they'd be tied together at both
ends, which makes them effectively the same as one wire. Audio is
such low frequencies that the return current doesn't know to take the
"correct" shield that is paired with its signal; it will go back on
any and all shields anyway. (By the time you get up into RF, signal
and return currents tend to stay tightly coupled to one another,
minimizing the loop area.)

The fundamental assumption, with unbalanced audio signals, is that the
signal (& return) current is small enough, so there is negligible
voltage drop on the returns, and little chance for crosstalk. But it
would be there, and maybe even measurable. Just hopefully not
audible.

Splicing an extra length of cable shouldn't change anything, aside
from making all 3 wires (tip, ring, and common) longer.

Andy
5604 2010-12-29 04:48:29 ae5ew Re: Speaking of Audio Cabling
I haven't really made such cables yet. None of my audio cables have been spliced of have any interconnects between the PC and FT-950.
My experience with RS-232 cables is extensive. 1968 to present. I use twisted pair, individually foil shielded pairs (Belden 8777). Indeed, each end is grounded. 25 contact connectors had pin-1 shield grounds which are NOT available on PC 9-contact (a contact could be either a pin or socket depending upon the connector in use) connectors.
Cables were made such that TX data was wired on one of the pairs while RX data was wired on another pair. The third pair was reserved for DTR say from a printer for buffer overflow control. One end was wired to a computer while the other end went to a monitor which usually ran at 9600, 19200, or 38400 baud. I could easily run 19200 baud out to 500 ft from the computer while 38400 baud could be ran to 250 ft. These are real life distances, not estimates. The shield of each pair was grounded at the computer end only. Some of the large systems with 25 - 100 plus terminals would use an interconnect panel located at the computer end. This made it much easier to wire up everything initially and to make port assignment changes easier.
Signal returns were wired to contact-7 at each end. Each cable went to its own assigned port connector at the computer. So there was a signal return for TX data (TX-R for TX Return) and RX data (RX-R for RX Return) and the active TX data (TX-D for TX Data) and RX data (RX-D for RX data) contacts and wires. I'm going to ignore the third pair. Ok, so we now have for any port:
Pair 1: TX-D, TX-R, TX-Shield and
Pair 2: RX-D, RX-R, RX-Shield.
The interconnect panel was approximately 15 to 20 cable feet from the computer. IF, the TX-R and RX-R wires were accidentally switched at the interconnect panel, the communication link was useless. Garbled data even when not TXing or RXing data. NOTE: only the Return wires were switched. Put the return wires back where they belong and the communication link was solid with no garbled data. Yes, the wires go to the same end point, but, they hate being switched. I never tested the minimum cable length at which this would occur. This situati