Car cooling fan bldc motor

Hello

I am Ruudi from Estonia and I am very interested in electric vehicles. This is the area where both of my main interests come together electricity and vehicles.
This is my first post in this forum. I have been reading a lot of your interesting topics here in this forum. But have not got answers for my questions yet.

I have Mercedes Benz engine cooling fan motor witch is BLDC outrunner with built in hall sensors. It is rated 850W and 13V. It has got 7 magnets on the rotor and 12 pole stator. It has usual ferrite magnets. Interesting is that this motor does not have any cogging at all if you rotate it by hand. What do you think is it electrically and magnetically a motor enough for big heavy 50cc scooter conversion?
On table with no load it was promising. I had kelly minicontroller connected to it and it worked fine at 48V 4,5A (from batteries) got 9200rpm even regen was working.
I have mounted it onto the scooter motor three different ways and it dissapointed me. Acceleration is wery poor top speed was 29 km/h. Tried everything: modified variator, locked it up, tried different belt drives. I think that this kelly mini controller is maybe too weak but it is rated 35A and my ammeter does not show over 20A I do not know why.
Anyway what do you think, what should I do to get more power out of this motor. Should I change the magnets to neodymium? Is it sensible to change the number of magnet poles to 8? More powerful controller from recommended e-crazyman is on its way

Ruudi said:

What do you think is it electrically and magnetically a motor enough for big heavy 50cc scooter conversion?

Click to expand...

I doubt it...

850W is a little over 1hp. A 50cc scooter could be ~3hp up to >6hp.

This is interesting, if this motor is truly good for 850 watts at 12 volts. I would think 3400 watts at 48 volts, with gearing (volt up and gear down theory) could work.

Would like to know the final gear ratios and the tire size.

learning by the seat of my pants.

Ruudi, what is the diameter and thickness of the motor, what is the diameter of the shaft, and what model of car/engine does this motor come from? Thanks in advance!

thanks for reply

motor diameter:127mm
overall thickness:66mm
shaft diameter:10mm

Shaft is very short because the fan is attached directly on the rotating bell. It can be found on mercedes benz C class (w203 body) cars and s-class (w220 body) I believe other models too and they have different power ratings some are 600w some 850W. They are with controller witch is the problematic part that is why I got these motors.

I suspect that the motor may not be able to produce enough torque to drive your scooter, despite the reasonable amount of current you can drive it with. You may well be able to increase the motor torque if you change the magnets. I'm not sure if it will make a massive difference, but my guess would be that it might well be significant, it depends on a number of other aspects of the motor's design.

The problem is going to be finding neo magnets that will fit. I'd suggest that you may well have to increase the number of pole pairs to 7 (14 magnets). This will reduce the rpm and may also reduce the hysteresis losses, as the motor is spinning a bit too fast for good efficiency at the moment, I think.

Keep in mind that it's able to do it's 850W while lots and lots of cooling air is being pulled past it. It may not be able to do anything close to that without that cooling. I doubt it can do more than that without even more cooling, and probably not enough to run that big a bike.

It is also going to be very high RPM, and if you do not let it run at that kind of RPM under load, it will probably get very hot very fast, cooling airflow or not. If you can measure the RPM it runs at with the fan blades attached, while running from a car battery, that would be best.

You'll probably have to use a belt or chain reduction from the motor to the wheel.

If you kept it at automotive voltages (13-14VDC) the high RPM and stayed within the 850W rating (which is probably peak, not continuous, at a guess), you could use the built-in controller. Otherwise you'll probably have to disconnect that and use an external controller.

Jeremy Harris said:

The problem is going to be finding neo magnets that will fit. I'd suggest that you may well have to increase the number of pole pairs to 7 (14 magnets). This will reduce the rpm and may also reduce the hysteresis losses, as the motor is spinning a bit too fast for good efficiency at the moment, I think.

Click to expand...

maybe these neo magnets would be good, motor calculators say that http://www.magnet-shop.com/Neodymiu...x-100-x-50-mm-N46SH-Nickel-150-%B0C::570.html will fit.
But in my scooter i need very high rpm, it has rear reducer ratio of 1:13 and I have 12" rear wheel. I would like to get top speed of 50km/h at that speed motor should rotate 7500rpm under load. There are a few topics where people have built these scooter with rc outrunners witch are fast and powerful enough but there is big controller problem. Maybe 72v 1500w ebike controller can drive this motor.

amberwolf said:

Keep in mind that it's able to do it's 850W while lots and lots of cooling air is being pulled past it. It may not be able to do anything close to that without that cooling. I doubt it can do more than that without even more cooling, and probably not enough to run that big a bike.

It is also going to be very high RPM, and if you do not let it run at that kind of RPM under load, it will probably get very hot very fast, cooling airflow or not. If you can measure the RPM it runs at with the fan blades attached, while running from a car battery, that would be best.

You'll probably have to use a belt or chain reduction from the motor to the wheel.

If you kept it at automotive voltages (13-14VDC) the high RPM and stayed within the 850W rating (which is probably peak, not continuous, at a guess), you could use the built-in controller. Otherwise you'll probably have to disconnect that and use an external controller.

Click to expand...

Cooling is not main problem right now if needed the scooter has water cooling radiator in front. And in the car this motor works in very hard conditions temperatures near 100 C degrees. The big fan witch I don't have is around the motor http://en.bildelsbasen.se/?link=item&post_id=23617929 I don't know if it gets very big airflow like that. But as you can see from the pictures separate motor controller does get big air flow but that is not enough and they stop working that's why I have these motors.
Very high rpm is good I need 7500rpm witch it doesn't do on the car with 12V but 35V gives 7300rpm without load. My scooter rear reducer has ratio of 1:13
I don't want to use original controller because it is built for 14V (and these witch I have do not work anymore) although they are very smart (comparable to your ebike controllers). If the controller detects bigger load then a fan at startup it doesn't start at all so it cannot be used for other purposes. There is current limiting and fan motor health control program inbuilt and it regulates fan airflow depending of car motor and air conditioning needs. And it gives feedback to car engine computer about fan rotation if it's stuck or not working it put's on check engine light and reduces internal computer engine power so it doesn't overheat and driver can drive to mercedes workshop slowly.
It should be continuous power this 850W because these big powerful cars with all electric and other equipment need very much cooling power on a hot day. And on my scooter the motor doesn't have to work at that high temperatures.
Maybe I have to put two of those motors on the bike maybe three?

The big problem with running this motor at more than 4 times its design rpm is going to be the one I've mentioned already, the increase in hysteresis loss in the stator. I strongly suspect that the stator of this motor will have fairly thick laminations, and this will result in high losses at such a high rpm. Thin laminations are expensive, and I am certain that a motor manufacturer, looking for low manufacturing cost, wouldn't use laminations that are any thinner than needed for the job.

Zibby at otherpower.com (RIP) had some luck replacing magnets like this with rows of cube or button-shaped neodymium magnets. But, I suspect if you went this route, you'd be better off just spending a little more and buying something that you know will work.

If you have money to burn and the time and tools, it may be "fun" to see what this would do at 24V...perhaps even 36V.

TylerDurden said:

Ruudi said:

What do you think is it electrically and magnetically a motor enough for big heavy 50cc scooter conversion?

Click to expand...

I doubt it...

850W is a little over 1hp. A 50cc scooter could be ~3hp up to >6hp.

Click to expand...

Yes, but an 850w motor might knock out more torque, pound for pound, and the weight savings could make up the gap considerably.

4.5A no-load @ 48V, does not sound too good to me.

spinningmagnets said:

Zibby at otherpower.com (RIP) had some luck replacing magnets like this with rows of cube or button-shaped neodymium magnets. But, I suspect if you went this route, you'd be better off just spending a little more and buying something that you know will work.

If you have money to burn and the time and tools, it may be "fun" to see what this would do at 24V...perhaps even 36V.

Click to expand...

Thank you for warning me, I really haven't got money to waste that's why I turned to you all, who are more experienced and can tell me right away what to do and what not.

What would be the right motor then, and where to get it? It should make 7500rpm under load.

Jeremy Harris said:

The big problem with running this motor at more than 4 times its design rpm is going to be the one I've mentioned already, the increase in hysteresis loss in the stator. I strongly suspect that the stator of this motor will have fairly thick laminations, and this will result in high losses at such a high rpm. Thin laminations are expensive, and I am certain that a motor manufacturer, looking for low manufacturing cost, wouldn't use laminations that are any thinner than needed for the job.

Click to expand...

Laminations are 0,65mm is it good or not?

Yesterday arrived 72v 1500w ebike controller. I connected it and it worked great motor spun over 9000rpm when rear wheel was in the air. Cannot tell you the current because my ammeter didn't work anymore. Battery voltage was 50V.
Then I rode with scooter and it went almost 40km/h. First test ride yesterday I had the variator working and I didn't like it because motor speed went up quickly and then slowly the speed started rising.
Second test ride I blocked the variator to 1:1 ratio and rode a few kilometers and then it stopped. Controller overheated, motor was cold. Then I pushed the scooter 2km. Do these controllers have overheating protection? Now I know that the problem was the very small kelly minicontroller witch isn't meant to give that much power.
I think that this motor has the potential maybe I put two of these in parallel with toothed belt. Then rebuild the controller with new mosfets and the belt drive system.

Ruudi said:

Laminations are 0,65mm is it good or not?

Click to expand...

Not very good at all, I'm afraid. Eddy current losses (hysteresis losses) in the stator are proportional to the square of frequency, so doubling the motor rpm means four times the loss. The motor was designed to spin at a much lower speed than you are now running it, so the manufacturer used those thick laminations. Motors designed to run at that high an rpm would typically have laminations that are around 0.2mm thick, or perhaps less.

Ruudi said:

Yesterday arrived 72v 1500w ebike controller. I connected it and it worked great motor spun over 9000rpm when rear wheel was in the air. Cannot tell you the current because my ammeter didn't work anymore. Battery voltage was 50V.
Then I rode with scooter and it went almost 40km/h. First test ride yesterday I had the variator working and I didn't like it because motor speed went up quickly and then slowly the speed started rising.
Second test ride I blocked the variator to 1:1 ratio and rode a few kilometers and then it stopped. Controller overheated, motor was cold. Then I pushed the scooter 2km. Do these controllers have overheating protection? Now I know that the problem was the very small kelly minicontroller witch isn't meant to give that much power.
I think that this motor has the potential maybe I put two of these in parallel with toothed belt. Then rebuild the controller with new mosfets and the belt drive system.

Click to expand...

I have that mini kelly, not in use as of yet. Have to mention the standard mini kelly is rated at 40000 erpm. 9000 rpm x 6 = 54000 erpm, you have exceded it.
something else to consider.

I dismantled one motor witch hall sensors failed to see how it's coils are connected. It seems to me that it is connected in wye and one phase has 4 coils witch are in parallel. What if I connect them in series? Do I get a 48V motor then?
https://skydrive.live.com/redir?resid=24AB5606EA901DFB!526

I draw the motor circuit diagram:

Interesting fan motor Ruudi.
I have a question: Is there a built-in BLDC controller? If so, can you post a picture of it?

No controller is not built into the motor but is one replacement part with controller with connection wires. If motor controller fails then you need to replace motor and controller together they are not separately available from dealers. Pictures of controller are possible, but later.
Controller is made very high quality and waterproof. But unfortunately it is meant to be a fan motor driver and that is why it assumes low motor starting current and if it detects high load it cancels fan motor startup and gives error signal to car engine control unit. And since it is made to work at 14V they use six low voltage mosfets 35V if I remember right.

I used kelly minicontroller and "ecrazyman" 72V 1500W 12 fet controller to drive it and it worked fine with both.

I see. Thanks for the explanation.

This motor might work well at 24 v on a bicycle.
Your project reminds me of my attempt to convert a Honda 80 scooter to electric using a relatively small motor. It never worked well but it lead me to make my first Ebike several years ago. For the Ebike I used two brushed Bosch radiator fans with a friction roller that drive the bicycle tire. The motors came from Mercedes cars in the junk yard. They worked well to drive a bicycle.
For the scooter an axial flux motor should work even with the variator locked. Something like a perm brushed golden brushless motor, mars brushless motor or even the Chinese brushed knock offs called eteck that are not really axial flux.

That is an interesting motor. Might be interesting to play with.

mr.electric said:

This motor might work well at 24 v on a bicycle.
Your project reminds me of my attempt to convert a Honda 80 scooter to electric using a relatively small motor. It never worked well but it lead me to make my first Ebike several years ago. For the Ebike I used two brushed Bosch radiator fans with a friction roller that drive the bicycle tire. The motors came from Mercedes cars in the junk yard. They worked well to drive a bicycle.
For the scooter an axial flux motor should work even with the variator locked. Something like a perm brushed golden brushless motor, mars brushless motor or even the Chinese brushed knock offs called eteck that are not really axial flux.

Click to expand...

I have this motor installed to my scooter and I have been riding with it. Voltage was 48V. Acceleration and hill climbing was poor but top speed was 45km/h (28mph) like it is allowed here in Estonia. Feels like one thiskind of motor is not enough maybe two or more are needed. But the question is how to put them together so that it would be good 48V motor?