Motor heat sensor

[Mathurin]

I'd like to mount some kinda heat sensor on the windings of whatever motor I end up using. I don't know how to do that, nor do I know at what temp I should back off. I know it's been done, but a quick googling yielded nothing.

Suggestions?

 

[roller]

brush hub or bldc shaft drive, no prob. As long as your windings aren't spinning.

glue a thermocouple on the coils, add strain relief, voila. no other part of the motor is worth measuring.

I think you can get lenghts of it in McMasterCarr, though likely through test/measurement catalogues.

you can read the V out for the given type of couple (J, T, K, etc).

I've also seen decent DMMs w/ thermocouple included for about $45.

if the coils are the spinning part, you have to go non-contact (Raytek, infrared).

 

[Lowell]

I wonder if a wireless sensor would work given the electrical noise inside a motor. Something similar to the tire pressure monitors fitted to modern cars.

 

[maxwell]

I use a thermistor, much cheaper than thermocouples, you can get a whole thing with display just find an internal/external thermometer, junk the outer case of the external sensor.

Sinclair in his C5 drilled out the motor shaft (normal brushed DC) to insert a thermocouple for test purposes. Do a search for Perrin Newman (head of the C5 electronics team, I worked for him for a bit not on the C5 though) and C5, there is an interesting article out there on the web.

 

[Ypedal]

At Canadian Tire this week they have the wireless Indoor/Outdoor digital thermometer for 9.99 !!!! I picked up 2 ! works on 900mhz, good for 100m range. ( or so they claim )

On the c-lyte brushless motors, there would be enough room inside the hub to fit the rig if you remove the plastic case.. powering it with 2xAA might be a hasle but a pair of batteries would likely last a whole season.

---

That said.. i don't think you have much to worry about.. i've pumped 1600w into my 406/409 without problems.. gets warm. but not hot. ( guessing aprox 45 celcius or so on the outside motor cover )

 

[fechter]

I have a couple of those cheap remote reading digital thermometers. The thermistor just uses a pair of wires, so if you need to cut and splice the wire to stuff it through the axle, it's not too complicated.

Like Roller says, you need to measure the temperature of the windings.
You can attach the thermistor to the windings with a blob of silicone glue. Silicone is a good heat conductor and can take very high temperatures.

Ideally, when the temperature reaches close to the maximum, it would automatically decrease the current limit in the controller.

 

[Mathurin]

Ok, good leads.

I'm thinking about 1.3Kw out, not in.



What's a safe temp limit for a hubmotor?

 

[Mathurin]

How about this thing, could feed it off the hall sensors and only need 1 extra wire out the motor:

http://www.facstaff.bucknell.edu/mastascu/eLessonsHTML/Sensors/TempLM35.html

 

[Lowell]

Looks good, and you could swap the hall sensor connector for a 6 pin to make it direct plug in.

 

[fechter]

How about this thing, could feed it off the hall sensors and only need 1 extra wire out the motor:

http://www.facstaff.bucknell.edu/mastascu/eLessonsHTML/Sensors/TempLM35.html

That's cool.
Yes, I think it could be powered by the hall supply so you only need to stuff one more wire. You still need a calibrated voltmeter to read it.

With an off-the-shelf digital thermometer you could share one if it's wires with an existing hall wire (like the ground) so you would only need to add one more wire.

 

[Mathurin]

Ok, how about this (from here):


Temperature Rise by Resistance Method

Degrees C Rise = Rh – Rc/ Rc (234.5 + T)

Where Rc = Cold Winding Resistance in Ohms

R h = Hot Winding Resistance in Ohms

T = Cold (ambient) Temperature in Degrees Centigrade

Note: This formula assumes that the ambient temperature does not change during the test.

Example: A small motor has a cold temperature of 3.2 ohms at 25° C (77° F) ambient temperature. After operating at full load for several hours, the resistance measures 4.1 ohms and the ambient has increased to 28° C.

Calculate the temperature rise:

Apparent rise = 4.1 – 3.2/3.2 (234.5 + 25) = 73° C

Correcting for 3° C increase in ambient:

Actual rise = 73° – 3° = 70° C

 

[fechter]

Hmmm....
Something like a DrainBrain, which is already able to measure current and voltage, could calculate temperature based on resistance.

It's difficult to make precision resistance measurements while under varying conditions, and things like leadwire resistance become issues.

With really good software, it could self-correct for lots of things.

 

[Mathurin]

Allright, guess it's gonna be a LM35CZ-ND then.

I figure the Clyte motor probably has a max temp in the ~85-100°C range.

 

[roller]

I'm a fan of the thermistor method. they are small, you can power them off halls, and they are quite accurate.

the app notes on the part (there are many) should give you a simple formula for V and T relationship. plot a little spread sheet and program the limits into your brain. when you see x voltage, you know you're hot.

100C should give plenty of head room. copper insulation starts smoking above 150C or so. the trick with smaller wire is that it heats so fast, and the rate of increase is more the bear than the temp. If you're bombing a big hill and see dangerous temp, even if you stop right away, the temp will still rise for a bit.

also, you can only measure one small region, and not every pole, or the hottest part (like that one mm of wire that scraped agianst the stator during build, and knicked a kink in it for higher resistance).

 

[Drunkskunk]

Thermisistor sounds like a great idea

But why export the temp? and why power off the Halls?

Instead, you have a huge sorce of spinning magnets and wire coils to chose from. a thermisistor needs Almost No currant, so a few coils of bell wire and it could generate it's own power localy from waste emf. a Ziener diod and a capacitor for voltage regulation, and instead of exporting the temp, have it shunt a hall sensor when it goes over temp. that would kill your motor untill the capacitor bled off.

Built right, it could either be a hard off untill the motor cools down, or could be built to cause the motor to cut in and out and allow you to limp along untill the motor cooled.

 

[Mathurin]

I can't seem to determine what class of insulation is used in Xlyte 4X motor's windings, the lowest ratings I've found were on small toy motors at 85c, but most motors seem to be rated for 100+. I've read that one should keep a 10 degree buffer for hot spots in the motor's windings. So with a winding temp of 85c, one should assume there would be 95c hot spots, and a motor at this temperature should be safe. However, I figure the windings should be able to take at least 100c without issues.

I don't know what temp the magnets used are rated for, all I can tell about the 4X motors is they have NdFeB magnets. The weakest ones have a Currie Temperature of 80c, better ones 130-ish. That is the temperature where it looses it's magnetism, permanently. In a hub motor, magnets get their heat sinked by virtue of being glued onto the outer drum thing, in my experience that part stays cool so I figure they should be safe.

That's how I figured 85-100c for max temp.


Power off the hall sensors seems convenient, and I was thinking of having an analog meter on the handlebar with a red zone, I've not figured how to do that yet though.