Rewinding 4kW scooter motor options

[ctirad]

Hello,

I have HUB motor that needs to be rewinded, because of shortcut to the stator body. I'd like to make it faster, because the current Kv allows the speed of 75km/h at 72V, but I'd like to have at least 90km/h. However, after studying some motor related materials, I'm not sure if it is possible at all.

The current parameters are:
Number of magnets: 56
Number of teeths: 63
winds per teeth (estimated): 2
winding pattern (estimated): AaABbBCcC
termination: WYE

Are there any possibilities to increase the Kv other than to make single wind per tooth (unusable high Kv and too low torque) or change termination to delta (generally discouraged due to worse efficiency)?

 

[flexy]

I am interested in doing the same.
I saw this on the Visforvoltage forum from M.Chen.

http://visforvoltage.org/forum/11326-erider-8000w-scooter-australia-similar-xm5000li?page=25

They appear to be using very thick copper with single wind. It looks quite challenging to get the wire fitted, it might be easier to enquire about the motor price and order one.

 

[John in CR]

What is your current no-load rpm per volt? I have a similar motor with a 50mm wide stator and same number of teeth, which has a Kv of about 17rpm/volt. When I looked at the copper I thought it was 1.5 turns per tooth. It is crazy hard on controllers due to the extremely low inductance, so I'd suggest staying with your current winding and raise the voltage for more speed.

 

[John in CR]

What is your current no-load rpm per volt? I have a similar motor with a 50mm wide stator and same number of teeth, which has a Kv of about 17rpm/volt. When I looked at the copper I thought it was 1.5 turns per tooth. It is crazy hard on controllers due to the extremely low inductance, so I'd suggest staying with your current winding and raise the voltage for more speed.

 

[ctirad]

I cannot raise the voltage much above 25 or 26s lifepo4 (100V components everywhere) not mention the safety issues.
I'll try to unwind a couple of theeth to explore the exact number of the turns.

 

[Thud]

you could add a turn & terminate in Delta.

 

[ctirad]

Yes, to add a (half)turn and delta termination is the plan B.

However after some more closer observation it seems there are 2.5 turns per teeth.

 

[ctirad]

The correct pattern looks like this, after all. Any ideas?

 

[fechter]

That winding arrangement is pretty common. Concentrated pole configuration. It reduces the amount of wire going from pole to pole, and therefore reduces the losses. That is a huge motor. I've rewound a few motors before. It's very important to carefully map out how the windings are arranged before taking them off. Carefully count the number of turns.

The number of strands in your winding is not that critical. More strands of smaller wire are generally better, but it won't make a huge performance difference. If you want to change the number of turns, it gets a little more tricky as you want to try to stuff the slots nearly full with copper and you need to calculate how what gauge of wire will fill the slots with a given number of strands, etc. Exactly matching the original is easier.

 

[Shenta]

The current parameters are:
Number of magnets: 56
Number of teeths: 63
winds per teeth (estimated): 2
winding pattern (estimated): AaABbBCcC
termination: WYE

OK, let's start with somewhere far far away... in China, that's what they call 13" Hub Motor, from the picture you showed, it is rather a 3Kw motor, and it is very early production, which has no rotor balancing, so over 60Km is dangerous.
the motor winding is actually 7sets of 8:9 motor.... 8pole x 9slot x 7set. so you could either serial them together or parallel them into 7 sets. My suggestion is 14 turns, AaABbBCcC winind should give you app 60Km or little bit faster. Do a good calculation on how many copper wire you could put into it. Good luck.

 

[John in CR]

The current parameters are:
Number of magnets: 56
Number of teeths: 63
winds per teeth (estimated): 2
winding pattern (estimated): AaABbBCcC
termination: WYE

OK, let's start with somewhere far far away... in China, that's what they call 13" Hub Motor, from the picture you showed, it is rather a 3Kw motor, and it is very early production, which has no rotor balancing, so over 60Km is dangerous.
the motor winding is actually 7sets of 8:9 motor.... 8pole x 9slot x 7set. so you could either serial them together or parallel them into 7 sets. My suggestion is 14 turns, AaABbBCcC winind should give you app 60Km or little bit faster. Do a good calculation on how many copper wire you could put into it. Good luck.

This will not work, but even if it spun up the motor it would be horribly inefficient.

CTIRAD,
You still haven't mentioned your no-load rpm per volt. I have experience with these types of motors, but without the Kv (rpm/volt) I really can't give you any guidance other than to say unless it's just for the learning experience leave the rewinding of a motor like than to the pros who do many a day.

 

[Shenta]

what u mean it will not work? It work better than the original wiring. And now motor manufacture is using this wiring now.

 

[ctirad]

CTIRAD,
You still haven't mentioned your no-load rpm per volt. I have experience with these types of motors, but without the Kv (rpm/volt) I really can't give you any guidance other than to say unless it's just for the learning experience leave the rewinding of a motor like than to the pros who do many a day.

I don't know how to find out the rpm per volt constant for a non functional motor The top speed was around 75km/h at 72v nominal, and it's 13" with 130/60 tire, but it gives only very rough estimation.

 

[fechter]

I don't know how to find out the rpm per volt constant for a non functional motor The top speed was around 75km/h at 72v nominal, and it's 13" with 130/60 tire, but it gives only very rough estimation.

Actually that's probably more useful than knowing the actual kV. If you change the wind, you can get a pretty good estimate based on the original and the ratio of new/old.

 

[ctirad]

Exactly. That is what I'm asking from the very beginning. I'm not going to make any complex electromagnetic analysis or simulation, I just want to fix my motor and make it 20-30% faster if possible. All the parameters are now known. The winding pattern including the correct number turns is on the drawing above. So please, do you have any ideas how to solve it?

 

[fechter]

You're kind of constrained to an integer number of turns. One turn would be too fast. Another possibility might be using the same 2 turns, but terminate it delta instead of wye.