Twists per inch for STP?

[Jim R.]

I had always planned to make my own solid core STP for my amp, but I've misplaced my original so I can't determin how many TPI it is.  If there is a more ideal number than what is included with the kit, I can also make it anyway I'd like.  I'm talking about signal wiring here -- AC power and filament wiring are going to be heavier gauge and those numbers I know.

-- Jim

[VoltSecond]

I've always used a low torque electric drill to twist up 10% + 2 feet more than I needed. When the cable just started to buckle or the insulation started to deform (which ever came first), I stopped twisting.  I never counted the TPI. The TPI depended on wire gauge and the type of insulation used.  I'd guess around 3 TPI for Teflon.

I suggest that you try to purposely over twist 2 feet of wires so you can see the insulation damage that occurs after the wire ties itself in a knot.

Use two different colored wires in the twist, it makes debugging later easier.

Don't twist teflon wires tight, they'll cold flow over the months and can short out. Murphy says they'll short out when you are showing off the system to someone you're trying to impress.

Notes from OTT on the WEB. OTT quotes 1 turn per inch.
http://www.physics.utah.edu/~kieda/ott.pdf

[Mudhiker]

I don't use any particular formula, I just twist them enough so that they stay twisted with no gaps.  That said, teflon wire will unwind itself a bit so it's work experimenting with some scrap.

[Paul Joppa]

I think 2 to 3 is "normal", whatever that means.

The purpose is to cancel magnetic fields, so in a uniform field there would ideally be an integral number of turns for each straight section, and sharp corners. Of course, fields, which mostly originate in the power transformer or those of other nearby equipment, the field varies a lot. For a varying field you would want "many" turns per space over which the field varies, and rounded corners so there are several turns as the orientation varies. So many turns per inch if the pair is close to the power transformer, an integral number of full turns if the pair has sharp corners, and fewer turns are necessary when the pair is far from the power transformer or has gradual direction changes.

So much for theory - very logical but completely impractical in the real world.  :^)

[Jim R.]

Thanks all for your replies.  2-3 is what I've always heard for filament wiring, and regarding the input signal wiring, my thought was that trying to balance between shielding and capacitance was the major concern, but really, I suppose that it really isn't all that critical except to get signal wiring pretty consistent from one channel to the other.  Thanks also VS for the link to Ott -- will take a look at that (and good to see you on the new forum.)

-- Jim