The "Danger Zone" Temp Wise

[LI-ghtcycle]

I know this will vary from motor to motor, and I have heard it said that a good rule of thumb is that if the motor is too hot to keep your hand on it comfortably for more than a few seconds it's getting too hot.

I have a IR thermometer, so I'm guessing that things start to get bad for the typical motor around 180 - 200F on the outside (I know this means things are MUCH hotter on the inside, so I am hoping to get a feel for where the limits are on my Golden Motor In-runner "Trike Motor") but how one translates that on this motor I'm not exactly sure where to start.

http://goldenmotor.com/

(It's the one on the far right, the "800W" model)

I plan on trying this motor at 74V (20S of Turnigy Nano Tech 25c) and I will most likely get some liquid cooling via a home computer style liquid cooling system with some hand bent copper tubing between the "fins" of a curved aluminum heat sink on the outside, using a small 12V pump with sump and probably a sealed system will be enough for my needs.

Just wanting to keep things on the cool side for some moderate over-volting.

I have been using this motor at 12S quite successfully with no active cooling and it has only gotten moderately hot going up the local hill (2 miles 5 - 12% grade IIRC) and has never really had me too worried, but at 74V I think it will be a bit more hot. :wink:

 

[amberwolf]

Well, you have a few variables that make it hard to guess at anything like this.

Power you put thru it vs it's efficiency, as well as the mass of the motor and it's design will change how fast it heats up inside vs how fast it heats up outside. So you might see a low temperature outside when the inside is already well past the point of molten solder or burned phase wires, or even damaged magnets (if it's an inrunner, for example). A few minutes later it might be smoking hot on the outside, too.


The only way I know of to tell what the danger zone is, measuring only from the outside, is to first heat it up on the inside to that danger point while monitoring it both inside and out, and seeing how long it takes for the heat to pass thru. But if you set it up for this test, you can already measure the temperatures at the necessary points, making the actual test unnecessary. :lol:

 

[dogman dan]

I found through experimenting, that below 200F inside will never cook anything. Trouble is, outside measurments vary a lot with weather. I've seen smoke pour out of a motor that was cool to the touch when it happened. 5 min later, you couldn't touch that case. An extreme case of course, but I try not to run motors past about 120F measured on the outside. By then, they can really be hot inside. So once you get above 100F measured outside, time to start backing down if you really want to avoid any damage. The trick is to notice if the temp stabilizes at a safe temp, or keeps climbing. If it keeps climbing, you are headed to meltdown if you don't stop.

 

[amberwolf]

Also, if it's a geared motor, there are more layers of thermal barriers. You coudl overheat it to burning while the outside is still completely ambient. Doesnt' apply to the ones in the OP, but as a general rule, it's pretty hard to make a general rule about this sort of thing.

 

[LI-ghtcycle]

Thanks guys, this motor is DD, and I have had it warm enough on the outside at 48V that I felt I needed to be careful going up the local monster 2 mile hill, but never so much that I couldn't hold a hand against it, but it was pretty close to where I wouldn't be able to keep it there long, and it seemed to survive unscathed, this is with a pretty heavy cargo bike and the motor doing 60 - 70% of the work, so I am hoping that it takes higher voltage with some active liquid cooling.

The first thing I need to find is a good way to add the "water jacket" and I have been thinking something along the lines of thermal paste between the motor's outer case and some form of aluminum fins that I can run 1/8 inch copper tubing back and forth through and then connect to a small computer radiator via the small sump & pump one such kit would have for a computer tower, and or lap top if there would be a more efficient for the pump motor at least.

 

[John in CR]

Two guys with a combined post count of over 30k are saying that an ebike motor can suffer a heat failure with an outer casing temp near ambient. Please, that's simply ridiculous. Only an absolute noob could pull that off, because other than a motor with a factory defect the only ways someone could possibly do that are:
1. Incorrect wiring combo - the motor has almost no output power, so almost all the power input is turned into heat, over heating the windings so quickly the heat has no time to spread to the outer shell.
2. Bogged down on a hill - the motor is turning such low rpm that almost all of the power input is turned into heat, over heating the windings so quickly the heat has no time to spread to the outer shell.
3. Setting controller drastically higher current than the motor can handle and taking off with heavy on the throttle with a high load, over heating the windings so quickly the heat has no time to spread to the outer shell.

The solution is simple. With anything new, a new component or new setting, test your bike first going easy on it at first. Increase the stress incrementally and check the entire system often, battery, controller, and motor. Problems show up as heat, sound or poor performance. It's not practical to put temp sensors all over your bike, so feeling it with your hand is very effective, but you need to do it with some regularity so you literally get a feel for when you system is normal making it easy to identify when there's a problem. It's especially important after higher load conditions like hills to gain an understanding of your system limits. Be sure to include the motor axle in that periodic temp check, because bad bearings don't always display other problem indicators such as squeaking until they're long gone. The first indicator is heat, and it shows up as abnormal axle temp, since they're directly connected to the axle. If you haven't felt the axle before, then you have no frame of reference.

One of those IR temp guns would be a cool tool, though since you wouldn't carry it all the time, it would be most useful in calibrating your hand, which is filled with an incredible number of highly sensitive very sophisticated temperature sensors.

The only useful number I saw above was Dogman's 200°F (93­°C) for an interior temperature limit. The high quality high efficiency motors I run now come with factory installed thermistors that have a switch temp of 95°. They're installed on the stator steel about 1cm away from the windings, so it gives a good indication of overall motor temp. Temp at the windings, where most install their temperature sensors, isn't the best location because winding temps can vary within a pretty wide range without issue. If you need to protect your motor from noob problems like 1, 2 and 3 above, then sure put it on the windings, but expect numerous alarms when your motor is fine, and even then 1-2-3 problems are nearly impossible to overcome without use of common sense.

120°F ??? Are you kidding? That's only 49°C and at the motor shell is just warm. That doesn't indicate a problem at all, unless you just took off with a cool motor and immediately bogged it down on a hill.

Using a thermometer and a pot of water you can easily calibrate your hand. Heat it up to boiling and take it off the stove. The outside of the pot is now your "motor", and the temp of the water will be very close to the temp of the outside of the pot, especially using a cheap thin pot. With a sealed motor about 65°C is the high end of acceptable outer shell temp. I factor in what I was just doing to decide whether I need to make some change. If I just finished climbing a long hill, then I know I'm getting close to the limit with that load. If I was riding around on flat ground, then there's problem I need to look into.

Since your motors are inrunners the stator is connected to the shell, so higher temps at the shell are to be expected since the primary heat source is connected directly to the shell instead of separated from it by essentially an insulating layer of air like in a hubmotor. 90°C is probably a good upper limit, but that's just an educated guess.

Don't ignore your nose. As the varnish on magnet wire approaches it's limit it changes in color and puts off a distinctive smell that generally leaks out even with a sealed motor. It does mean you've gone to far, and probably weakened that wire varnish insulation and made it more fragile, but many motors are still fine at that point, but at a minimum you definitely danced with the failure limit.

John

 

[dogman dan]

Did I say I wasn't abusing the motor that melted yet had a cool to the touch case as smoke poured out of it?

Rest of the story, I was bogging that motor down hard in deep sand, AND was running it at about double it's intended wattage. It was a brushed motor too, that I had just gotten. My suspision is that a brush turned into an arc light inside the motor, creating sudden and intense heating. The motor basicly combusted itstelf, to the point where the magnets would no longer stick to steel.

It did happen very fast. And the motor cover was cool to the touch as the smoke poured out. Seldom will the thermal lag be that severe, but it can be enough to put a motor to melty temps inside while the cover feels only moderately warm on the outside. This is particulary possible in cooler weather. I've melted the solder off the phase wires off of a motor that was reading 114F on the outside. I've stopped recomending the outside thermometer, because after several years of doing it that way, I became less and less trusting that it meant shit about actual interior temps. But most of the time, 120F outside means it's a pretty consevative 180F inside, under that 200F limit.

Bottom line, outside temp measurements seem much less reliable the more you are flogging the motor. Under heavy loads, up long steep hills, the motor heats faster so the thermal lag increases. Under moderate loads, outside temp is about 40-50F cooler than inside. That was the number I found in my tests, using moderate loads and 800w of power. Typical commuting at 20-25 mph stuff, not dirt riding, racing, or heavy cargo hauling. For sure, many run motors hotter than 200F inside. I just don't have the actual data at what point the varnish starts stinking and turning black. I just know one motor I never let get hotter than 200 never lost the nice amber color of the varnish.

Another number to remember. Right about 451F, the twine tying together the motor windings will definetly burn to ash. That was the temp taken from my motor at the track, when flames were shooting out the inadequate vent holes. So for sure 451 F is a limit. :wink:

 

[itchynackers]

Dog, don't take offense so much to what John (self appointed expert) says. It boggles my mind to think he is being so technical in some areas, but so lackadaisical in others (hand method). I'm not sure I'm willing to stop and feel my motor in the middle of riding ('waves to Optibike as they pass me by), or ride backward to use the "smell" method. Like John says, it IS simple, monitor temps accurately at the source points of heat. After all, the source will ALWAYS have the highest PEAK temp in the motor. That is where a conservative/logical person will put the temp sensor, and simply glance down at their temp readout instead of stopping, taking off your glove, putting hand on motor, deciding how hot is too hot, did I calibrate my hand?...After all its a bit cooler out today, not too sure, puts glove back on, hops on bike, delivers sandwich late, gets to work late, wastes more time etc... If air cooling were "better" (since ultimately air is the end place for heat right) then all our cars, diesels, tractors, etc. would be air cooled. For some reason he cant grasp that. Not only that, but like you tried to tell him elsewhere, he is lumping apples and oranges together (monstrous motors and small motors) and thinking because results work with large motors, then they MUST work with smalls. Simply not true, and quite obvious to most.

Defense rests, begin John attacks...

 

[itchynackers]

A predictable child. How original.

 

[John in CR]

Go ahead keep digging a deeper hole for yourself.

 

[itchynackers]

I'm not quite sure what that means. Maybe you can explain it to the "simple-minded" folk round here. In an effort to provide some kind of MEANINGFUL discussion, I will say...regardless of your BELIEF, cover temps have very little to do with the real issues. That is, the failure methods of the motors we are talking about. Just try (if for a nano-second) to think that another person on this planet (besides yourself) may have any kind of useful data. Just curious John, what kind of formal education do you have? Why is this relevant? So that it will be easier for the non-believers (me) to believe what you are saying.

Defense rests. Begin meaningless attacks, AGAIN (boredom setting in).

 

[Ratking]

I would only trust a temp sensor glued with thermal glue to the windings of the motor. As itchynackers said, is the only accurate way to really know how hot the motor really is.
My 80-100 is open, so I have the possibility to touch the windings with a finger, but that is still inaccurate. Why not use a temp sensor when it is so easy for those with a C-A?

 

[neptronix]

Inner temps can differ extremely with outer temps. There is a definite delay of time where heat equalizes between the copper and the metal surface that the heat will escape to ambient air through.

I put a temp sensor in my magic pie and rode it in the cold up a hill at about 3kW continuous. The cover remained warm, but the motor was about 70C inside. I let it sit for about a minute and the cover got rather hot, then gradually cooled down. Suprise suprise! The differential in temperature can effect this delay. Short spikes of power followed by gentle cruising will effect this delay too. The only real way to measure things is to get a reading right off what's generating the heat: the copper coils.

The more layers, the worse the 'feeling the cover' situation is. Feeling the case of a Geared hub motor for example, is a complete waste of time.

You have an inrunner motor, which is different than the outrunner motors which are what hub motors are. The outside case temp will give you a pretty good idea of one side of the copper temp, but the magnets in the center of the motor are what you need to worry about the most. You can demagnetize the neo magnets with intense heat rather fast without even knowing it. The heat on the neo magnets is impossible to determine by feeling the case.

There is an amazing amount of additional thermal mass on that motor, which might further distort what's really going on inside.

My recommendation: volt it up and gear it down as much as possible. Go buy yourself a cheapo $15 walmart BBQ temp meter. Pull off the metal probe, and stick the copper sensor 'tween some windings. Extend the wires from the temp sensor if need be, using wires that are a little thicker than the stock ones so that the temp reading does not weaken from resistance. Stick the temp sensor header on your handlebars and you will have an exact idea of when to pull back.

[youtube]dWJSYGAqm0A[/youtube]

I installed this on my magic pie when i was doing hill tests to avoid blowing the motor prematurely. See the temp sensor on the right? it even has an alarm you can set when you hit a certain temp. Very helpful!

 

[LI-ghtcycle]

I would only trust a temp sensor glued with thermal glue to the windings of the motor. As itchynackers said, is the only accurate way to really know how hot the motor really is.
My 80-100 is open, so I have the possibility to touch the windings with a finger, but that is still inaccurate. Why not use a temp sensor when it is so easy for those with a C-A?

Yeah, that sounds like a good idea, didn't realized it was set-up to show temp too!

 

[LI-ghtcycle]

Thanks everyone for your recommendations!

I will put them to use in near future when I get a chance to volt this baby up to 72V+ 8)