Cooling Solutions?

[John in CR]

It wouldn't be "dangerous" for the motor unless you went up to it's physical limits. It's just another source of heat always occurring when the motor spins, even coasting (if it doesn't have a freewheel), and cheaper motors have lower quality iron and make more heat. Revolt motors are a good example, and everyone using them has had heat problems directly due to low quality steel laminations in the stator.

 

[speedyebikenoob]

Okay, I've just always assumed that the controller will limit the rpm before it actually becomes dangerous for the motor, I guess not.

Generic controllers genrally don't check or limit anything at all, other than battery current (and even that can be significantly different than what it's rated for, above or below, because of inconsistent materials or parts used, or poor manufacturing causing too little or too much solder at the shunts).

There *are* controllers that do various checks and limits, but they generally are significantly more expensive, and also require setup for the specific motor you're using them with, which can be kind of complicated for some of them.

Some of the Kelly controllers, Lebowski's, Incememed's SFOC, Phaserunner / ASI, Sevcon, others I cna't think of off the top of my head.

Yup I have the cheap stock controller lol, don't think it has anything fancy going on in it. So when the controller shuts off when you're really hauling it it's probably because the controller senses the internal heat inside the casing is too high right? It's not for the motor?

 

[speedyebikenoob]

It wouldn't be "dangerous" for the motor unless you went up to it's physical limits. It's just another source of heat always occurring when the motor spins, even coasting (if it doesn't have a freewheel), and cheaper motors have lower quality iron and make more heat. Revolt motors are a good example, and everyone using them has had heat problems directly due to low quality steel laminations in the stator.

I see. It's been raining the past couple days and I don't have anything to protect stuff from shorting out, but once it gets better, apparently this weekend, I'll do some tests.

 

[amberwolf]

Yup I have the cheap stock controller lol, don't think it has anything fancy going on in it. So when the controller shuts off when you're really hauling it it's probably because the controller senses the internal heat inside the casing is too high right? It's not for the motor?

It's probably not even sensing controller temperature--like I said, in the generics there's literally nothing sensed or limted except battery current (well, voltage usually has an LVC, and occasionally an HVC). (there are some that do, but most do not, until you start getting into more expensive versions, like, for an instance here on ES, the Powervelocity series, which are still "generic" controllers, but that have been customized, and have more features than typical ones. )

If a motor wiht hall sensors gets hot enough, the hall sensors can begin to fail, sometimes just temporarily, sometimes permanently damaged, and that can cause a sensored controlelr to stop working correctly (or at all), because it isn't gettting position from the mtoro so it can't send phase current in timing that works.

If the FETs get hot enough they may stop working correctly--but often this is a complete catastrophic failure.

If the capacitors get hot enough, they may leak or explode.

If the MCU gets hot enough, it might stop working like the hall sensors, in a temporary way, but past some point it'd just fail permanently. Similarly, other non-power-handling semiconductors in there could do that.


But...the most common thing that causes shutdown while pulling a lot of power is the battery. The cells sag in voltage more then, and if any of htem are weak enough they'll go below the LVC of the BMS, and that will shutdown everything, at least momentarily. If there is no BMS, or it doesn't go quite low enough to trigger it, it could still go below the LVC of the controller (if it has one), and the controller will stop working utnil voltage goes back up once the load is removed.

 

[John in CR]

It wouldn't be "dangerous" for the motor unless you went up to it's physical limits. It's just another source of heat always occurring when the motor spins, even coasting (if it doesn't have a freewheel), and cheaper motors have lower quality iron and make more heat. Revolt motors are a good example, and everyone using them has had heat problems directly due to low quality steel laminations in the stator.

I see. It's been raining the past couple days and I don't have anything to protect stuff from shorting out, but once it gets better, apparently this weekend, I'll do some tests.

Tests of rpm creating heat are best done inside with the chain off, so current is limited and can't be blamed for the heat. No load current X voltage gives you the amount of heat created just to spin the motor. My guess would be that's in the several hundred watt range, but if it's 500W or more, then it's a major contributor to your motor temps.

 

[amberwolf]

There'd have to be something wrong, or some significant load, to be pulling several hundred watts just to spin the motor.

Typical unloaded motor currents at full speed should be in the 1-digit amps, for most of the "ebike-sized" motors, from my own experiences and what I've read here on ES that I can recall.

Let's say the cyclone's is high, at 5A unloaded motor current. Let's say 55v pack voltage under that load. That's 55 * 5 = 275W.


I poked around the big Cyclone thread and found this post:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=69867&p=1389773&hilit=unloaded#p1389773
that indicates that even with the gearbox it only takes a couple hundred watts at 18s, whcih is comparable to the Tangent motor running it's gearbox.


The last time I tested it, my MXUS 4503 hubmotor, offground but with the wheel on it, at that same voltage, was about 2.5A, IIRC. Round that up to 3A, * 55v, is only 165W.

 

[speedyebikenoob]

It wouldn't be "dangerous" for the motor unless you went up to it's physical limits. It's just another source of heat always occurring when the motor spins, even coasting (if it doesn't have a freewheel), and cheaper motors have lower quality iron and make more heat. Revolt motors are a good example, and everyone using them has had heat problems directly due to low quality steel laminations in the stator.

I see. It's been raining the past couple days and I don't have anything to protect stuff from shorting out, but once it gets better, apparently this weekend, I'll do some tests.

Tests of rpm creating heat are best done inside with the chain off, so current is limited and can't be blamed for the heat. No load current X voltage gives you the amount of heat created just to spin the motor. My guess would be that's in the several hundred watt range, but if it's 500W or more, then it's a major contributor to your motor temps.

Should the motor be getting hot if it's just spinning the wheel at the back? Mine doesn't get hot but it reaches maybe 80-85 fahrenheit on the outside, it's just warm to the touch. Also, how do I measure the watts? I don't have a cycle analyst or anything

 

[speedyebikenoob]

There'd have to be something wrong, or some significant load, to be pulling several hundred watts just to spin the motor.

Typical unloaded motor currents at full speed should be in the 1-digit amps, for most of the "ebike-sized" motors, from my own experiences and what I've read here on ES that I can recall.

Let's say the cyclone's is high, at 5A unloaded motor current. Let's say 55v pack voltage under that load. That's 55 * 5 = 275W.


I poked around the big Cyclone thread and found this post:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=69867&p=1389773&hilit=unloaded#p1389773
that indicates that even with the gearbox it only takes a couple hundred watts at 18s, whcih is comparable to the Tangent motor running it's gearbox.


The last time I tested it, my MXUS 4503 hubmotor, offground but with the wheel on it, at that same voltage, was about 2.5A, IIRC. Round that up to 3A, * 55v, is only 165W.

How did you test yours? I don't have a cycle analyst or any one of those computers, but I do have a multi meter though :lol:

 

[speedyebikenoob]

So I compared the amount of heat at the outer casing to the amount of heat at the shaft of the motor and there is a significant difference. The shaft is about 35 degrees fahrenheit cooler than the casing of the motor. When the casing was at 110, the shaft was at 75. That sound like a healthy safe temperature?

 

[John in CR]

So I compared the amount of heat at the outer casing to the amount of heat at the shaft of the motor and there is a significant difference. The shaft is about 35 degrees fahrenheit cooler than the casing of the motor. When the casing was at 110, the shaft was at 75. That sound like a healthy safe temperature?

Yes, that's a good sign, but not definitive. You can determine no-load losses (heat just to spin the motor) with a multimeter. With the chain disconnected and a cold motor, measure the current to spin it at wide open throttle, while going easy on the throttle to get there so current doesn't spike while getting it to max rpm. That current X pack voltage is how many watts required (all heat) just to spin the motor.

Since the motor is relatively small and your temps aren't all that high to begin with, you should see some temp improvement simply using something to deflect more air flow from moving toward the motor. Anything you do that results in a cooler motor is a good thing. The bulk of the motor losses on our ebikes is always losses due to resistance of the copper. Resistance of copper increases by 0.22% for each degree fahrenheit, so a cooler motor always increases efficiency (lowers losses). On the motors I run at extreme performance my cooling efforts decrease motor temps by at least 50°F, so it lowers losses in the motor by over 10%, which enables me to run at much higher power and I get more out of my battery pack.

 

[speedyebikenoob]

So I compared the amount of heat at the outer casing to the amount of heat at the shaft of the motor and there is a significant difference. The shaft is about 35 degrees fahrenheit cooler than the casing of the motor. When the casing was at 110, the shaft was at 75. That sound like a healthy safe temperature?

Yes, that's a good sign, but not definitive. You can determine no-load losses (heat just to spin the motor) with a multimeter. With the chain disconnected and a cold motor, measure the current to spin it at wide open throttle, while going easy on the throttle to get there so current doesn't spike while getting it to max rpm. That current X pack voltage is how many watts required (all heat) just to spin the motor.

Since the motor is relatively small and your temps aren't all that high to begin with, you should see some temp improvement simply using something to deflect more air flow from moving toward the motor. Anything you do that results in a cooler motor is a good thing. The bulk of the motor losses on our ebikes is always losses due to resistance of the copper. Resistance of copper increases by 0.22% for each degree fahrenheit, so a cooler motor always increases efficiency (lowers losses). On the motors I run at extreme performance my cooling efforts decrease motor temps by at least 50°F, so it lowers losses in the motor by over 10%, which enables me to run at much higher power and I get more out of my battery pack.

So where do I connect the leads of the multi meter to to find the current? Battery?

 

[Chalo]

So where do I connect the leads of the multi meter to to find the current? Battery?

You don't use a multimeter for that. (Usually. If your multimeter is super precise, you can measure voltage drop across the ends of a calibrated shunt and calculate current from the result.) Most of us use a watt meter, power analyzer or ammeter to measure current. The meter, or at least the shunt for the meter, has to be placed in series between one of the battery leads and the load.

 

[John in CR]

So where do I connect the leads of the multi meter to to find the current? Battery?

You put the multimeter probe in the unfuse hole with the higher current limit (typically 10A). That's for measuring current above the very low current limit of the rest of the multimeter. Be sure to set the multimeter dial on the 10A limit setting also. You probably wondered what that 3rd probe hole on the multimeter was for, and now you know. While you need to go to the trouble of a shunt and measure voltage drop across it to measure high current, and I wouldn't use the multimeter to measure close to 10A continuously, it's perfect for measuring no-load current on our motors as long as you don't really goose the throttle and make current spike much higher. For measuring up to 10A for a few seconds, it's perfect and easy. In fact, I have a cheap multi-meter with a broken dial that is now dedicated to measuring current under 10A with the dial superglued on that setting and the probes glued in place with alligator clips on the end.

Yes, complete your battery connection through the multimeter to measure the no load current.

 

[speedyebikenoob]

So where do I connect the leads of the multi meter to to find the current? Battery?

You don't use a multimeter for that. (Usually. If your multimeter is super precise, you can measure voltage drop across the ends of a calibrated shunt and calculate current from the result.) Most of us use a watt meter, power analyzer or ammeter to measure current. The meter, or at least the shunt for the meter, has to be placed in series between one of the battery leads and the load.

I see. No idea how precise mine is anyways, will probably end up getting a wattmeter soon regardless.

 

[speedyebikenoob]

So where do I connect the leads of the multi meter to to find the current? Battery?

You put the multimeter probe in the unfuse hole with the higher current limit (typically 10A). That's for measuring current above the very low current limit of the rest of the multimeter. Be sure to set the multimeter dial on the 10A limit setting also. You probably wondered what that 3rd probe hole on the multimeter was for, and now you know. While you need to go to the trouble of a shunt and measure voltage drop across it to measure high current, and I wouldn't use the multimeter to measure close to 10A continuously, it's perfect for measuring no-load current on our motors as long as you don't really goose the throttle and make current spike much higher. For measuring up to 10A for a few seconds, it's perfect and easy. In fact, I have a cheap multi-meter with a broken dial that is now dedicated to measuring current under 10A with the dial superglued on that setting and the probes glued in place with alligator clips on the end.

Yes, complete your battery connection through the multimeter to measure the no load current.

Sorry for the really late response, I actually tried and the multimeter just read zero. I basically tied some speaker wire to the wires that go from the battery to the motor controller, and as I ran the motor it just didn't give me any reading other than the voltage.

 

[Chalo]

Umm....

I said earlier you'd need to put the meter in series with the load. Do you know what that means, in practical terms?

One of the implications is that if you use a cheap meter with super thin lead wires, they'll get really hot, maybe even fuse open. Another implication is that you will leave either the + wire or the - wire alone.

 

[speedyebikenoob]

Umm....

I said earlier you'd need to put the meter in series with the load. Do you know what that means, in practical terms?

One of the implications is that if you use a cheap meter with super thin lead wires, they'll get really hot, maybe even fuse open. Another implication is that you will leave either the + wire or the - wire alone.

I actually don't, atleast not exactly. But I'm guessing I have to put the meter where the motor is drawing the current from, so the wires from the controller? The multimeter looks like it's rated for 10 amps, if it's not underload, I don't think it would exceed that?