Hub motor rewinding, what thickness of enamelled copper wire?

Status
Not open for further replies.

Chalo

100 TW
Joined
Apr 29, 2009
Messages
12,407
Location
Austin, Texas
jamiejackherer said:
and have 6 kids

Bless your heart.

Some adversities are self-inflicted.
 
Then i challenge you to measure it. Take a slot, without insulation and measure it. Calculate area exactly. Divide wire area by slot area. You’ll get a fat apology if above 30%, a normal one if above 25%

If your motor is ready (and i guess it is) then the next time you do a motor you can PM me the results. I’ll save the apology for then. Alright?
 
I’ve done a few motors. It’s surprising how the slot fill can look so full but fill be so low, and this motor does not look so perfect, this is due to the bundled enters and exits, low turn count and crossed wires. This costs fill.

my comments come across as negative and you have the reason to be happy with this winding no matter what. Sorry. I think you made a great effort and it’s really good that you finished it :bigthumb:


It seemed like it hasn’t the same wind on every tooth, is it like this?
 
I'm not convinced this is that incredible. We all know very well that these motors can be run at several times their boiler plate power.

I've got the exact same 1500W voilamart motor, stock winding. I'm feeding it 16s at 80 phase amps. I've been blatting around the countryside at 30+mph and the motor is barely warm to touch. Like... You could be excused for thinking it caught the sun a bit.

I've also got some 8080 motors that I can easily feed 50amps, 2kW and they weigh literally 1/5 as much as that voilamart motor.

Given your original burn out pic, I suggest it burnt out because there was some winding short, a few wires broke and the rest took the full load, the wire was extremely thermally insulated at that point... something, like that since the burn was very localized.

I mean... Good work rewinding a motor, that's hard enough, great achievement, but I don't think you've done any magic here. And while you're using shunted square wave controllers there's still quite a way to go. Sin wave isn't "slow" as you've claimed, that's a ridiculous statement. Maybe you mean "I got a shit sin wave controller and it was slow". Sin wave can mean many things. You need DTC or FOC. Sure, dumbly drawing a sin wave is going to be not much better than square wave.
 
jamiejackherer said:
As I said though, it's would not be possible for you to estimate the copper fill based on the pics. I used an old perferated PCB sanded down to be smooth edged and hammered each turn in to the slot. It's packed in solid mate. There will obviously be some gaps as it the nature of round wire but I can assure you it is packed as tight as it possibly can be. Not one wire is able to be moved in the slots. No light penetrates through any slot either. I'd estimate the fill to be closer to 60 or 70%

Commercial motors are only around 65% fill, so if experienced builders are suggesting less, your estimate may indeed be optimistic.
 
Battery amps=Phase amps/√3*duty cycle.

Duty cycle is proportional to speed, going from nearly zero at stationary to 100% at max throttle.

If you pulled 110 battery amps at stationary you'd be forcing thousands of amps through the motor, with the 6kW going almost entirely into heat in the FETs and windings. Controllers primarily control phase amps, because that's what destroys MOSFETs and motors. Your one does too.

I don't know what I'm talking about. Hmmm. I've literally designed the boards and written the code for sin and square wave controllers. Guess it's blind luck that it works, I was basically typing random characters.

I've only rewound one motor, a little rc one. I'll defer to larsb and so on for their knowledge on this.

Let's have a video of this. Your destroying surrons with a 200$ motor on a pedal bike? With one of those rubbish little square wave boxes you can buy for like 20$ with a bit of solder on the shunts? This wants some evidence. It'll be revolutionary. You'll open a business with people posting motors from all over to get some magic Jamie windings. Hell, I'd send you one.

So why do you think one wire is completely black and charred and the rest look absolutely fine?
 
fatty said:
Commercial motors are only around 65% fill, so if experienced builders are suggesting less, your estimate may indeed be optimistic.

Actually numbers are a lot lower. Good machine wound rc motors are around 30% when i’ve measured them

65% is an exceptional motor, often hand wound
This is an example of a high fill motor, best ebike motor winding i’ve seen:
BFCB06AD-01A5-46DA-B47F-BE5A0F4BA50A.jpeg

Zero crossings, minimal lost space due to bundled wires

Compared to this one you see the difference:
070F1928-6965-49B1-AB6F-83FC934543A5.jpeg
 
Honestly looking forward to this video. You've got some big claims beating the surron. If it's as capable as that, I'll be buying some copper wire and glass tape.

What exact controller did you start with? The stock one from the voilamart kit? I stopped using that because the throttle delay was about a second. It does have a lot of MOSFETs though.

My code is available at
https://github.com/davidmolony/MESC_Firmware/blob/Add_ADC_inputs/Core/Src/MESCfoc.c

Help yourself. Free for all licence. No flashy interfaces etc like VESC, just takes an ADC or ppm input and maps it to a quadrature current on a 3 phase inverter.

No jealousy, I'm also riding around with a fully home built bike pushing far more than the 1500W boiler plate rating. I actually mainly ride one with a belt drive 80100 motor because it weighs so little, battery and motor at about 4kg overall and I get a (low) few kW.

Its just big claims (surron) need evidence, and it looks like you started out with a broken motor, which would have seriously under performed until you fixed it. My estimation of your improvement from the pics is that you've added about 50% extra copper, so can expect about 33% less heat generation over a properly working motor. This is good. Fantastic even, but it's not really indicative to say by doing this you've improved from 1500W to 6, 10...kW, you've probably gone from a 4-5kW capability to a 6kW capability, depending how long the period you measure is.
 
Just gathering some quotes here from user jamiejackherer:
Yeah, Ive been trying to learn as much as I can from video's and reading what I can.

Ive learned that I need a higher number of turns for greater torque, which will reduce the max RPM.
what length of wire shall I buy?
I have 2 jobs, 6 kids and I am studying a degree aswell... It's not a case of "can't be bothered". Its a case of literally not having the time.. I don't have transport until I can get this done
Im a professional programmer mate (python for fun, C for work) and I've been rewriting the VESC code so
I've been through about 50 motors in the past 5 years
I build and sell custom builds, about 1 a week atm

I find these more than a bit contradicting.
You build ebikes for a living but when one motor burns you’re desperately out of transport? Meanwhile you have used and burnt 50 motors in the last years

Wound two motors but didn’t have time to spend an hour reading basic motor knowledge?

Programmed VESC but cannot estimate a copper wire length?

I’m calling you: what’s the true story? Would be nice to hear what’s behind the lies.
 
You've got to understand, we're not being dicks. The problem is, this forum is read by literally tens of thousands of lurkers, and you can guarantee someone will try rewinding a voilamart when they see this. And if they've got a broken motor, doing exactly as you have is a great way to save $$$ and get a bit better performance.

They will be very disappointed when they don't get 4x the power handling, they get 1.3x.

You'll find there's a disproportionate amount of people here who really really know what they're doing compared to other ebike forums, from all over... There's zero motorbikes engineers, about 5 companies actively selling high end controllers, people who've built electric dragsters... I do this because I went into lockdown with an attitude that i wasn't going to fill it with Netflix, I was going to learn power electronics and learn to code in C... my coding is still pretty poor :lol:

Please, keep at this, just... Be a bit less angry, and cite like for like numbers.
 
jamiejackherer said:
And bare in mind, I'm running this with a 48V battery, as we should all know, lower volts and higher amps is the way to destroy any motor. If I was running higher volts, say a 60v or 72v and still the same wattage my amps would be lower. It's much harder to get 190 amps through 48v than it is to do the same at 60v..

This is an interesting topic and i think it is not true, at least in a general sense. We are looking at the motor side so there can be some different cases

1) power to the wheel is the same --> doesn't matter what input to controller is. Power to the wheel is torque*RPM and torque depends on phase current, if phase current is the same then motor is driven the same when it comes to losses and heat.

2) motor is stalled and fed same power --> phase current limiting is done, phase current is the same, motor is driven the same

3) Motor is driven but current limiting is not taking place. Controller output corresponds to throttle input. To be able to compare in one use case i guess requested acceleration must be the same --> phase current is the same, motor is driven the same.

Would like to hear what @mxlemming has to say about these lines, if you programmed controller drive routines then this must be a known for you.
 
jamiejackherer said:
Go frock yourself pal... Trying to call me out as a lier... You're a prick fella

It might seem that way but it's obvious that, due to some reason, you played us.
I notice you didn't respond to the question i had? Why?
 
Status
Not open for further replies.
Back
Top