What causes a motor to get hot?

[auraslip]

I finally re-installed the temp sensor in my motor to find that it gets VERY HOT at what I would consider cruising speeds. At around 1kw it gets up to 95c! This is odd because the ebike.ca simulator says it should NEVER overheat at this power level.

I'm using:

HS3540 (ventilated)
lyen 12 fet
22s10ah lipo pack
throttle limited to 55% for crusing
temp sensor is a lm35 using the hobby king VT

I don't think the new temp sensor is malfunctioning as the motor case does get very hot. Much more so than my 9c 9*7 motor running at 1000w @30mph.
I don't think the motor is malfunctioning as it plugs along at the correct speed (1000w = slight less than 30mph)

It just gets very hot. Even ventilated.

Hint: the ambient temp in texas is finally below 100f today!

[Chalo]

P=I^2R. That's roughly the amount of heat you have to dump.

Heat transfer to the environment is proportional to hot surface area in contact with ambient air, times the difference in temperature between hot thing and adjacent air. To lower operating temperature, get lower I^2R, more heat transfer surface area, colder air, or more air movement.

Chalo

[e-beach]

I finally re-installed the temp sensor in my motor to find that it gets VERY HOT at what I would consider cruising speeds. At around 1kw it gets up to 95c!

I didn't know it got centigrade anywhere in Texas USA.
I guess the 95c is about 205 degrees Fahrenheit anywhere else around Texas, and that sounds warm to me.

When you say ventilated, how is it ventilated? Do you have a picture?
Are you putting the motor under stress...like a lot of hills?
Is your wiring heavy enough to carry loads without getting hot?
Is there corrosion inside your hub?
Is the hot sun just plane shinning on your motor?

I guess I am asking for a little more info.

[neptronix]

100f outside.. there it is. The lower the differential between ambient temperature and the motor itself, the worse cooling you get. Certainly in the winter time, you will see a completely different picture.

All motors will create some heat; that's just resistance from copper turning it into a heating element. A hub motor can shed a certain amount of heat constantly, and often there is a point where the heat sort of equalizes rather than continuously building up to a crescendo of motor failure.. but your temperature does that no favors.. and hub motor spinning at only a few hundred RPMs is going to be one of the least effective heat discarders of all electric motors.

I don't think ventilating will help much in this instance, maybe out in the cold it will, but not here.

Putting that motor in a smaller wheel would help. But ultimately i would either oversize or ditch hub motors entirely in your situation :/

[auraslip]

P=I^2R. That's roughly the amount of heat you have to dump.

Only 10a at these conditions because it's mostly flatlands.

When you say ventilated, how is it ventilated? Do you have a picture?
Are you putting the motor under stress...like a lot of hills?
Is your wiring heavy enough to carry loads without getting hot?
Is there corrosion inside your hub?
Is the hot sun just plane shinning on your motor?

Ventilated as per the guide in my sig.
Mostly flats.
12awg wiring inside the motor. 8awg outside.
New motor.
Gets this hot on a cool night (80f)

I'm baffled because my old 9c 9*7 didn't get this warm, so I'm wondering if the higher turn count has something to do with it, but that doesn't make sense because don't higher turn counts have lower resistance? Maybe it's the hard dozen or so take offs on the 8 mile commute that does it?

Perhaps enclosing the lm35 in epoxy is the issue?

[amberwolf]

What's your phase currents in the controller vs the old setup?

I'm using an 18FET "80A" that Methods beefed up (and I think reprogrammed for higher phase current ratio) as a prototype, and even if I keep the watts/battery amps down to the same as with the old 12FET 40A by being gentle on the throttle, it gets the 9c 2807 (ventilated) way hotter than I would like; hot enough I can't touch the windings and hot enough I can smell the overheated phase wire insulation/jacket plastic after almost any ride.


With the 12FET, even at WOT, the same motor on the same battery pack and bike didn't get very hot unless I abused it pretty seriously, except on really hot days and when i would have to sit in traffic with hot car engines running all around me, on hot asphalt where the ambient air would be a lot hotter than normal.


FWIW, the heat problem appears to be the same even if I'm just cruising along for long distances (3 miles) without stopping--a trip I took today let me ride that far with no stops thanks to green lights just as I got there, out of the 12 miles total, and the motor was still too hot to touch it's windings when I stopped and waited at the next red light (whcih takes forever at night to change). Ambient was probably 105-110F or more at motor height above the still-hot asphalt, at a guess; ambient when I got home 15 minutes later was still 102F at about 930pm, 4 feet off the ground near a wall of the house that is always in shade.



So I'm mostly convinced (prior to experimentation to prove it) that it has something to do with phase current settings, although I don't really understand it exactly, so that even at lower throttle settings (and thus lower average motor voltage), and the same battery watt and amp levels, teh phase currents experienced by the motor are much higher at least during high loads, and higher for longer than they would be with the other controller or with different settings on this one.


I need to dig out the programming cable again (misplaced it in my house shuffling recently), move the computer to wehre it can reach teh bike, and try different phase current and other settings in there to see how it behaves after that.

[dogman]

Faster wind motor. That's it. That's the primary reason I kept buying my next motor in slower and slower windings.

And the heat this year is not helping it shed the heat it makes.

[Russell]

An interesting experiment would be to change a couple of parts, if you have them of course, and see if there is any effect on temperature.

First, change the battery to 12S and run at 100% throttle (which should yield about the same 30mph)

Second, swap out the controller and again run on 12S

-R

[e-beach]

don't higher turn counts have lower resistance?

Turn count can actually add to the resistance due to the extra length of wire the electricity has to run through.

Given the same amount of current:

If the higher turn count is made with the same gauge or thinner wire than the lower turn count, then, the extra amount of wire will cause the resistance to go up.

If the higher turn count is made with thicker gauge wire then the resistance will go down due to the thicker area for the electrons to pass through.

Kind of like a two lane highway getting stuffed into a 1 lane tunnel versus a 2 lane highway going into a 3 lane tunnel.

[dogman]

Your 30 mph is slower relative to your no load speed. That means more heat than a slower wind that would be running faster relative to your no load speed.

On that motor, in 26" wheel, you are already beginning to stall the motor, even at 30 mph. S why everybody keeps putting those fast winds in small rims. My love of a big rim has led me to love slower motors.

Plus, it's just a bigger motor. So I bet you pull more amps starting up and such. so more heat there too.

Then the weather. It's why I yank the 3000w controller off my dirtbike for the summer. Heat just doesn't flow into 100F dry air good at all.

It's possible you have your wiring wrong, but I doubt you'd be doing that without noticing it, nor does it sound like it. You just went from a motor with a lower equilibrium temp at 30 mph, to one with a higher equilibrium temp, combined with the weather. More motor so it would tend to produce more heat getting to 30 mph.

[Miles]

that doesn't make sense because don't higher turn counts have lower resistance?

No. Resistance goes up as the square of the turn count (assuming equal fill factor). This is balanced by the increase in the torque constant and the fact that losses go up at the square of the current. So, you end up with the same amount of heat for a given torque.

[Kingfish]

Are we forgetting that magnets and electric steel become hot because of Induction Heating by way of Eddy Currents? Ironless cores eliminate ½ of the heat – leaving copper resistance (or bad winding design) as the source for heat. However with more traditional DD motors, there are two distinct causes for heat.

What we have is:

• Hot Day & 100*F differential between environment and motor
• Motor heating due to copper resistance
• Motor heating due to high frequency switching – electric steel and magnets and eddy currents

You can't escape the environment or the eddy currents, but - short of replacing the motor I agree that increasing the wire gauge from the controller would help.

Although easiest solution is to slow down a little bit; you know… stop and sip the lemonade

In the shade, KF

[Russell]

The HS3540 RPM/V is almost identical to a 9C 2806 so it's not an especially "high speed" wind however Auraslip is running it with a rather high supply voltage (22S). Because of this at the 30 mph cruise speed (55% throttle setting) the controller is doing a lot of PWM (phase current is about twice as high as the battery current per ebikes.ca simulator). It's for this reason I suggested the first thing he do is run the motor on 12S which will hit about the same 30 mph but at 100% throttle at which point the very little PWM will be done by the controller.

-R

[neptronix]

Yeah; from what i've seen going off of the ebikes.ca simulator, partial throttle seems to be less efficient than full throttle at the correct voltage.

[auraslip]

Now that I think of it, I never really put a temp sensor in my old 9*7 (running them same setup) during the hot part of the year. Only in the fall and winter. So that may be a big issue. Temp isn't 100f. More like 105f. And 6" off the tarmac who knows how freaking hot it is....

Another thing I noticed is that the motor ventilation doesn't seem to produce as much air movement. In my old motor it seemed to actually spit out air and suck it in. I could hold my hand to the running motor and go, "yup, the centrifugal fan is working." This one not so much. Maybe the fact that the covers are NOT vaned and much closer to the windings has something to do with it? The 9c motors were quite cavernous on one side, and had vanes...

I'd love to try 12s, but I'd have to completely disassemble the bike. I expected the controller to get warmed running at partial throttle (it does! Like 60c!), but I didn't think the PWM would cause the motor to heat up.

[John in CR]

To make sure it's not a motor or connections problem:
No load speed?
No load current?
Power consumption with mostly cruising at that 55% throttle?

Phase/battery current settings of the controller?
Are the covers getting hot too, which would indicate a lack of flow?

FWIW, the difference in ambient in C° should roughly equal the difference in motor temp in C° all other conditions equal, so something is wrong....maybe just the partial throttle, but I doubt it, not at steady state cruise.

You didn't accidentally get your exhaust holes angled forward did you? When spinning it up you should be able to feel air coming out of the exhaust, but be careful not to get your fingers or hand tangled up in anything feeling for flow. Note that I've become pretty confident that the inside blades play a big part in the success I've had with ventilation. Based on El Steak's results I don't think it's so much that they increase flow, but more that they deflect air flow at the stator that will otherwise naturally follow the path of least resistance (along the surface of the smooth covers to the exhaust vents). The covers don't need cooling, the stator does, and increasing turbulence at the stator very significantly increases the coefficient of convective heat transfer. The blades are really the only thing that explains such a cool motor at the top of a high power climb like I did with my son. On Hubmonster my interior blades pass within a couple of millimeters of the stator. In fact, I had to pull the cover twice when finalizing the vent mod, because blades were just barely touching the stator and I had to bend them flatter. Don't think vanes, instead think a blade shape like a regular axial fan to deflect air away from the covers. My motor gets thousands of little puffs of air from the blades per minute all around the end windings. If space is too limited, the blades can be at a slightly lesser radius than the stator like I did with my the motor on my son's bike.

[veloman]

I think ventilating would still help a LOT, even if the ambient temperature is 105 F. That's what, 40 C? So you have a 60 C temperature difference between danger zone and ambient. That's still enough to cool the windings I"m sure.

My Mac just saturates horribly in this hot weather. Long rides are impossible to keep it under 90 C without stopping. I tried doing 45miles one morning, and on the way back I was pedaling moderately in spandex (carrying a bike and into a headwind too), going about 20mph and ambient was around 85 degrees. Motor just kept climbing into the mid 80s so I stopped to let it cool. I was averaging at most 300watts battery power. Pathetic. The Mac is a 200 watt motor IMO.


I really want to see some improvement in the way ebike motors handle hot weather, even at reasonable speeds. I'm not pumping multi kw into it at all. What a shame..... good thing most of my riding is only 15-30 minutes at a time.....

My controller gets pretty hot too. I can smell it when I bring it in the house, if I take a sniff right near it.

[TylerDurden]

I really want to see some improvement in the way ebike motors handle hot weather, even at reasonable speeds. I'm not pumping multi kw into it at all. What a shame..... good thing most of my riding is only 15-30 minutes at a time.....

My controller gets pretty hot too. I can smell it when I bring it in the house, if I take a sniff right near it.

Not uncommon for the Canuks among us to de-rate their batteries for cold weather...

Y'all down there have the opposite problem and likely de-rate other types of equipment for high ambient temps - from PV modules to industrial motors, eh?

[John in CR]

In my previous post, the ambient temperature of concern is the air temp the motor sees, which I'm sure does vary quite considerably on a hot Texas day about a foot above the asphalt. A quick test for an overall system problem would be to get out and do the same ride just after sunrise. If it gets hot, then you've got a problem that needs to be addressed before any more far hotter daytime riding, though I'd first try some different speeds on that morning ride to see if faster is cooler. My latest big hubbie really loves to fly, and gets warmer with around town errand rides than it does at WOT up moderate grades.

I thought you already knew that a smaller wheel is better. What are you doing with that 26"?

[auraslip]

Went with the tried and true 24" - originally I wanted 17" moped rims, but it would of been quite a mistake to spend the time lacing this motor in a serious wheel.
Vent holes are angled correctly.

No load speed?
No load current?
Power consumption with mostly cruising at that 55% throttle?

~75mph full throttle ~3a
~40mph 55% throttle ~1a
~1kw

Can someone tell me what a hub motor should be at temperature wise for continuous loads? I'd like to hear a what is considered a proper temp range for this sort of motor as a manufacturing/industrial/engineering point of reference.

[veloman]

In my previous post, the ambient temperature of concern is the air temp the motor sees, which I'm sure does vary quite considerably on a hot Texas day about a foot above the asphalt. A quick test for an overall system problem would be to get out and do the same ride just after sunrise. If it gets hot, then you've got a problem that needs to be addressed before any more far hotter daytime riding, though I'd first try some different speeds on that morning ride to see if faster is cooler. My latest big hubbie really loves to fly, and gets warmer with around town errand rides than it does at WOT up moderate grades.

I thought you already knew that a smaller wheel is better. What are you doing with that 26"?

I would have to redesign my whole ebike and get a new frame with disk brake mounts if I went to a 24" wheel. Tire choices are slim. Would need to relace the motor. But yeah, run higher volts on a smaller wheel to keep the motor cooler.... I could run less volts to keep the controller cooler, but then I need to run more amps for the same power. I've invested in a 50v a123 pack anyway.

The motor heating is fairly similar just after sunset. How fast I ride makes a big difference. Also, the motor casing temp is not directly proportional to the winding temps, as I've seen lower winding temps with a hot casing on a hot sunny day of more extended easy riding, than compared to hot motor temps on a fast short night ride.

Last night I rode the tandem 9c 2810 26" wheel on 62 volts nominal, 1700w peak. When we got home, the motor casing was hotter than I had felt on any motor of mine, actually hot enough to hurt a little after a few seconds of holding my hand on it (hottest near the axle). This sort of surprised me. I hope a hall sensor isn't lose because I hear something rubbing when I spun the wheel no load at the end of my ride.

[auraslip]

When we got home, the motor casing was hotter than I had felt on any motor of mine, actually hot enough to hurt a little after a few seconds of holding my hand on it (hottest near the axle).

It can get much, much hotter.

[John in CR]

75mph no load is geared way too high for that power of motor. No wonder it gets hot. To give you an idea, my new big hubbie does right at 75mph no load, but it has enough power and is geared low enough with a 19" wheel that it can do 65mph. You need to back way off on the voltage, but even then the motor really belongs in a much smaller wheel. No wonder so many have heat problems with those motors.

John

[dogman]

Like I was saying, wrong winding or wrong wheel size for the spot on the face of the earth.

Why does it get hot? Cuz you're killing it. You need one of Johns motors if you want to ride like John.

[wesnewell]

10A on 22s is a max 924W. Your motor is rated at 1000W or higher. It shouldn't be getting hot on the flats. Either your sensor is wrong, or there's a problem somewhere. Dry bearings. Wrong phase or phase settings. Bad halls.