Why is my new hub motor making my controller super hot?
- Thread starter ClintBX
- Start date
On a semi related note, I'm getting a new controller that spec'd at 1200 watt. The amp rating is a little higher than my current controller. It's a 35 and I'm currently using a 30 amp. Will the new controller burn through my battery faster? I really don't want to reduce the mileage I get from my new 30 ah battery.
Voltron
1 MW
Sure it's allowed... And low amps, like less than 5. Not 15 or 20.
And the new controller won't reduce your mileage... as long as you don't go punching it every block. But that is tougher than it sounds with a nice new more powerful controller
And the new controller won't reduce your mileage... as long as you don't go punching it every block. But that is tougher than it sounds with a nice new more powerful controller
Alan B
100 GW
If the motor phases and hall sensors are not sequenced properly the motor may (or may not) run but it can draw a lot more power than it should, heating both motor and controller. Hence:
1) get the phasing right first. with no load it should run smoothly from zero to full speed at very low power. Color codes may differ.
2) then observe the temperature of the controller and see if it needs more airflow. It should get some but in normal circumstances won't get very warm.
Controllers are made to withstand moisture and shock. They are not designed to be operated without airflow, though they don't need a great deal.
The important controller ratings are voltage and current. Voltage must be adequate (with a safety margin) and current determines the power at your operating voltage.
It sounds like the new motor is a lower voltage motor, so it will spin faster and draw more current. This may make the controller slightly hotter. Much hotter indicates a phasing problem.
1) get the phasing right first. with no load it should run smoothly from zero to full speed at very low power. Color codes may differ.
2) then observe the temperature of the controller and see if it needs more airflow. It should get some but in normal circumstances won't get very warm.
Controllers are made to withstand moisture and shock. They are not designed to be operated without airflow, though they don't need a great deal.
The important controller ratings are voltage and current. Voltage must be adequate (with a safety margin) and current determines the power at your operating voltage.
It sounds like the new motor is a lower voltage motor, so it will spin faster and draw more current. This may make the controller slightly hotter. Much hotter indicates a phasing problem.
Alan B said:
Lower voltage? The only specs I see for both controllers and motors are their watts and the amp discharge.
And I'm pretty sure that my pairing a 1200 watt motor with a 500-600 watt controller is the source of the problem.
That said, I still haven't played around with the phase order yet. The plastic of the controller's yellow phase has sortta fused itself. I don't really want to mess with it until I have the new controller. I figured I'll test out the phase order then. Still skeptical that the phases are wrong. Its just too smooth.
Wheazel
10 kW
ClintBX said:
Are you well understood about U x I = P??
So voltage times current is power. That is why you can see controllers with different amp ratings have the same power rating for example.
Or the other way around. The voltage is a factor.
Also, if the new motor is a faster wind, you might be pulling a lot more power overall just by going faster.
You spoke about 10km/h faster, but faster from what speed? This matters as air resistance is in a squared relation to speed.
Meaning 20-30km/h is not the same power increase as 30-40km/h. As an example, my cargobike will be reasonably efficient up to 40km/h.
But going to 50km/h or even 60km/h, its like racing a barndoor thorugh a fruit cake.
Alan B
100 GW
There are lots of specs for motors (and controllers) that they're not telling you, such as RPM per Volt, and the windings. You are basically throwing darts blindfolded when buying components this way.
ClintBX said:
No, it's not.
All "a 1200w motor" means is that the motor has been "rated" for handling up to 1200w by the seller (and maybe by the maker, but who knows?). It doesn't have anything to do with how much power it actually uses.
That is limited by the controller, and determined by the load on the system (the work it has to do pulling you down the road).
If the controller is rated at 500-600w (by the seller, not necessarily the maker), then that's what it will be able to output to the motor, depending on the system load.
But if the system load, for whatever reason, is higher than that, the controller would have to provide that amount, only cutting back at whatever it's current limit is programmed to be (which in combination with whatever voltage it's run at, might be a lot more watts than the controller can actually handle).
So you could have a "1200w motor" and try taking it up a really steep hill, and have it draw 1800w, if it's running on ~52v (under load) (just a figure, I don't know your actual battery voltage under load; you'd have to test that yourself), and the controller is limiting at ~34A (they aren't usually exact, and mostly err a few amps higher than the stated limit).
That would be about three times the watts your controller is "rated" for, even though the controller is actually limiting current as it should.
If you used a controller "rated" for 1200w+, it might have an even higher current limit, that would in the same situation allow even more power thru the motor (probably enough to cook the motor); it might have an 80-100A current limit, for instance, which would also greatly abuse most ebike batteries. If it did have say an 80A limit, then at that same 52V (actually, let's say 48v because the battery would sag more at the higher current), the controlller would actually have to provide almost 4000w, and the motor would be heating up quite a lot at that point (though you'd probably be going up that hill nice and easy
). Anyway, the above is just postulation, since we don't know any of your actual system's usage for current draw, or what the voltage sag is at various current draws, to determine what watts are actuallly being drawn from the system.
This means that connection point has gotten very hot, which usually means it is a poor connection. It can also mean it is arcing inside the connection, which can also cause controller heating due to RF from the arcing feeding back into the FETs and disrupting their normal operation. (can also blow them up).
Alright. A lot of information here.
Okay, just to be clear, I'm running a 48v 30 ah battery with a 1200 watt motor and a 30 amp 500-600 watt controller.
It's true that while my controller is getting super hot, the motor never feels more than Luke warm to the touch.
Soooo... if the amount of power a motor draws is determined by what the controller is programmed to dish out, why is it a bad thing to have an under spec'd controller that gets a whole lot hotter the moment you pair it with a bigger motor?
By that logic, the motor shouldn't be getting anymore than what the controller is programmed to give it. There should be no additional stress on the controller than usual, right?
Okay, just to be clear, I'm running a 48v 30 ah battery with a 1200 watt motor and a 30 amp 500-600 watt controller.
It's true that while my controller is getting super hot, the motor never feels more than Luke warm to the touch.
Soooo... if the amount of power a motor draws is determined by what the controller is programmed to dish out, why is it a bad thing to have an under spec'd controller that gets a whole lot hotter the moment you pair it with a bigger motor?
By that logic, the motor shouldn't be getting anymore than what the controller is programmed to give it. There should be no additional stress on the controller than usual, right?
Another thing I didn't mention is that my battery connectors have been melting down a lot since the new motor.
I use anderson power poles. They were fine with the old motor. Only had to replace one once due to bad crimping. Since the new motor, the plastic melts and has been fusing the 2 connectors together repeatedly. Finally got around to swapping the red PP to a bullet connector today. (Haven't had any melting black connectors yet for some reason)
I use anderson power poles. They were fine with the old motor. Only had to replace one once due to bad crimping. Since the new motor, the plastic melts and has been fusing the 2 connectors together repeatedly. Finally got around to swapping the red PP to a bullet connector today. (Haven't had any melting black connectors yet for some reason)
ClintBX said:
amberwolf said:
What are the actual measured voltages under load, and what are the actual currents drawn from the battery, and what are the actual wattages?ClintBX said:
If you don't know those things, then you should test them, because troubleshooting based on "specifications" that someone decided to print on a part or a web page is not reliable--everything anyone can say to you is a guess, at best.
If you don't answer/don't test things people suggest, all they can do is make more guesses, speculate more in random directions, wasting more time for everyone. Sometimes it turns out to be a waste of time to do the tests, too, but if they're not done there's no way to know what the problem is, or what to do to fix it, without just swapping stuff out till it goes away.
That gets expensive and frustrating.
amberwolf said:
How am I supposed to test under load? I have a watt meter (can't currently connect the load end due to yet anotger bad connector) but while I can use it to tell my ah usage after a ride, I can't look at the numbers while riding.
Would I need some longer wires? I think I read something about using a shunt.
Its hard to do the tests people ask me to do when I don't have the equipment to test with,
I don't know what equipment to get,
and when I do, I can only get it online and wait 3-6 weeks to receive it because my local electronics store doesn't sell it (or even knows what it is when I ask them).
So, if you can understand my frustration.
Anyway, my previous comments were (aside from answering and providing extra deets on my experience) just about trying to understand some contradictions.
Alan B
100 GW
ClintBX said:
To get back to basics here, there is one reason the controller will get hot. It is handling too much current. What heats a controller is I squared R. The I is battery current and motor current, and the R is the resistance of the FETs, which goes up with temperature. So as the FETs get hot they heat even more. But notice that the current term is squared, so doubling the current makes four times the heat.
Now why would there be more current?
If the motor is producing the same top speed and the same acceleration then it should consume the same current. If the performance is significantly different, eg it goes much faster or accelerates much more aggressively, then it will draw more current. But if it is about the same, or a little more, then we have to look elsewhere for the greater current that we know is flowing.
This is where we get to the electrical timing. If the timing is wrong, we can have a situation where 2/3 of the motor current is driving it forward and 1/3 is driving it in reverse. It might be rough, or it might be fairly smooth depending on the motor's design. But in this case 2/3 of the current is causing cancellation of motor torque (in other words, turning to heat), and only 1/3 of the current is doing useful work. Since controller heat is from current squared, this means the current, which is tripled, would cause 9 times the heat in the controller. That would heat it up in a hurry.
It could just be that you've exceeded the current that this small controller can handle. This seems unlikely as the difference you are reporting in controller temperature is so great. But it is possible. Cheap controllers monitor battery current, not motor current. So it could be operating at 30A of battery current, but really high motor current, and the controller might not be able to protect itself, so be careful.
On the subject of Anderson Powerpoles. It is an excellent connector, but realistically the 15/30/45 amp rating is peak, not continuous. You can get the ratings curves from Anderson and review the temperature rise data. It depends on the airflow, wire size, and duration, but 15 to 20 amps is all that is safe for continuous current. Move to the SB50 or PowerPole 75's if you are going to run 30 amps continuous. Also note that motor current is essentially always greater than battery current, slightly mitigated by being 3 phase, but those connectors should also be suitably increased in current capacity.
But there's a good chance that adjusting the phasing (electrical timing) will fix the majority of your excess current.
Incidentally, weight is a fair indicator of hubmotor power handling. If your new motor weighs similar to your old motor then they likely have similar power ratings (there are exceptions, but not likely found in cheap motors). The advertised power ratings are not to be trusted (until verified). One may be a continuous rating, the other peak. You just don't really know.
You may want to move to a better instrumentation system on your ebike. Something like a cycle analyst that can measure a lot more of what you need to know.
The winding speed of the motor has a big effect on how hot the controller gets. A high-speed motor will draw the maximum current to a higher speed than a low-speed one because the back EMF is lower. Let's say you had a 30 mph 1200w motor and a 30 amp controller. At 25 mph, you'd be past the point when the back emf limits the current more than the controller, so you'd be down to something like 25 amps, which would reduce to zero at 30 mph. A motor wound for 40 mph would still be drawing the full 30 amps at 25 mph and at 30 mph. Only after that would it start to reduce, so at 28 mph, you could be drawing ten times the current than you were before. Basically, it's not just the motor that overheats when you have a high-speed motor. If everything else is equal, the controller will run hotter too.
Alan B
100 GW
If the top speed is the same or similar then the winding is not much different. What were the before/after top speeds? I see there was some mention of 10 km/h, but no reference as to from what speed to what speed. That's a pretty small change in top speed, not likely enough to explain the tremendous delta T in controller temps.
If the performance is significantly different then yes, that could be a major factor. The OP has not made clear how much difference there is. The controller temperature difference (as described) between the motors is so huge it would likely require more than a 2x motor difference.
It would be nice if the OP told us more about the speed / acceleration differences between these two motors. The nameplate values aren't very helpful.
In any case the data we have points more to incorrect timing than to the motor change. Going from what is probably a 6 FET controller to what is likely a 12 FET controller will help, but not solve the problem if the phasing is incorrect. Of course the new controller will need to be rephased anyway.
If the performance is significantly different then yes, that could be a major factor. The OP has not made clear how much difference there is. The controller temperature difference (as described) between the motors is so huge it would likely require more than a 2x motor difference.
It would be nice if the OP told us more about the speed / acceleration differences between these two motors. The nameplate values aren't very helpful.
In any case the data we have points more to incorrect timing than to the motor change. Going from what is probably a 6 FET controller to what is likely a 12 FET controller will help, but not solve the problem if the phasing is incorrect. Of course the new controller will need to be rephased anyway.
All the watt meters I've used will retain a trip statistic for at least:ClintBX said:
Ah
Wh
Peak Watts
Peak Amps
Lowest Voltage
Sometimes other stuff.
You don't have to look at the info while riding (though that is useful to know what the W, A, and V are at any particular point so you know what is happening under different ride conditions).
Some have a battery (or place for one) to retain this info even when the system is turned off, but most people don't use them that way, so you have to check the stats and write them down (or take pictures of each screen) at the end of each ride before you turn the system off.
So just noting that info down and keeping a log of it will help you know when things change, and then you can compare that info to the problem you have and see if it tells you where the problem is. If it doesn't help you directly, you can then post that info (both previous "normal" plus new "problem" data) along with the problem description, and it will help us with definite data to help you with more definite solutions, rather than guesses.
Without the load end connected, then the only function it can do is voltmeter--it can tell you what the volts are right now, and it can tell you what the lowest volts (sag) were during the ride.
To use it for Ah, Wh, W, A, the load end must be connected so that current passes thru the wattmeter from battery to controller.
If you want to mount it on the bars to read realtime, then you'd need to run thick-gauge power wires from the battery, to the bars at the wattmeter, then back down to the controller.
Alternately you'd need to open up the wattmeter, take it's shunt off, and run thinner wires from the place it was inside to where you will then mount it inline with the original wires from battery to controller. This is not that hard, but you'll want to look up how to do this, in threads like the ones about the original Turnigy Watt Meter:
A wattmeter (which you already have), or a multimeter, or separate voltmeter and ammeter.
They can be used instead of a wattmeter, but have to be setup the same way, so it's easier to just use the wattmeter if you already have it.
They can come from auto parts stores, harbor freight, walmart/kmart/etc, quite a few types of places carry the cheap ones. Even some computer stores carry them (fry's electronics, for instance). Even thrift stores and pawn shops probably have them. Then there is Craigslist and Freecycle, etc.
A little looking around at things outside the "boxes" and labels people put around things can find sources for stuff you need.
A "cycle computer" or other type of speedometer is also useful, so you can mount the magnet on the driven (motor) wheel, because it can be used to test both on-road speed and off-ground no-load speed, which is useful for a number of tests (to find info like that speculated on in posts above this).
Alan B said:
There are other factors too, like whether the motor has the power to pull its maximum speed (or near).
As an example:
A Q100 motor with a standard 15 amp controller. The 201 rpm one will max out at about 15 mph on the flat and the 328 rpm one can do about 20 mph (no-load speeds about 17 mph and 26 mph). The back emf starts cutting the current of the 201 rpm one at about 11 mph and the 328 rpm at about 20 mph, so at 15 mph, the 201 rpm is getting about 7 amps and the 328 rpm gets the full 15 amps all the way to 20 mph. That means that the 328 rpm one goes 5 mph faster, but draws double the current from the same controller. The two motors are identical from the outside.
I tried that setup with the 328 rpm one, but the controller overheated all the time, since 15 amp max controllers are only rated at 7 amps. It needed a controller rated for and limited to 15 amps. I've never had a problem with the 260 rpm or 201 rpm version. The same principle applies to any motor. The biggest single factor that affects its performance is the winding speed. Surprisingly, most people don't even know what theirs is!
If you use the simulator to compare any two motors with different winding speeds and the same battery and controller, you'll see how the higher speed variant always draws higher current in the high speed range (look at torque, which is more or less proportional to current). The higher speed motor.
http://www.ebikes.ca/tools/simulator.html
Here's an example. At 30 kph, the current is more or less the same. At 35 kph, the higher speed one is drawing double the current. At 37 kph, it's drawing ten times the current:
Alan B
100 GW
What you suggest is possible - someone could make a really low voltage motor, perhaps designed to have the same performance at 24V that a motor made for 48V would normally have, an extreme "fast wind" motor. That would draw extremely high current at 48V, but it would also produce crazy acceleration and speed while it overheated the controller. In this case the motor behavior would be so different from the previous motor that the OP would not be surprised by a hot controller. There would likely be trouble keeping the front wheel on the ground, for exampe, and the motor would tend to get hot as well.
In the case you have shown above there is a difference in power, but it is only different by the amount that the load has increased due to higher speed. The difference in motor current is largely due to the motor being a lower voltage motor, so to achieve the power with the lower back EMF the current must then be higher. There is an effect there, it just doesn't have enough magnitude (less than 2:1 in this case) to account for such a huge increase in controller heat.
The only two points where the data in that graph represent actual continuous operating conditions are where the torque curve crosses the load line (for each case). The power in these cases differs there by a relatively small percentage. At all other points the motor is either accelerating, or it is beyond equilibrium and would not operate continuously there. Going beyond the load line intersection indicates an over speed condition where the motor is not producing enough power to continue at that speed, so it is slowing down, so the power consumed at that point is low but meaningless. During acceleration there is more power being used, however this is a transient condition and would not tend to heat the controller so much due to the short time involved.
There could also be other causes, like a shorted turn inside the motor, or incorrectly installed Hall Sensors, or a motor with some other defect. We are working with very little data here.
If the OP can measure the power of both motors unloaded and running at full speed, and what the unloaded full speed for each motor is, we could learn a lot.
In the case you have shown above there is a difference in power, but it is only different by the amount that the load has increased due to higher speed. The difference in motor current is largely due to the motor being a lower voltage motor, so to achieve the power with the lower back EMF the current must then be higher. There is an effect there, it just doesn't have enough magnitude (less than 2:1 in this case) to account for such a huge increase in controller heat.
The only two points where the data in that graph represent actual continuous operating conditions are where the torque curve crosses the load line (for each case). The power in these cases differs there by a relatively small percentage. At all other points the motor is either accelerating, or it is beyond equilibrium and would not operate continuously there. Going beyond the load line intersection indicates an over speed condition where the motor is not producing enough power to continue at that speed, so it is slowing down, so the power consumed at that point is low but meaningless. During acceleration there is more power being used, however this is a transient condition and would not tend to heat the controller so much due to the short time involved.
There could also be other causes, like a shorted turn inside the motor, or incorrectly installed Hall Sensors, or a motor with some other defect. We are working with very little data here.
If the OP can measure the power of both motors unloaded and running at full speed, and what the unloaded full speed for each motor is, we could learn a lot.
I don't think you have it right there, Alan, IIUYC. The controller provides the current rather than the motor drawing it, so a different winding motor won't make a significantly different overall power. The motor wants to take a lot of current, but it's limited by the controller until the speed where the back emf reduces the net voltage to a point where the current will be less than the controller limit. If you have the same motor with different kVs and everything else the same, the lower kV one will produce slightly more power until it reaches its maximum power. After that, the higher speed one makes more power than the low-speed one, but not more power than the lower one's maximum. It just produces the power for longer (speedwise).`
you can see on those charts above the point where the back emf starts to limit the current. it's where the torque/current line kings downwards and ramps down to zero.
you can see on those charts above the point where the back emf starts to limit the current. it's where the torque/current line kings downwards and ramps down to zero.
Alan B
100 GW
The controller provides motor voltage proportioned by PWM and the current that flows is a consequence of motor parameters - motor voltage minus back EMF divided by total resistance. The controller measures battery current and calculates estimated motor current and limits PWM based on whichever is the more limiting value after the block time has elapsed.
You are correct when the controller behaves like the sim. But the sim does not actually model the controller behavior during block time and during the motor current limiting phase of the acceleration. The small controller will tend to mask the power surges that are normal for speed wind motors and more capable controllers. But many folks run small 6 FET controllers with DD hubmotors without this level of heating.
In any case we are making a lot of assumptions that are affecting our predictions of the behavior.
We need more data from the OP. Either the larger controller results or the unloaded speed/power would tell a lot.
You are correct when the controller behaves like the sim. But the sim does not actually model the controller behavior during block time and during the motor current limiting phase of the acceleration. The small controller will tend to mask the power surges that are normal for speed wind motors and more capable controllers. But many folks run small 6 FET controllers with DD hubmotors without this level of heating.
In any case we are making a lot of assumptions that are affecting our predictions of the behavior.
We need more data from the OP. Either the larger controller results or the unloaded speed/power would tell a lot.
Alan B said:
My previous motor was topping around 35 km/h. Though I didn't pay attention to the absolute max speed on flats.
I may have exaggerated that 10km/h increase. Most of the time, my current motor peaks around 39-40. It takes a while to see it creep up to 42-43.
chas58
10 kW
Boy, I was thinking of the best way to reply, and then D8Veh said it all quite eloquently.
yep, the controller determines the power, not the motor, and yes if the motor is not operating past the inflection point (where the current beginns to drop off) it will need more power from the controller than it should.
I guess the real question is, what is the noload speed of your motor, and what is the load speed? If there is a large difference, you may be pulling max amps from the controller for too long.
You could get a controller that allows you to adjust current to keep it from getting too hot, and you could get a cheap watt meter from Luna cycles.
Either way, get the controller out of the bag. A good place to put it is behind the seat tube, above the rear fender, although I like putting it on the underside of a rear rack mount (again, above your rear fender.
600 watts mountain bike
377 watts my bike (road bike)
265 watts recumbant
yep, the controller determines the power, not the motor, and yes if the motor is not operating past the inflection point (where the current beginns to drop off) it will need more power from the controller than it should.
I guess the real question is, what is the noload speed of your motor, and what is the load speed? If there is a large difference, you may be pulling max amps from the controller for too long.
You could get a controller that allows you to adjust current to keep it from getting too hot, and you could get a cheap watt meter from Luna cycles.
Either way, get the controller out of the bag. A good place to put it is behind the seat tube, above the rear fender, although I like putting it on the underside of a rear rack mount (again, above your rear fender.
Never say never.
600 watts mountain bike
377 watts my bike (road bike)
265 watts recumbant
Similar threads
