Using a coathanger as a digital cable (1 Viewer)

[Matt_Doug]

Bruce, I applaud your thoroughness. I tried something similar with a pair of db9 connectors, picture wire and my serial printer but couldn't get anything to print. I should mention that I was referring to Cat 5e (350 mhz) You'll be hard pressed to find a workstation with an NIC that isn't rated as such. But I admit that cat 3 may be enouugh for some simple file sharing duties on short cross over runs. You'll have problems though connecting to a typical network with cat 3.

 

[brucek]

John, yeah if I can find the time I'll post the results..kinda meaningless though.....and boring.

Matt, yep, I agree. It was only for a little experiment. I just couldn't let that one foot challenge stand...

 

[brucek]

John,

Here are the results of the cable test I promised. I put the little explaination of each test above the results. The data really isn't that useful other than for cursory interest. The useful information is that each test passed the 10MhzCAT3 specifications. Of course, the cable was only a meter long.

(1) Near-End Crosstalk (NeXT) is a measure of noise coupled (excessive crosstalk) between two pairs measured at the same end of the cable as the disturbing signal source. The results show the worst value found within the entire bandwidth tested (10Mhz) at both ends.

NeXT (LOCAL)
PAIRS FREQ NeXT SPEC
1,2/3,6 6.7Mhz 43.2dB 27.0dB PASS
1,2/4,5 8.5Mhz 34.4dB 25.4dB PASS
1,2/7,8 9.4Mhz 31.5dB 24.7dB PASS
3,6/4,5 9.4Mhz 30.1dB 24.7dB PASS
3,6/7,8 8.5Mhz 39.7dB 25.4dB PASS
4,5/7,8 8.6Mhz 37.7dB 25.3dB PASS

NeXT (REMOTE)
PAIRS FREQ NeXT SPEC
1,2/3,6 8.5Mhz 41.3dB 25.4dB PASS
1,2/4,5 9.4Mhz 33.5dB 24.7dB PASS
1,2/7,8 9.5Mhz 31.2dB 24.6dB PASS
3,6/4,5 9.4Mhz 30.4dB 24.7dB PASS
3,6/7,8 9.5Mhz 39.0dB 24.6dB PASS
4,5/7,8 8.5Mhz 37.7dB 25.4dB PASS


(2) Length test locates significant impedance changes or anomalies in a cables length.

Passes all leads.

(3) Propogation Delay is the time in nanoseconds for a pulse to travel the length of the cable.

PAIR DELAY
1,2 4nsec PASS
3,6 3nsec PASS
4,5 3nsec PASS
7,8 4nsec PASS


(4) Attenuation provides loss in the cable at all frequencies and provides worst frequency results. Tested up to CAT4 - 17Mhz.

PAIRS FREQ ATTEN SPEC
1,2 17Mhz 1.7dB 13.7dB PASS
3,6 17Mhz 1.9dB 13.7dB PASS
4,5 17Mhz 2.6dB 13.7dB PASS
7,8 17Mhz 1.6dB 13.7dB PASS


(5) Attenuation to Crosstalk Ratio is the difference between crosstalk and attenuation, measured in dB, at a given frequency. A positive ACR is necessary to assure that a signal at the receiving end of a channel is stronger than the crosstalk interference imposed from adjacent wire pairs.
ACR is therefore an indication of the expected reliability of data transmissions. The results display at the frequency where the worst crosstalk occurs.

ACR (LOCAL)
PAIRS FREQ ACR
1,2/3,6 6.7Mhz 42.9dB PASS
1,2/4,5 8.5Mhz 34.0dB PASS
1,2/7,8 9.4Mhz 31.1dB PASS
3,6/4,5 9.4Mhz 29.5dB PASS
3,6/7,8 8.5Mhz 39.5dB PASS
4,5/7,8 8.6Mhz 37.5dB PASS

ACR (REMOTE)
PAIRS FREQ ACR
1,2/3,6 8.5Mhz 41.0dB PASS
1,2/4,5 9.4Mhz 32.9dB PASS
1,2/7,8 9.5Mhz 30.8dB PASS
3,6/4,5 9.4Mhz 29.8dB PASS
3,6/7,8 9.5Mhz 38.6dB PASS
4,5/7,8 8.5Mhz 37.5dB PASS


(6) The Noise Test counts noise impulses (spikes) greater in amplitude than a minimum threshold on an idle pair cable segment.
Impulse noise is non-repetitive electrical spikes or pulses caused by external electromagnetic fields and is one of the primary causes of inconsistent performance on unshielded pair networks.

Passes with 0 errors


(7) Characteristic impedance is a measure of resistance to AC current flow by a cable.

PAIR Z(ohms)
1,2 98ohms PASS
3,6 97ohms PASS
4,5 97ohms PASS
7,8 94ohms PASS


(8) Resistance is the simple DCR of the cable.
PAIR DCR(ohms)
1,2 0.1, 0.1 PASS
3,6 0.1, 0.1 PASS
4,5 0.2 ,0.3 PASS
7,8 0.1, 0.2 PASS

That's it.

brucek

 

[Matt_Doug]

Bruce, I'm curious as to why crosstalk would matter in your experiment. I understand why it would be an issue in a tightly wound network cable but at 10 gauge per conductor that I presume were each insulated I have a picture of wires hap hazardly strewn on your work area. How much crosstalk can you expect from that?

 

[brucek]

Matt, You're right about the mess on the bench and that crosstalk is not a huge concern at this short distance. These are the standard tests to run to verify a cable passes the spec, so I ran them. There will be mutual inductance between any conductor that passes near another conductor and in longer runs and higher frequencies it is necessary to twist the pairs and interweave the pair sets to reduce the effect......

And so, haven't you hit upon the entire purpose of the experiment to prove that at short distances in relation to the frequency, a cable becomes more of a jumper than a cable and you can ignore some of the rules. Same holds true for the digital audio interconnect that prompted this discussion.....so of course, the coathanger digital audio interconnect works fine...