Cat 5 has 8 strands. For internal wiring, is it best to use a single strand for each terminal or four strands for each, or something in between? Or wires of 8 strands each?
What about stripping the outer sheathing off and using the wires loose or keeping it as one 'big' wire except at the ends?
What's the consensus? I see this mentioned a lot, but no one explains exactly how they use it.
Doug
What about stripping the outer sheathing off and using the wires loose or keeping it as one 'big' wire except at the ends?
What's the consensus? I see this mentioned a lot, but no one explains exactly how they use it.
Doug
All of the above have been done, and more variations besides. There's no consensus because audiophiles never seem to agree on anything!
In the pair of BIB's I'm building I'm using all eight strands, 4 colors for +, 4 white for -. For a longer run, I might braid three runs of Cat 5, and then terminate them the same way (12 for +, 12 for -).
In the pair of BIB's I'm building I'm using all eight strands, 4 colors for +, 4 white for -. For a longer run, I might braid three runs of Cat 5, and then terminate them the same way (12 for +, 12 for -).
I would braid even more, perhaps 24 strands for + and another 24 strands for ground, using braided 3×CAT5 as ls connect sucks, not enough bass, braided cable made from 2×12×0,75 mm2 sounded ways better
http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&category=81950&item=7602573999
http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&category=81950&item=7602573999
You didn't state what drivers you are using, but let's assume that you are using reasonably efficient full-rangers. That means that you max power levels are in the 5 watt range. A 3 ' run of single pair Cat5 gives you about 0.15 ohm for the out and back. Assuming an 8 ohm driver, that gives you a 2% voltage drop in the wiring. Some one else can finish the math, but this is an inaudible drop in SPL. There is no need for more than a single pair of Cat5 for internal wiring in a reasonably high efficiency speaker.
Using the same logic, my 12' speaker cables are 2 pair star wound Cat5. This is probably over-kill, but I've at least done SOMETHING more than just a single run of Cat5. The high count Cat5 cable formulas make no sense except to jack the cable capacitance out of sight.
Bob
Using the same logic, my 12' speaker cables are 2 pair star wound Cat5. This is probably over-kill, but I've at least done SOMETHING more than just a single run of Cat5. The high count Cat5 cable formulas make no sense except to jack the cable capacitance out of sight.
Bob
Hi,
I have seen the CAT5 recommendation. I like it because you have separately insulated cores.
Try adopting different core combinations for the different drivers.
Two or three pairs (.4sqmm/.6sqmm) for the treble driver. Four pairs (.8sqmm) for the midrange driver. Six or eight pairs (1.2sqmm/1.6sqmm) for the bass driver.
For the short runs inside the speaker these should adequately cope with voltage drops and heat generation.
Have you considered using the CAT5 option for the external cables? They are relatively cheap, even if you go for 8pairs upto 3m long and 12pairs for 3m to 6m long. 12 pairs at 6m long per channel comes to 36m of CAT5 or about $10 for stereo. Bi-wiring only adds $5 for the treble units or a total of about $20 to tri-wire at 6m long to each side.
I have seen the CAT5 recommendation. I like it because you have separately insulated cores.
Try adopting different core combinations for the different drivers.
Two or three pairs (.4sqmm/.6sqmm) for the treble driver. Four pairs (.8sqmm) for the midrange driver. Six or eight pairs (1.2sqmm/1.6sqmm) for the bass driver.
For the short runs inside the speaker these should adequately cope with voltage drops and heat generation.
Have you considered using the CAT5 option for the external cables? They are relatively cheap, even if you go for 8pairs upto 3m long and 12pairs for 3m to 6m long. 12 pairs at 6m long per channel comes to 36m of CAT5 or about $10 for stereo. Bi-wiring only adds $5 for the treble units or a total of about $20 to tri-wire at 6m long to each side.
I've used Cat 5, just splitting up a single run as stated above, 4 wires per side. When doing this, it's been said it's best to split the solids and stripes to + and -. Each twist inside the outer housing twists at a different rate, so their shielded by design.
I've tried one wire per side, (+ and-) and didn't care for it, no low end. Could work for supertweets though.
However, I'm now using 47 labs OTA with 1 wire per + and -. Has to be about the same gage as Cat 5, strange thing is, this seems to work realy well with plenty of low end. Truly wonderful wire if you can swing it, or find smaller pieces for sale.
I've tried one wire per side, (+ and-) and didn't care for it, no low end. Could work for supertweets though.
However, I'm now using 47 labs OTA with 1 wire per + and -. Has to be about the same gage as Cat 5, strange thing is, this seems to work realy well with plenty of low end. Truly wonderful wire if you can swing it, or find smaller pieces for sale.
FYI Bob Brines: Right now, I'm working with a Fostex FE108 Sigma, 92 dB. Given your input, it sounds like any full ranger or two-way would be covered by one 8-strand cable (or more if the builder chooses).
Well, it's a relief to know that just about any combination will work for short runs. So I'm just going to stick with four strands each in the future for full rangers.
one1speed: I just realized, aren't the solids and stripes of each color in the same twist? If so how could they be twisted differently. Now I'm confused. It's almost lunch, so I will go look at some cable and get right back.
Doug
ps: I edited my original reply to one1speed.
Well, it's a relief to know that just about any combination will work for short runs. So I'm just going to stick with four strands each in the future for full rangers.
one1speed: I just realized, aren't the solids and stripes of each color in the same twist? If so how could they be twisted differently. Now I'm confused. It's almost lunch, so I will go look at some cable and get right back.
Doug
ps: I edited my original reply to one1speed.
With Cat 5 each pair has a slightly different rate of twist.
I use 30' of 4 pair parralleled (i.e. one Cat-5 cable) on a FE-166E BK-16, I was a little worried about capacitance but no noticiable issues with amp stability (med-low feedback Dyna-mod) or High Frequency response when compared to 8ft of 14 gauge.
Sean
I use 30' of 4 pair parralleled (i.e. one Cat-5 cable) on a FE-166E BK-16, I was a little worried about capacitance but no noticiable issues with amp stability (med-low feedback Dyna-mod) or High Frequency response when compared to 8ft of 14 gauge.
Sean
Ok, I'm back. I stripped a one foot section of cable.
The twist sequence of Cat5 is as follows, from loosest (or slowest twist) to tightest twist:
Brown pair: Loosest twist
Orange pair.
Blue pair.
Green pair: Tightest twist.
The green pair is twisted at twice the rate of the brown pair.
So how would you implement one1speed's methodology? I realize it probably makes absolutely no difference but I'm curious just as an exercise in electrical engineering (of which I know nothing about).
Doug
The twist sequence of Cat5 is as follows, from loosest (or slowest twist) to tightest twist:
Brown pair: Loosest twist
Orange pair.
Blue pair.
Green pair: Tightest twist.
The green pair is twisted at twice the rate of the brown pair.
So how would you implement one1speed's methodology? I realize it probably makes absolutely no difference but I'm curious just as an exercise in electrical engineering (of which I know nothing about).
Doug
Oh yeah, that part makes sense. I guess what I was asking is how to utilize the twist patterns of the four pairs?
For instance, for two sets of four strands each to the driver + and - :
1. Use four strands of solid and four strands of striped: This would give you two identical twist patterns for + and -.
2. Use the two most loosly twisted pairs (brown/orange) for the + and the two most tightly twisted pairs (blue/green) for the -. This would give you the largest dissimilar twist configuration for the + and -.
That's kind of what I was getting at. I was just wondering how the shielding came into play here.
Doug
For instance, for two sets of four strands each to the driver + and - :
1. Use four strands of solid and four strands of striped: This would give you two identical twist patterns for + and -.
2. Use the two most loosly twisted pairs (brown/orange) for the + and the two most tightly twisted pairs (blue/green) for the -. This would give you the largest dissimilar twist configuration for the + and -.
That's kind of what I was getting at. I was just wondering how the shielding came into play here.
Doug
Considering the frequencies CAT 5 was designed for I doubt there is any capacitance issues at even dog, cat and bat audio frequencies. Inductance would be more of a concern and that gets dealt with by the way CAT 5 is constructed to minimize inductance also. Each twisted pair should be + & - to minimize inductance. EIDE 40 conductor ribbon cable is terminated +-+-+-+-+-+- etc. to get very low inductance.Bob Brines said:
http://www.venhaus1.com/diycatfivecables.html
http://www.audioholics.com/techtips/setup/interconnects/DIYSpeakerCablesp1.html
For my long runs, I cut off the sheathing and remove one of the twisted pairs and braid the other three.
I do this 3 times leaving me with 3 wires each having 3 twisted pairs of wire. Another way to say it is that I am left with 3 wires each having 6 total wires, 3 striped and 3 solid in color.
I take those 3 wires and then braid them to make one cable. I split the stripes and solids and that give me one run for positive and negative.
I will be using these only for the midrange and tweeter runs. I will use standard OFC wire for the bass drivers.
http://www.audioholics.com/techtips/setup/interconnects/DIYSpeakerCablesp1.html
For my long runs, I cut off the sheathing and remove one of the twisted pairs and braid the other three.
I do this 3 times leaving me with 3 wires each having 3 twisted pairs of wire. Another way to say it is that I am left with 3 wires each having 6 total wires, 3 striped and 3 solid in color.
I take those 3 wires and then braid them to make one cable. I split the stripes and solids and that give me one run for positive and negative.
I will be using these only for the midrange and tweeter runs. I will use standard OFC wire for the bass drivers.
A couple of weeks back I made cat 5 speaker cables and interconnects, this has proven very interesting, in my case the cables are probably total overkill 24 strands in total using three lengths still in the blue sheath. All the white striped ones joined and all the coloured ones joined. THe three cables were braided together.
The results are incredible, 7 people have now heard these in A/B tests and everyone picked the cat 5 as being utterly superior to the regular heavy duty cables I was using, no hesitation at all. I imagine a much smaller number of cables would be fine, just as Bob said. Absolutely no downsides sound wise that I can detect.
BTW I am using a pair of very efficient 4 inch based speakers and a 20 watt per channel gainclone.
The interconnects use a whole cat 5 with 4 strands for plus and 4 for minus, with wrapped foil inner sheath (earthed) teflon plumbers tape outside this and then shoelaces as an outer sheath. The difference these made was almost as great as the speaker cables once again A/B tested with 4 subjects and the same 100% verdict! only downside here was they were a bit tricky to make.
I'm certainly not one to be up on all the maths so I'm not sure why it works so well on my system but "Bang for Bucks" it is utterly amazing! No matter how you decide to use them I feel confident they would do the business.
The results are incredible, 7 people have now heard these in A/B tests and everyone picked the cat 5 as being utterly superior to the regular heavy duty cables I was using, no hesitation at all. I imagine a much smaller number of cables would be fine, just as Bob said. Absolutely no downsides sound wise that I can detect.
BTW I am using a pair of very efficient 4 inch based speakers and a 20 watt per channel gainclone.
The interconnects use a whole cat 5 with 4 strands for plus and 4 for minus, with wrapped foil inner sheath (earthed) teflon plumbers tape outside this and then shoelaces as an outer sheath. The difference these made was almost as great as the speaker cables once again A/B tested with 4 subjects and the same 100% verdict! only downside here was they were a bit tricky to make.
I'm certainly not one to be up on all the maths so I'm not sure why it works so well on my system but "Bang for Bucks" it is utterly amazing! No matter how you decide to use them I feel confident they would do the business.
audioaficionado said:
You are correct about out-of-the-box Cat5 cable. But, that is not what I was referring to. THIS is the 27 wire braided formula I had in mind: http://www.venhaus1.com/diycatfivecables.html. Tell me that this stuff doesn't have high capacitance!
You are also correct that all of these tweaks etc only apply way above the audio frequency range and that was sort of my point anyway.
Bob
Bob Brines said:
Okay, Audioaficianado and that link at least answers my question in Post #14 of whether you want an identical twist pattern between + and - or a dissimilar twist pattern. From what I read in that link, the author clearly favored the identical pattern by virtue of not separating the twisted pairs and then using the solid for one polarity and striped for the other.
Ahhh, but what if you only used one of the four colors of the twisted pairs to make up your cables, huh? If you used all brown or all blue, would you hear a difference between the two due to the different rates of twist? C'mon guys, this can go a lot deeper
Doug
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.