Phase cable upgrade with magnet wire

[auraslip]

Finally got the money to order the wire - 20awg is the largest they have - lets see if the 300c temp rating is pixie dust.

Any idea for a north american supplier of stuff similar to the HK 26awg silicon wire rated at 200c - possibly 30awg wire? I don't want the halls to be magnet wire too.

edit: found some mil spec stuff on ebay - 30awg 200c and 100ft of it for $16 shipped http://cgi.ebay.com/100-30-AWG-Teflon-High-Temperature-Voltage-Wire-White-/120642013985?pt=LH_DefaultDomain_0&hash=item1c16d30b21

 

[auraslip]

OMG IS THIS A PAIN IN THE ASS.

EDIT:

I just reread johns original post where he talked about staggering the ends of the phase wire to help with the bend at the bottom of the axle. Well to late for the last bundle: the magnet wire is all scrapped up. Even inside the tape which is still fine.

 

[John in CR]

OMG IS THIS A PAIN IN THE ASS.

EDIT:

I just reread johns original post where he talked about staggering the ends of the phase wire to help with the bend at the bottom of the axle. Well to late for the last bundle: the magnet wire is all scrapped up. Even inside the tape which is still fine.

That's why my posts are often so long. I try to be thorough. Look on the bright side, at least you proved that staggered ends was the way to go.

I don't understand "last bundle". Didn't you put all wires in tape and then in the single shrink to send it all through the axle in one shot protected by tape and shrink? No way sending each phase through one at a time will work. No way would I try to feed through a bundle with tape as the outer layer either, because the edges of the tape would catch. I don't understand what scraped the wires.

 

[auraslip]

Just one bundle wrapped in tape. Don't have the right size shrink on hand at the moment. Heck, I was willing to bet before this that the setup would be safe with out out shrink or tape.


Funny thing is it went through fine the first time - except I broke one of the 30 awg wires yanking it through with needle noses.
The second time I had trouble at the bend and inner lip of the axle exit hole scrapped some of the insulation of the magnet wires. That lip is sharp, and I can't really fit a file in there to smooth it down. (well maybe I can figure something out)

Anyways, thanks for the encouragement. Staggering seems the way to go. How does that work though? Three different lengths of magnet wire? Trim it down once it goes through?

 

[Pure]

I don't know about using this stuff for long term aps. It seems to me, vibrations causing the wires rubbing against each other might become an issue with the varnish.

 

[John in CR]

I don't know about using this stuff for long term aps. It seems to me, vibrations causing the wires rubbing against each other might become an issue with the varnish.

What do you think your motor is wound with? Each turn must be insulated from the others. Obviously you would need to do something to avoid jiggling wires, but you don't want that in your motor or axle with regular wire either.

 

[John in CR]

Here's a perfect example of why this works so well, as long as you can put up with baring the ends to electrically join strands to form phases.

I'm reharnessing a motor that originally had 12ga, but it's simply not large enough for the power we're using. Here are 3 pieces of the thinly insulated 12ga taped as tightly together as I could. Next to it is 30 strands of 18ga magnet wire, which works out to 8ga. They are effectively the same size to fit through a circular hole. After considering the thick outside insulator on the original harness, I'll be able to easily add another strand or 2 to each phase as well as double up on the shrink and slather high temp silicone so it won't be loose. IOW I'll be able to get up to close to 7 gauge wire for my phase wires in this motor, and still have a better insulated wire than the original.




EDIT- Here's the 8ga equivalent harness with 2 turns of electrical tape at the end and 5 at the overlap, compared to the original motor harness. I've got plenty of room for even more copper, or maybe just a nice high temp silicone injection into the axle tunnel. Note that phase wires will be an insignificant addition, since I'm using 23ga magnet wire for them (screw wire insulation ever melting or cracking).



Anyone trying this, don't forget to stagger the phase wire ends to make it easier to get past the bend at the end. You can even pre-curve the ends when you do it. Sorry but I don't have an easy answer still for getting the insulation varnish off, so get 16 or 18ga to keep the work to a reasonable level.

Outside of this particular motor I am running dual 10ga normal wire for each phase down to the magnet wire.

John

 

[hillzofvalp]

why isn't this done from the factory...

I'm skeptical of the insulation on that wire and how necessary it is to have 7ga phases. 9-10ga is pushing it.

 

[John in CR]

why isn't this done from the factory...

I'm skeptical of the insulation on that wire and how necessary it is to have 7ga phases. 9-10ga is pushing it.

I wish the factories did do it. No, it's not necessary for whimpy ebike hubbies, but this motor will see 400-600amps on the phase wires and heats up 10ga even. Also, it's a low turn count motor, and I'll know for sure soon enough, but it looks like the windings have at least 10ga.

I fail to see what you guys worry about the insulation for. This harness will be protected by the magnet wire varnish, a double layer of electrical tape, a double layer of shrink wrap, and finally a thick layer of high temp silicone. On top of that, the thick rubbery plastic original harness is problematic at the exit due to the contraints of the bike frame and the spinning interior portion of the freewheel. It's not an axle end exit that really suck if your bike goes down on that side. My new harness will be easier to bend into a shape that it will retain, so rubbing the freewheel is much less a concern than the original. Pics to come of the effort, but insulation will be far less a concern now, especially since the wiring won't heat up.

Even if somewhere between 7 and 8ga is overkill (though it can't be actual overkill, since if it is from a current standpoint it will sink heat away from the windings via a copper route), I have room for it so why not?

 

[The Mighty Volt]

I'm intrigued by this, but I don't know exactly what is happening here.

My basic understanding: High-Power applications require bigger phase wires, and people are having difficulty getting new, enlarged phase wires down into the axle of the hub.

So you guys have switched over to a new type of wire.....which you are taking the insulation off....for some reason.

Why are phase wires insulated in the first place then???

Sorry to be so noobish about this. I am intrigued as I have a sort of low-grade version of the Moped motor JohnCr has, and this sort of job is probably in my very near future.

Cheers.

 

[John in CR]

MightyVolt,

This is magnet wire, like what motors are wound with. It has insulation, a very thin varnish. Because of this insulation, I don't need a separate insulator between the 3 phases going through the axle, which enables a nice copper fill that is round, something 3 already made insulated wires can never accomplish. That's why I shared the comparative pics above. I can easily get a harness of 3 electrically separate phases larger than 8 gauge through a hole that was a tight fit for the 12ga phase wire harness from the factory. This way I cut resistance by way over half, and I have magnet wire, but no suitable 12ga wire anyway and 10ga wouldn't fit.

 

[The Mighty Volt]

MightyVolt,

This is magnet wire, like what motors are wound with. It has insulation, a very thin varnish. Because of this insulation, I don't need a separate insulator between the 3 phases going through the axle, which enables a nice copper fill that is round, something 3 already made insulated wires can never accomplish. That's why I shared the comparative pics above. I can easily get a harness of 3 electrically separate phases larger than 8 gauge through a hole that was a tight fit for the 12ga phase wire harness from the factory. This way I cut resistance by way over half, and I have magnet wire, but no suitable 12ga wire anyway and 10ga wouldn't fit.

Cheers. That's cleared it up a lot. So then....why are some people mentioning removing insulation at all in the first place?? Are they replacing it with something superior?

Basically you have the thinnest, least-volume occupying insulation available, which allows for more copper, which allows for more successful conducting of current.

So....you have phase wires overlapping everywhere, safely, thanks to this insulation.

 

[jonathanm]

You need to remove the varnish insulation at the ends to solder the wires together and onto your connectors etc. The baked on varnish can be very tough to get off.

 

[The Mighty Volt]

You need to remove the varnish insulation at the ends to solder the wires together and onto your connectors etc. The baked on varnish can be very tough to get off.

Ah...so only a very small amount is removed, at the ends, cheers.

 

[John in CR]

You need to remove the varnish insulation at the ends to solder the wires together and onto your connectors etc. The baked on varnish can be very tough to get off.

Ah...so only a very small amount is removed, at the ends, cheers.

It may seem like a small amount, but it's the main issue with this method if you value your time. This time I'll have 66 or 72 ends plus another 10 for the hall sensors, so I'll be trying some new ways other than scraping with a razor knife. Hopefully something like a dremel + sanding disk does the trick to do more than one at a time and make quick work of it.

 

[The Mighty Volt]

A chemical sounds like the best solution, no pun intended.

Dip to the required depth, remove, hang up and come back later, all nicely eroded away.

Any ideas anyone? Some sort of a solvent??

 

[hillzofvalp]

I think I read of people using chemicals for this purpose? dremel jig seems like most likely option

 

[amberwolf]

If you wanted to add some sealant to it that will flow into all cracks and crevices, places like this:
http://www.kyocera-chemi.jp/english/prdct/list/cm/varnish.html
offer non-solvent varnish you can put right in there that shouldn't damage the existing varnish.

If you wanted to be doubly sure about insulating the phases from each other, with minimal waste of space, you could add a Y-shaped core insulator like this:

if it was thin enough material, like Kapton, it could even be made to conform a bit to the shape of the wires it's sandwiched between, so that you could still interlace the shapes of phases between each other (assuming diameter of wires vs total space in axle is correct for this).

Even without that, it would not reduce the active area by too much.

Then wrap the entire bundle in something really thin but tough, like Kapton tape, or similar.

Once it's all done and in the axle, force in some liquid but non-solvent varnish for vibration stabilization and further insulation.


I had hoped they'd have a document on stripping the ends, but I didn't see one. I did find this site:
http://www.wiretron.com/free.html
that has software supposedly with that in it, but for some reason I can't access the actual download to test that (it times out).

 

[John in CR]

I tried the aspirin+heat trick, and some couple of the other acid ones I read about all to no avail. Heating didn't work either. It's been 6 years or so, but I think the wire shop I bought it from only had the high temp kind for motors in the gauges I bought, which worked out perfect because I wasn't buying for motors. It was to wind chokes for speaker projects where I didn't need the more expensive high temp kind. It seems that the quick and easy methods are more for the really thin stuff for other purposes. Next time I get to the other side of town, which is rare, I'll stop by a motor rewind shop I know of and ask. The way they slather around the extremely toxic smelling vinyl ester resin to glue windings in place on brushed motors, I wouldn't doubt they have some industrial strength acid or solvent that eats anything but the copper.

I'll be happy if I can go at multiple ends at once with a dremel without eating away much copper.

 

[John in CR]

If you wanted to add some sealant to it that will flow into all cracks and crevices, places like this:
http://www.kyocera-chemi.jp/english/prdct/list/cm/varnish.html
offer non-solvent varnish you can put right in there that shouldn't damage the existing varnish.

If you wanted to be doubly sure about insulating the phases from each other, with minimal waste of space, you could add a Y-shaped core insulator like this:

if it was thin enough material, like Kapton, it could even be made to conform a bit to the shape of the wires it's sandwiched between, so that you could still interlace the shapes of phases between each other (assuming diameter of wires vs total space in axle is correct for this).

Even without that, it would not reduce the active area by too much.

Then wrap the entire bundle in something really thin but tough, like Kapton tape, or similar.

Once it's all done and in the axle, force in some liquid but non-solvent varnish for vibration stabilization and further insulation.


I had hoped they'd have a document on stripping the ends, but I didn't see one. I did find this site:
http://www.wiretron.com/free.html
that has software supposedly with that in it, but for some reason I can't access the actual download to test that (it times out).

The divider thing is cool, but I see that and other fancy time munchers as really not needed. On my first attempt at this I thought my magnet wire was 20ga, not 18ga, so I ended up with larger than 10ga through my 9C axle with room to spare. I wrapped the bundle in electrical tape first, mostly to get it in a nice tight bundle, and then the shrink fit perfectly. No, it's not the kind of wire bundle you want to beat around or flex back and forth a lot, but it seems like everyone forgets this stuff is used for the windings in our motors where each turn needs to be electrically insulated from each other, and the wire has alternating magnetic forces on those plain magnet wires. The Chinese put varnished string to tie windings in place, but that stuff always disintegrates over time. Tightly taped and double shrink will surely prevent any chaffing between strands, so this is easily more secure than motor windings.

My wire bundle will have multiple layers of protection, so as long as you secure it well, something necessary with any wire through the axle, and add extra protection to areas with potential to wear, it should be better than original in all respects. To put my money where my mouth is, I've decided to use it for my entire phase wire run. For the segment outside the motor I'm going to separate each phase with color coded shrink just in case I ever need to cut it down the line. In my only exposed area, which is the drip loop, I'm going to add some extra layers of protection to the exterior.

I've got my bundle ready for skinning the ends, and it looks like 37 strands plus the halls will fit nicely with room for double shrink and a good cushion of silicone between the metal of the axle and the outer shrink. That's 11 for each phase, which is larger than 8 gauge and almost equal to double 10ga, leaving 4 extra strands for things like fans and temp sensors. Since this is to be a life of the motor fix, it's prudent to include some extra strands. Maybe I should include a spare set of 23 gauge hall wires too just in case. 8)

I can't wait to see the kind of phase wires I can do on Hubmonster, whose 10 or 11ga phase wires were getting hot.

John

 

[johnrobholmes]

The way I see it you have three main options to remove the insulation- scraping, sanding or heat. Sanding is hard to be precise unless the wire is thick enough to work on something like a belt sander. A knife ground to the shape of the wire is the quickest when scraping off the insulation, but is really damn aggravating as you already know.

Heat will be the fastest and least labor intensive overall, it will require a very hot torch or iron so you don't damage the insulation too far down the line. A welding arc or direct contact with tungsten can be used to burn if off too, but requires some tooling and prep on your behalf. The faster you can apply enough heat to burn the insulation, the better. I would suggest burning the wires as a bundle, then separating them and beating or rubbing the burned insulation off. A stainless or brass brush makes quick work of cleaning it well enough to solder or braze.

Directing an oxygen enriched torch to the crown of the wires would be my method of choice. As soon as the copper shows a hint of dun red color you have done enough. Then take the bundle and hit it on the bench, scrape with a brush, then do the other side.

 

[Toorbough ULL-Zeveigh]

never tried chemical stripping off insulation but now that i found out how to i'm sure that i never want to.
another chemical method found googling was that acrylic lacquer thinner supposedly works better than regular lacquer thinner, in hours of soaking rather than a day.


http://en.wikipedia.org/wiki/Enameled_wire

For ease of manufacturing inductive components like transformers and inductors, most new enameled wire has enamel that acts as a flux when burnt during soldering. This means that the electrical connections at the ends can be made without stripping off the insulation first. Older enameled copper wire is normally not like this, and requires sandpapering or scraping to remove the insulation before soldering.

http://en.wikipedia.org/wiki/George_Jacobs_(inventor)

The three Fort Wayne companies, Rea, Phelps Dodge Magnet Wire, and Essex account for 2/3 of the world's production of magnet wire.

and they say there's no more manufacturing going on in america.

that last bit about flux added in the enamel reminded me of the importance of proper flux.
on diyaudio forums someone pointed out that some solders have more aggressive flux (not acid flux) that strip the varnish better than others during soldering.
kester 88 being one that's preferable for biting thru.
i have no idea how you can tell what you got wrt flux except try different types.

my suggestion on a 4th method...
if you know anyone that you can borrow a one watt blu laser, maybe give that a shot. :wink:

 

[John in CR]

While I'm sure a torch might burn it off, I have 2 problems with that route:
1. You're still left with scraping/cleaning anyway.
2. Within about 3" the phases will be joined as one bundle, and copper conducts heat so well that there's no way I could trust that the heat of burning at the end doesn't weaken and damage the varnish far enough up the wire to make me nervous. Holding the wire in ice water with only the end exposed for burning would do the trick, but it sounds like enough trouble combined with #1 that I'll leave it to someone else.

My big advantage with the dremel sanding route is that with my 25yro son I have 4 hands to make short work of it. That's 2 or 3 hands to hold the wire in parallel groups and roll and or rotate them.

I will pick up some paint and varnish stripper/remover and give that a go first, since dip-wait-wipe off would be ideal.

 

[The Mighty Volt]

Slightly off topic....but what happens if phase wires cross and short??

 

[John in CR]

Slightly off topic....but what happens if phase wires cross and short??

Adios controller and immediate plug braking. That's what I'm working to avoid with using magnet wire. Heat is the enemy. The strands can no more rub together and chafe the varnish off wrapped tightly in a well secured bundle than they can wound on the stators of our motors.