Hub motor internal temperature protection

[Hyena]

Hey guys,
We started talking about internal temperature cut outs over the the stealth owners thread HERE but it was getting off topic so I'm starting another thread here.

Obviously there's a few solutions, the ideal one being fitting an internal temp sensor that sets off an external warning or better still dials back the current through a CA or limits the throttle or what ever. What I'm aiming to do is have a safety cut out that's strictly inside the motor, the main reason being that the axles are often already cramped and don't allow room for additional wiring or if it does it's a fair undertaking.

What I'm looking at is interrupting the 5v supply to the halls using a thermostat or thermistor.
The negative arguement for this is that it could be dangerous to suddenly cut the power but my view is if you melt a phase wire or cook a winding etc the failure is going to be just as abrupt and then you're in for a long push home.

Aaanyway, the 1 option I've tested is a normally closed thermostat switch that goes open when the desired temperate is reached.
The downside to these is the reset temperature is often quite a bit lower than the trip temperature, so if you have to wait for the temperature to drop 20 degrees inside the motor that'd be a pain.

The other option is a thermistor wired in series with the 5v hall line. These are smaller and could be mounted direct to the windings and I've just done some tests which look like they'd work well.

I wired a 1k pot into the 5v hall supply from the controller
Voltage to the halls at rest is 4.3v
Increasing the resistance to 130 ohms causes a slight and consistent audible vibration. The motor maintains speed but if you were riding you'd certainly detect it.
Increasing to to 150 ohms and above causes harsh stuttering that quickly brings the motor to a stop
The voltage drop required to cause the vibration is 3.5v and it has a spack attack when it hits 3.4v. So it's probably not going to be possible using the output of a thermisor to get this mild vibration as an early warning. The increased resistance will cause a sudden shut of - which is all I was expecting anyway.
The second the resistance decreases and the voltage comes back up the motor continues to work as normal, so there's no need reset the controller (to answer your question full throttle )

Here's a quick video showing the test - further discussion and comments welcome


[youtube]uzY6EyanNck[/youtube]

 

[Alan B]

Have you tried interrupting only one hall signal?

 

[Hyena]

No, but it would have the same end result - ie like when you blow a hall or damage the wiring.
Either pulling down the whole 5v rail or the supply to any 1 sensor would do

 

[nechaus]

Smart Thinking!

 

[dogman dan]

The 24v Heinsmans had a similar feature, that cut power to the controller power on switch. I kind of hated it, and my first temp sensor was aimed at knowing when to stop before it happened. Such a long wait, in desert heat, to get it turned back on again. It definitely worked great, but I just hated getting suprised by the motor cutting off. You never know if it's the motor cutoff, or you just bounced the wiring loose, or maybe you just smoked the controller. 15 min later, Yay, it was just the cutoff. But I'd worry for 15 min.

In this case, when the motor stopped, it would freewheel, so it wasn't a big deal. With a dd, you'd get a bit of unexpected "regen brake effect" so it might take a bit of getting used to. I doubt it would toss you over the bars though.

Definitely I'd rather have the motor just stop running, than have phases short each other at say 40 mph! I don't like the idea of just cutting one hall signal, and having a huge stutter going at 40 mph! Cutting the halls power to all three sounds fine to me.

I just think that you do still want some kind of warning that it's about to happen. Perhaps a wire still needs to exit the motor, so you can still have a temp readout. You could do an external temp sensor, like I often do, but internal temps can rise so fast on a hot rod bike.

Having the backup of a cutoff that is entirely internal sounds real smart.

 

[Hugues]

... You never know if it's the motor cutoff, or you just bounced the wiring loose, or maybe you just smoked the controller. 15 min later, Yay, it was just the cutoff. But I'd worry for 15 min.
...

same impression here,
some kind of warning is needed, then shut off if no response at some higher treshold.

ideally:
i'd rather add an extra tiny wire, going to the CA, with a neat little display of the temp, with programmable treshold and alarm ? or why not, pulling down the amps automatically

i don't remember seeing much CA programming on the forum, is it possible at all ? there would be other things to add to that little box...

 

[Alan B]

No Wire Motor Temperature Measurement Techniques

1) make a circuit that goes in the motor that modulates the current the hall circuit draws based on motor temperature. Sense this current in the +5 lead with an external circuit. A simpler version could use a thermal switch and resistor to make the hall circuit draw more current, and detect this change to give a simple "overtemperature" output.

2) Measure the resistance of the motor copper. Easiest to do when the throttle is off. Update a display each time a reading is taken, and after a time fade the display out if no new reading has been made. To see a new reading simply let off the throttle momentarily.

3) transmit the temperature information with RF.

4) transmit the temperature information with a small AC current on the hall power lead.

 

[Spicerack]

I would love to have an internal temp sender sending signals to a CA which could both display the temp on screen and also reduce or cut power at preset levels. I think the biggest hurdle for that is there is simply no programming space left in the CA so you would have to lose some of the other functions. Justin went into that in the post about the rc ca.

I ended up putting the sensor in the hub just under where one of the phases comes off a tooth and the signal goes to a bar mounted display (the good old meat/BBQ thermometer mod).

 

[Hyena]

You never know if it's the motor cutoff, or you just bounced the wiring loose, or maybe you just smoked the controller. 15 min later, Yay, it was just the cutoff. But I'd worry for 15 min.

Yeah that's a point - though better to be put out for a few minutes on the side of the trail though than push a bike with melted phase wires all the way home, then have the hassle of trying to fix it and/or the expense of replacing it.

In this case, when the motor stopped, it would freewheel, so it wasn't a big deal. With a dd, you'd get a bit of unexpected "regen brake effect" so it might take a bit of getting used to. I doubt it would toss you over the bars though.

Yep there's an initial jolt then nothing. If you just interrupted one hall the symptoms would be the same as when a hall sensor dies (or a wire comes loose) I guess on the side of the trail you can check your hall wiring over and by the time you've finished all your checks hopefully the motor has cooled down and starts working again! Either way the second you let off the throttle the symptoms stop, so even if you were at high speed and felt the jolt/stutter it's just a matter of letting off the throttle and coasting to a stop as normal.

I just think that you do still want some kind of warning that it's about to happen. Perhaps a wire still needs to exit the motor, so you can still have a temp readout.

Yeah that's ideal, and if you're going to run wires there's certainly more options and better ways to do it.
I guess if I was really dedicated to the cause (or more so felt like "investing" in this) I'd sacrifice a motor to see at what temperature it failed. Put an internal temp sensor and watch it closely, slowly increasing the temperature until something failed. Technically the winding insulation is supposed to let go at what - 180oC ? Would a 160oC thermal cut out then be sufficient ? I'm not sure what temperatures they get to sealed as most of mine are vented but even 140 seems to hot to me. I guess it's a fine line between having it cut out on you often vs saving you that one time when you're abusing it and get carried away...

Then there's the other arguement that the magnets start to lose their mojo and much lower temps.

No Wire Motor Temperature Measurement Techniques
1) make a circuit that goes in the motor that modulates the current the hall circuit draws based on motor temperature. Sense this current in the +5 lead with an external circuit. A simpler version could use a thermal switch and resistor to make the hall circuit draw more current, and detect this change to give a simple "overtemperature" output.

That's not a bad idea! Based on this thinking, I think I could actually do it by measuring voltage drop rather than current draw, which is slightly easier to do given what I now have in mind. In my tests above I found that adding 130 ohms of resistance inline was enough to drop the voltage and stop the hall from working - but at 100 ohms it still worked. So maybe a thermostat that generates a 100 ohm drop - this would drop the hall voltage to 4v or less and could then sound a simple low voltage alarm across the hall supply at the controller.
This will be my next line of investigation - thanks

transmit the temperature information with RF

You think the components to generate this RF signal would withstand the heat ? It would need to be able to operate at over the 100oC it's going to be inside the motor.

That's what stopping me from using an internal buzzer - my other though was to use a normally open thermostat connected to the 5v hall line, connected to a piezo buzzer. When the desired temperature was reached the thermostat would conduct 5v to the buzzer and you'd get an audible warning - which you could ignore at your own risk or stop when it's safe to do so (if you're on the road for example)
You'd need a drilled hub for this to work though.
My other thought was that if the buzzer was placed close to the side cover below the air cooling vents it'd probably stay a fair bit cooler than where the thermostat is next to the windings. They're only rated to operate up to 70oC though so I'd probably be pushing my luck...

 

[GMUseless]

I think most of the ideas AlanB listed are pretty good. Especially since at 180°C it's much easier to get the electronics out to a cooler place. But, a thermocouple can use a pretty small wire after all. At a few feet it doesn't really matter what you use...as long as you calibrate it.

180oC ? Would a 160oC thermal....

By the way, something I never forgot from typing up engineering labs in Word Perfect 5.0....the escape sequence for the degrees symbol:

[ALT] 0176 (typing the numbers on the numeric keypad. Doesn't seem to work on the QWERTY numbers.)

Still works from DOS 5.1 all the way through Windows 7.

 

[Drunkskunk]

I live in a area where people like Pit bulls for pets. Some of the neighbors are into illegal dog fighting, but not really interested in fence maintenance, or remembering to close the door when they crawl home drunk. I've got a fast bike for that reason.

The last thing I'd want is my motor to decide it didn't want to burn it's self out for me while I was running from from a pack of dogs, or for that matter, jealous husbands. Or Cops. Or bill collectors. Or door to door Jehova's witnesses.

There are times when you need to GO, and you don't care if your motor blows apart as long as you get away.

I think I would want control of when my motor shuts down, or at least a warning, and a chance to over ride the shut down if needed. I've been in too many situations where having the motor decide to shut down right then would have been trouble.

 

[dogman dan]

Thinking some more, maybe just a cutoff that is fairly high would do it, provided the phase wiring through the axle wasn't particualrly vulnerable. I had one motor melt and stop because the solder melted and disconnected the phase wires when I hit a big bump. So somewhere just short of that temp for your shutdown whatever that is . At the death race, when smoke poured from my motor, a remote thermometer read about 450f inside it. In both cases the motor had burned away all the string that ties up the windings, so above 451F makes sense for both events. Maybe a sensor that cuts off, but not till 400F.

On that heinsmann, the sensor cut off a lot cooler, because the brushes needed it for that motor. So it would stop for the overheat trip fairly easy. You could overheat that motor fast.

 

[Alan B]

Unfortunately it is hard to know what temperature is safe.

We have a number of things that are heat sensitive:

Magnets
Hall sensors
Wire insulation
Epoxy
Wire retainers (string, etc)

Each of these elements is subject to slightly different temperatures, and can have significantly different thermal capability. Some magnets are rated only to 80C and will be weakened at temperatures above that. So the only "safe" temperature if we don't know all the actual ratings might be 80C. The highest it might be, if all materials are good, is probably the Hall sensors, around 160C if I remember right.

 

[Alan B]

I agree that it is better to let the Captain decide when to burn up the dilithium crystals.

Thermocouples require two wires and are not too easy to read out, so I looked for a 200C capable thermistor and found them for 35 cents at Digikey. We can use the ground wire for the hall sensors so only ONE additional conductor is needed. Can we get one more very small conductor out of the motor? Some hall sensor connectors have 6 pins, perhaps this can become a standard and we can agree on the thermistor so we have similar setups.

If we want to keep the electronics in the motor to a minimum the RF method is probably not a great choice. The temperature near the axle should be lower than near the windings so I would put electronics near there, the sensor would go on the windings to see the max temperature.

 

[Hugues]

... Can we get one more very small conductor out of the motor? ...

on a X'lyte HS / HT -> definitely

...

If we want to keep the electronics in the motor to a minimum the RF method is probably not a great choice. The temperature near the axle should be lower than near the windings so I would put electronics near there, the sensor would go on the windings to see the max temperature.

agreed, no RF needed for this. if the electronic is that small, one could have only the temp probe inside the hub and the rest outside, bundled with the anderson plugs and hall sensors plugs...

 

[Alan B]

The thermistor I have been looking at:

Vishay NTCLG100E2103JB

-40 to 200C
small glass encapsulated axial 2mm by 4mm

10K ohms at 25C
1256 ohms at 80C
339 ohms at 125C
145 ohms at 160C
95 ohms at 180C
64 ohms at 200C

There are other resistance values of this part, perhaps a higher value should be chosen. They to up to 220K at 25C.

These are 35 cents each in 10 quantity at Digikey. If folks agree that this is a good choice perhaps I should get a bunch and make them available in small quantity. It is easy to get these if you order stuff from Digikey, but the minimum order etc makes it less practical for folks who want just a couple of them.

 

[Hyena]

Alan if you do what I suggested and measure the voltage drop across the hall line you don't need to run any additional wires outside the hub.
I spent nearly 2 hrs yesterday going through different spec sheets trying to find the right thermistor and as yet haven't found one so I suspect I might have to do a bit of trial and error.

The one you posted is a NTC type - or negative temperature coefficient. The resistance starts out high and drops as the temperature rises. If you're using an external circuit to translate this into a temperature reading then that is fine, but if you want to drop the voltage in the manner I'm suggesting you need a PTC type - positive temperature coefficient that increases resistance with temperature. Aside from that the resistance is too high.

Most list the resistance at 25C but what you need to know is the resistance at just over 100C - because IMO that's when you want it to give you the warning.
As I said previously, 100 ohms will drop the hall voltage about 0.5v, enough to easily measure but not enough to stop the sensor working.
I reckon this is the best way to go - you'll get an audible warning at what ever voltage you nominate (determined by the type of thermistor you use) and then if you ignore it and the temperature keeps rising then eventually the resistance will rise enough to cause a voltage drop that will cause a hard cut to the hall line.

So yeah, lets hunt for a PTC thermistor that generates 100 ohms at 100 degrees or there abouts. Or even less - 60 ohms would probably go and we can run 2 in series. Given how they are typically quite non-linear in how the resistance increases we probably need one rated quite low at 25C (which is how they seem to be named) - maybe around 10 ohms ?

 

[full-throttle]

What he's saying is: the Halls consume 1mA (just for example) normally, when the thermal switch trips it could add a resistor across the power rail adding another 1mA. All you'd need to do then is to make a circuit on the controller's side to measure this current and react to the change.

 

[icecube57]

How come we cant use a thermal breaker like the ones used in nicad packs to just kill the whole 5v. It would break the connection and kill power to to the motor and to restore power would require the motor to cool down. Having a breaker thats around 120C-150c. This would leave the bike dead in the water but by you pulling the signal down with the resistor it seemed kinda harsh and abrupt .... have u rode with this implementation yet?

or better yet just have 2 small wires going going into the axle and use a thermresistor in or near the windings and wire that inline with the return voltage of the throttle. As it warms up ang gets critical it will increase in resistance to bring the throttle voltage down by a small amount to limit speed and in a round about way limit current giving the motor some im to cool. That way you arent left dead in the water and still have power to get some where to let the motor cool down.

 

[Alan B]

I selected the above thermistor earlier, it is not for your technique but instead for measurement as you suspected.

I have one question on your approach Hyena. If you connect the thermistor inside the motor in series with the power lead to the hall sensors, where can you measure the voltage at the node? The node you want to measure is inside the motor, no wires from that node leave the motor. The regulated 5V line coming from the controller should not be affected by this thermistor, it is a regulated voltage. In any case it would differ between controllers. You could measure the signal voltage coming back from the hall devices as his is a function of the node voltage, that is an interesting approach but would require the use of a peak reading meter as the hall signals are switching. Is this what you meant?

The next challenge is to find a thermistor that meets your other requirements and is small, cheap and works up to 200 degrees C.

 

[John in CR]

Hyena,

I'm in a different situation with unknown temp sensors already in the motor with wiring for the sensor coming out with the wiring harness. I need to figure out how to take advantage of it.

I like the idea of containing it in the motor due to lack of space on the route out. My only issue with it is that the motor will cool far quicker in low load use than sitting in the off position. I've always worried about roasting the mags and wiring by stopping, and now I always include a good half mile of easy riding at the end of any aggressive rides.

John

 

[itchynackers]

This is what I was planning on using... http://dkc3.digikey.com/PDF/US2011/P2225.pdf but I haven't installed it yet for reasons mentioned above. I believe it is better to coast or use low power rather than stop. I have found that after stopping, the internal motor temp rises for about 2 minutes, then slowly decreases.

 

[Hyena]

What he's saying is: the Halls consume 1mA (just for example) normally, when the thermal switch trips it could add a resistor across the power rail adding another 1mA.

OK that's a good option, that way I can simply switch in a known resistance when things get hot rather than trying to find an exact thermistor with the resistance I need. Although being able to watch the voltage drop by using the thermistors increasing resistance would be a good warning that things are creeping up.

How come we cant use a thermal breaker like the ones used in nicad packs to just kill the whole 5v. It would break the connection and kill power to to the motor and to restore power would require the motor to cool down.

This is what I suggested initially but it does cause a harsh cut off - if I you just pull down the hall voltage a little without cutting it out all together you can run a warning buzzer from this.

Having a breaker thats around 120C-150c. This would leave the bike dead in the water but by you pulling the signal down with the resistor it seemed kinda harsh and abrupt .... have u rode with this implementation yet?

Having a breaker yank the 5v line should result in exactly the same result as dropping the voltage down (or not having the hall plug connected at all for that matter)
I haven't ridden with it, only bench tested. But it should work fine.

If you connect the thermistor inside the motor in series with the power lead to the hall sensors, where can you measure the voltage at the node? The node you want to measure is inside the motor, no wires from that node leave the motor. The regulated 5V line coming from the controller should not be affected by this thermistor, it is a regulated voltage.

During my tests (video in first post) I was measuring the voltage drop at the 5v connector on the controller. I put it on cruise controlle so the motor was happily spinning on its own, then slowly increased the inline resistance in the +5v wire and the voltage dropped accordingly. So it can't be too regulated...
Adding the resistance 6" from the controller or 2 feet away inside the motor shouldn't make any difference.

Agreed, this will vary from controller to controller though.

I'm in a different situation with unknown temp sensors already in the motor with wiring for the sensor coming out with the wiring harness. I need to figure out how to take advantage of it.

Aren't those scooter motors the gift that keeps on giving
How many wires come off the temp sensor ? If they're a simple 2 wire thermistor you can connect a multimeter across the wires and see how the resistance varies and then use that to run a temp gauge circuit (or at the simplest learn what the resistance is when it's really hot and simply watch for that on the multimeter)

You're right though, riding slowly is better for cooling (especially with a drilled motor). On one one my bikes that does have an internal temp probe I notied the temp jumps a good 5 degrees when I get home and stop (on a hot day I threw a wet towel on it to cool it down and it sizzled and steamed!)

 

[amberwolf]

I came to the thread to suggest using the CA to limit power rather than cutting it off, but Hugues already has:

ideally:
i'd rather add an extra tiny wire, going to the CA, with a neat little display of the temp, with programmable treshold and alarm ? or why not, pulling down the amps automatically

i don't remember seeing much CA programming on the forum, is it possible at all ? there would be other things to add to that little box...

The code is not publicly available, AFAIK. At least one person has essentially made their own, though (heathyoung, with the "anal cyclist"). But Justin has stated that an in-progress version might have more inputs for things like thermal sensors and torque sensors, and whatnot (assuming it's not already done, or possibly partly available via the RC type CAs).

But yes, this is what I would want. Still, you can do a "thermally-scaled current-limit" even now if you like, by building an op-amp circuit that takes the thermal sensor analog output and scales it so you can then add it to the shunt voltage, and pass that on to the CA. This will result in incorrect A, Ah, and Wh readings, though, so an alternate method is proposed below it.

The positive shunt wire to the CA would be cut.
The first op-amp would read the thermal sensor, and possibly scale it (or use a second one to scale it). (variable scaling is needed because it is how you set how fast the thermal input begins to interfere with the normal operation of the system without compromising the actual current limiting you may already be using.)
Another op-amp would read the shunt voltage, simply buffering it.
Another would add the shunt voltage and the scaled thermal sensor voltage together.
The output of that would go to the CA on the positive shunt input wire.

The result of this is that as the temperature rises, the CA will read a higher current than is really there. The CA would be setup so that it can limit throttle input based on the current draw, so once the motor heats up to the point where even "normal" usage pushes it's reading over the limit, it'll throttle back to keep the current down so you don't keep pushing the motor too hard and destroy it.


Because the above method screws with the main purpose of the CA (keeping track of A/Ah/Wh), you could instead use the speedo input for this. It's just a bit more complicated to do the electronics for. Naturally, this sacrifices a different function--speed monitoring/distance.

Disconnect the CA speedo input from either the hall or the wheel sensor
Thermal sensor analog output goes to a variable scaling circuit (op-amp?) so you can alter the amount of effect the temperature has on the system, and how soon, without going into the CA settings all the time.
Scaling circuit goes into a voltage-to-frequency converter, which can probably made from the 555 timer. Basically you are emulating the "speed" of the wheel with the pulses from this.
The output of that is connected to the CA's speedo input.
Then set the CA for a maximum speed limit of whatever works for your scaling factors.

Bascially this will send pulses more often for higher temperatures, emulating a faster-spinning wheel. Once the pulses reach the specified rate, the CA will trim your throttle back to prevent further overheating. I don't know if this will work as well as doing it via the current-limiting, but it is at least an option.

You could probably OR the two signals together, so it can count wheel speed still, and can still limit actual system speed if you like, but it will not be accurate because it will be counting thermal sensor pulses, too.

 

[parabellum]

What if thermal breaker activates circuit that acts like blinker shutting off hall signal for short instant and restoring it for longer while, repeatedly doing that until temp breaker recovers on lower temp? Short 1 sec cutoff following by 10 sec normal operation or so.