CAT3 Speaker Cables

[Am386SXL]

I bet many readers have seen plenty of speaker cables made with CAT5, but in my approach I only had CAT3 available to me and I didn't want to make a huge effort of it. The difference between CAT5 and CAT3 is not much, CAT3 is isn't used for LAN networks anymore, but for voice networks (PBX/PABX) instead. CAT3 consists of an X amount of pairs of PSTN cable strands, in my case I had 8 pairs giving me 16 strands of copper.

It's so easy to make, I hardly needed any tools. With some effort, the jacket and individual wires can be skimmed by hand and twisted together by hand. It only need to be tinned (soldered) a bit to keep the twisted wires together when putting on banana plugs. Everything was a perfect fit except for tightening the screws on the plugs needed a little effort.

I took a multimeter and did a continuity test, all four cables passed first time. Switched over to resistance test. To make the test easier, I added the very same banana plugs to the probes of the multimeter. The probes alone tested 1.0 ohms. Each cable, at 1.8m each, tested 1.1 ohms within margin of error. I would say, moving to the CAT3 cable (from Kirlin SM-18 OFC Cable) made a difference to the sound. I really wish I had better instruments to measure these cables properly, but I am happy with the results. I must say, them cables really looks boss on my NAD C300.

Probe test.

Cable #1 resistance test.

Cable #2 resistance test

First completed pair.

Connected to NAD C300

 

[TheWarrior]

I'll bet they do make a 'difference' in the sound quality!

This is one situation where speaker cables WILL make an audible difference!

Ever notice all the threads going on about speaker cable gauge? Well, 12 gauge is the standard recommendation, 14-16 can be used for short runs... but your cable is something like 24 gauge, which is WAY too small for handling audio!

Do NOT use these cables with any higher power amplifier, and do NOT EVER run them in walls!

Seriously! This is an electrical fire waiting to happen!

 

[TLS Guy]

The resistance of your cables is too high by a factor of 10 or more and are totally unsuited to be speaker cables.

 

[highfigh]

I'll bet they do make a 'difference' in the sound quality!

This is one situation where speaker cables WILL make an audible difference!

Ever notice all the threads going on about speaker cable gauge? Well, 12 gauge is the standard recommendation, 14-16 can be used for short runs... but your cable is something like 24 gauge, which is WAY too small for handling audio!

Do NOT use these cables with any higher power amplifier, and do NOT EVER run them in walls!

Seriously! This is an electrical fire waiting to happen!

Each strand is 24 ga, not all of them combined. Eight strands of 24 ga is about the same as 15 ga.

I think the meter's leads need to be addressed- 1 Ohm is too high and I would be suspicious of any readings.

 

[TheWarrior]

Each strand is 24 ga, not all of them combined. Eight strands of 24 ga is about the same as 15 ga.

Not debating... trying to make sure I understand this correctly:

Cat 3, as I understand it, is only 4 strands. How do you sum that?

 

[Speedskater]

The wire math for 24AWG:
1 strand = 24AWG
2 strands = 21AWG
4 strands = 18AWG
8 strands = 15AWG
16 strands = 12AWG
32 strands = 9AWG

************************
For speaker cable purposes, there is no difference between Cat3 and Cat5 or any other Cat with 24AWG strands.

Best way to make the speaker cable is to tape 2 or more Cat3 cables together and connect all the solids to one terminal and all the stripes to the other. No weaving or braiding is needed.

************************
Most budget DMM have problems making accurate reading a below 1 Ohm.

 

[TLS Guy]

Each strand is 24 ga, not all of them combined. Eight strands of 24 ga is about the same as 15 ga.

I think the meter's leads need to be addressed- 1 Ohm is too high and I would be suspicious of any readings.

Cat 3 is not intended as a speaker cable. It would be most unusual for any meter to be off by a factor of 10. He shows a resistance of 1 and 1.1 ohms, whereas his resistance needs to be 0.1 ohm. With those readings his frequency response will follow the impedance curve of the speaker. Not good at all.

 

[Speedskater]

A 24AWG conductor has a resistance of about 25 Ohms per 1000 feet.

 

[highfigh]

Cat 3 is not intended as a speaker cable. It would be most unusual for any meter to be off by a factor of 10. He shows a resistance of 1 and 1.1 ohms, whereas his resistance needs to be 0.1 ohm. With those readings his frequency response will follow the impedance curve of the speaker. Not good at all.

The resistance of the Cat3 was .1 Ohm- 1 Ohm was the meter's leads.

I read an article in one of the audiopile rags by some clown who was incredibly proud of himself, for braiding three Cat5e for his speaker wires, then braiding the striped wires at the ends as positive and the plain ones for negative. Said they were the best he had heard. Makes me want to bang my head on a wall.

 

[highfigh]

Not debating... trying to make sure I understand this correctly:

Cat 3, as I understand it, is only 4 strands. How do you sum that?

That was my first thought, but the OP wrote that he had four pairs, so I worked from that and ignored the 3. Four conductors would be 18 ga.

 

[Rickster71]

This thread got me thinking about that CAT-5 braided speaker wire link and wire shootout that was going around about ten+ years ago. Anyone remember that? I even think it was posted in AH.... didn't Gene do a comparison? Thought I saved to favorites... still looking.
I found one, but it's not the comparison with the measurements:
http://www.tnt-audio.com/clinica/cat5questions_e.html

 

[Rickster71]

A little search and I found this:
http://www.audioholics.com/gadget-reviews/diy-speaker-cable-faceoff

 

[Am386SXL]

Thank you for the replies.

The cables seems to work fine, I didn't notice a rise in operating temperature of the amplifier. I know that the multimeter isn't the best, I still want to test the cables on a LCR Meter, my father has a Toptronic T24.

Each strand is 24 ga, not all of them combined. Eight strands of 24 ga is about the same as 15 ga.

I think the meter's leads need to be addressed- 1 Ohm is too high and I would be suspicious of any readings.

The wires of the leads are extremely thin since it is a cheap multimeter. That is why the impedance of just the leads alone is 1.0. On the meter's specsheet is says +/- 0.8% of reading + 3 digits for resistance test.