How many Amps can this 3500W rated QS hub motor actually take? (build planning)

redmouse

10 W
Joined
Aug 3, 2017
Messages
65
EDIT: conclusion of this thread: https://endless-sphere.com/forums/viewtopic.php?f=12&t=105413


Hey folks, I need expert advice.

I plan to build a moped using this hubmotor. It's a QS motor with 3500W stamped on it, 13" rim, originally ran with 20S LiFePo (72V max) and what I assume was 50A controller. The donor bike is pretty large and heavy (guessing ~130 kg + 2 passengers)

The bike I'm building will be much smaller and lighter (80kg), it's actually using a pitbike frame, and only one passenger. It will mostly be used for city drivimg, but with steep hills. I will probably use Nissan Leaf cells, they can provide 200A easy.

What is maximum reasonable Amps I could run the motor on under these conditions, and still keep it reliable?

Picture:
PwRjggM.jpg

2AT7ydN.jpg
 
That version was produced with different windings (different Kv), and current handling varies directly with Kv, because resistance of the copper windings varies. Kv (rpm/volt) is a pretty good indicator, but the best method is to measure the phase to phase resistance, so you can computer heat created in the copper.

I have one of those motors in that fastest winding they did, and powering a 60kg ebike with at the time me weighing 130kg, I pushed my controller to 425A battery side and 550A phase. Of interest is that the motor ran hottest cruising on the highway at 100kph+ with my 20s pack. That's because back then they used .5mm thick stator laminations with a very high slot and pole count, which drove up iron core losses. To get a handle on that heat generated, simply measure no-load current at max rpm and multiply it by pack voltage.

In addition to quite extreme torque potential, a good thing about the motor despite it chewing through juice at speed is that the large and thick side covers with the raised design details that increase surface area help it dissipate a lot more heat than the common 220mm diameter hubbies with smooth side covers I ventilated mine with intake on one side with holes closer to center and exhaust holes at the very perimeter of the other side, which ended my issue of getting hot at high speed.

If you have the fast winding like mine (over 18rpm/volt), one concern is the risk to controllers due to very low motor inductance. The motor itself is a difficult load. I ended up ventilating my 36fet controller with fans to achieve solid reliability at the extreme power I was running.

Motor heat depends so much on how you ride, the total load, the terrain, and ambient temperatures that only experience can tell you the true limits.

If you do a light build, use a long wheelbase and a low CoG. There's no way you can push the potential of the motor with any of the typical ebike layouts seen here on the forum. With my custom swingarm on my Cannondale SuperV that extended the rear about 8" and gave me a low riding position, I could lift the front wheel while leaning my fat self out over the handlebars as far as I could.
 
425A, holy shit, I guess I don't have to worry about using 200A then :shock: .

I think the bike originally did 70km/h @ 72V, not sure what it says about windings, but since my battery is well below what you ran it on, I shoudl be fine. Thanks John!

How about controllers - this is a budget focused build (lookign for best performance for the buck). Can I use some controller from Aliexpress? Somethign like this?
https://www.aliexpress.com/store/product/Kelly-KLS7222N-30V-72V-220A-SEALED-SINUSOIDAL-BLDC-MOTOR-CONTROLLER-For-2000W-3000W-Motor/109978_32895907636.html?spm=2114.12010612.8148356.19.4f30286bWt0J3w

Or should I aim for a bigger one, to use full potential of my motor and battery (probably 9 Nissan Leaf cells (74V and 250A capability)
 
redmouse said:
425A, holy shit, I guess I don't have to worry about using 200A then :shock: .

I think the bike originally did 70km/h @ 72V, not sure what it says about windings, but since my battery is well below what you ran it on, I shoudl be fine. Thanks John!

How about controllers - this is a budget focused build (lookign for best performance for the buck). Can I use some controller from Aliexpress? Somethign like this?
https://www.aliexpress.com/store/product/Kelly-KLS7222N-30V-72V-220A-SEALED-SINUSOIDAL-BLDC-MOTOR-CONTROLLER-For-2000W-3000W-Motor/109978_32895907636.html?spm=2114.12010612.8148356.19.4f30286bWt0J3w

Or should I aim for a bigger one, to use full potential of my motor and battery (probably 9 Nissan Leaf cells (74V and 250A capability)

You can go bigger if you have the budget. I'd say the limit might be around 5-600A/phase or slightly higher. The 220A you plan to use seem a bit on the weak side to me.
I was pushing 400A/phase on my 2000W hub motor and it still works fine now, 4 or 5 years later. So if yours is a real 3500W there should be no problem.

Just make sure it is a V3 (version3), QS offers 3 different versions of their motors at a given power rating. V1 uses a steel stator plate, V2 uses a machined aluminum one and V3 uses the same stator plate as V2, but with beefier wires.
 
I think it's a good idea to take it apart to find out which version you have and also to check the internal condition. if there is any rust etc.
 
If it only did 70kph at 72, then it's a slower winding, more turns on the teeth, so longer thinner copper, which can't handle current as high as the one I have. You need those measurements and opening the motor won't tell you. Opening would only enable you to measure the stator width to verify it's the same 50mm wide, but that doesn't impact current handling. Measure phase-to-phase resistance, by putting a known low current through a pair of phases and measure the voltage drop. That's the ONLY way to predict current handling for ANY hubmotor. Cooling can help some, but not all that much, since usage, load, and terrain are far more important.

Those older versions of that motor like we have can't handle current like Dui is talking about unless it's the fast winding AND you upgrade the thickness of the phase wires coming out of the motor. The 273's were never even made with steel stator supports, and V1, V2, V3 info doesn't apply. Scooter manufacturers bought motors from QS based on whatever specs they ordered.
 
if you open the motor: replace the bearings and seals.
its a quick job if you have the motor open and it makes it more reliable.
get 2RSL bearings, those have low resistance seals.

i have a simmilair motor on the self as well, if you can share some pics of the inside and bearing sizes that would be great.
 
Alright, I cracked it open, here it is:

8Y9xb0V.jpg


L52bqOr.jpg


surprisingly there is very little rust (considering it was sitting outside in the rain for 3 years)
And I hope it wasn't overheated, otherwise the hot glue inside would be burned, right?

One thign that's mildly concerning is cracked insulation on phase wires. I suppose it's from old age? So it looks like I'm buying new ones either way.
8uxhNXo.jpg


The current one have 3mm diameter (the copper part). I suppose if I'm gonna be running 200A+ I should be thicker, but how much thicker? 4mm is pushing it? The one right now came through the axle very easy.
 
Alright, so it turns out you can actually stuff 3 additional wires into the axle. I reused good sections of cable from controller and rest of the cables. So now each phase would use two parallel wires, hopefully effectively halving the resistance. From what I can tell the wires are probably 8AWG, so doubling them would make them ~5AWG?

Any reason doublign cables won't work? I know it's a bit redneck solution, but I live at the end of the world and specialty stuff like teflon coated thick gauge wires is really hard to get here.

aFky5h8.jpg
 
No reason it won't work; it's been done before, when a thicker single wire can't fit thru some space it has to go thru.

If you need even more conductor, you could replace the wires with "magnet wire", so the insulation would be very thin. But it should have a secondary insulation layer, like kapton, wherever it passes up against metal like the axle, especially if vibration/etc could cause rubbign over time.

I did this for my MXUS 3k 4503, and it worked fine for the few days until the axle broke again and cut into stuff. :/ (I fixed the axle but didn't have time for rewiring that night, and haven't been able to get back to it since, so the wheel is still unpowered ATM).
 
Thanks Amberwolf!

I'm currently trying to solder the wires, they are very tough, solder doesn't want to stick at all. I'll try some other solder and fux tomorrow.
Maybe cleaning them with acid would help?
 
Good Job!
It's not visible at the photos how you try to solder them so I suggest you to solder the two pair of wires straight away on the winding.
When I have similar soldering issues I use high power iron and lot of good flux paste.
 
Thanks! I tried a regular soldering iron at first, but it was clearly not enough, so I tried a heat gun. Covered everything around so I don't do any damage, and it heats it up nicely. Except the solder doesn't stick.. maybe it's the extra oxidation from airflow? I'll keep practicing on some wires aside, and worst case I'll go borrow some massive iron.

I do flux the hell out of it, maybe I'll try different brand.
 
Can you show some pics of the conductors of the wire you're using? It might help us understand why you're having the problem with solder. As long as the wire is copper, whether it is tinned or bare, it should easily solder with flux and any solder that's meant for copper.

If it's not copper, then it depends on what it is which solder / flux would work. If it's a joint between copper to some other metal, that may not work at all without a crimp lug.

If it's magnet wire, then you have to remove the coating first; I just scraped it off with a file for the 10g wire (it's really stuff so easy to do that way), but it can be burned off or various chemicals will also work (I think APL's thread on his motor has some info, there's a few others around ES and elsewhere that give other ways).


FWIW, with large gauge wire, it may take considerable heat to bring it to the melting point of regular tin-lead solder. If it is lead-free solder it may take even more heat.

What size is your soldering iron? Not just the wattage, but the tip size?

It usually takes a large-mass chisel-tip iron (like the size of a couple of joints of your finger) to have enough heat-containment available to quickly heat up a thick wire fast enough to be able to melt solder before the wire conducts too much of the heat away. The iron doesn't have to be high-wattage, but it helps reheat the tip quicker each time you use it. (I use an 80w Weller with a large chisel-tip, and it will do 8g easily, probably much larger; I just use regular 60/40 rosin-core solder).
 
Thank you Amberwolf for being so helpful! And others too! I have been visiting ES for some time, and it's amazing how much wisdom you guys tirelessly share!

Anyways, I managed to solder it with a butane torch and different brand of solder (probably higher quality?) I also crimped it with copper, to make it more failsafe.
 
So in the end I went with 350A peak, 100A continuous controller. It wans't that much more money, and I can always detune it if it turns out too much.


How about power wires. I got 4AWG, do you guys think that's enough, or should I go thicker? They will be pretty short, 40 cm from the battery and 50 cm phase wires. (probably shorter)
 
4awg is fine. Just take care of the connections. They also have to be able to handle those currents.
 
While you have it open put some kind of temperature sensor in there, even if it is just one of the cheap stand along eBay modules with cable sensor.

That way there is no guessing about temperature
 
I hope you replaced those original phase wires. That cracked insulation could quickly lead to burning up a controller, and the crack on that blue wire makes all of the original suspect. It's always best to do it all right the first time. There's no better transportation that an absolutely reliable EV, and there's few things more frustrating that breakdowns with a vehicle that is relatively easy to make 100% reliable.
 
Thanks guys, I will use 4AWG then. Will be checking on them, if they get hot, I'll double them or upgrade.
Yeah, I'm pretty anxious about making good connection. I'm thiking I should get myself a crimping tool, or perhaps rig up one with a vise or something.


bamitsram: yep, already done! Just the cheap one, that's what I had. It can go only up to 100 celsius, but I hope the motor is big enough to not even get there.


John in CR: I replaced the bad ones with good sections from rest of the cables. I'm keeping only first 20cm (from the winding to outside of the axle). I gave them pretty tough test by stuffing them into the axle, and the insulation is holding up great, so I'm not too worried. They are also siliconed around the axle and additionally insulated with fiberglass tube and kapton.
 
Good day fellas, I have another question:
I got this BMS from ICGOGOGO, that's rated for 400A (peak I assume)
Do you think it's ok to pull 350A 30S peak through it? I have some doubts about the traces, it seems like they are too thin. Would it be ok in your experience, or is it worth soldering on some reinforcements?

Pictures here:
ICGoGoGo 400A BMS https://imgur.com/a/DZ9X4tx
 
Back
Top