Is my MAC 8T temp sensor lying to me?

So I have a MAC 8T running at 48V on 26" wheels. I use it to get to/from work. This is my first ebike build and I did a lot of researching of components and never before did I read these motors overheat easily.

ANYWAYS my route is approx. 15 minutes of nearly constant WOT @ 31MPH. With the Cycle analyst displaying the temps/regulating over 90C-110C I am see temps in the 80's as soon as 8 minutes into riding. I pull a continuous 1200-1300watts at speed. At the end of riding the motor peaks at 110C and within a minute is back down under 100C. Touching the motor it feels BARELY lukewarm, this doesn't seem logical. Is the motor really this greatly insulated? I am tempted to disassemble the motor while it's hot to verify internal temps at the coils...

That sounds like the correct temps to me.

The core of the stator is a fat chunk of aluminum, but it has a dry fit onto a steel shaft. The connection (core/shaft) is small in area, and steel does not conduct heat as well as aluminum. The steel stator laminations are the hot part, and the aluminum core might absorb a short spike, but...eventually the heat has to go somewhere.

The aluminum sideplates are connected to the shaft through bearings, so that is a thermal wall. Air separates the stator from everything around its tip and sides.

I've had smoke belch out from a motor you could put your hand on. 5 min later, you couldn't put your hand on it.

Amazing how fast that copper heats up eh? Faster than it can pour the heat into the case.

m52 power! said:

So I have a MAC 8T ... Is the motor really this greatly insulated?

Click to expand...

In case SM's explanation wasn't clear - the 'motor' for the MAC is much smaller and inside the case separated by air from the hub you see. It drives the outer hub via gears. This is not the same as a DD motor where the hub actually is the motor.

So - when you touch the hub, you are just feeling the temp of a big box in which the real motor is running... sort of the difference of putting your hand on the hood of the car vs the engine block.

Right. FWIW, the motor I fried while the cover was cool was also a gear motor. So the covers take longer to warm.

The ONLY place to judge the heat on a hubmotor from the outside, is the axle. But even there, you can melt a motor before the axle heats enough to be noticed on a thermometer sensor. Axle temp lags stator temp by at least 5 min.

And you can heat a motor up good in 5 min, if you try.

I regularly ran bursts of over 3-4 kw from my 8T, I climbed steep hills etc etc, windings still looked good, I did damage the clutch though and it never went back together properly so I retired it. That was the old clutch, I suspect the keyway is the next weak link.

That mac is amazing how much power it can produce !

I suppose you'd have to worry more about the magnets than the windings. So I think the safe limit is 140 deg C ?

I couldn't imagine overheating it on continuous 1200 watts.

The new clutch on the mac high power kit should tolerate a lot more, but boy is that mac fun and pretty efficient too at sane speeds.

You mention 31-MPH at WOT on hills. I wouldn't recommend that. Its winter right now, and cold air might be helping "just enough" that it doesn't quite fry. Fast forward to the middle of summer and another 10-degrees Celsius can make all the difference.

I would swap-in a 10T (26-MPH @ 48V?), and get a CycleAnalyst to roll-back the amps when it starts getting hot. For a fairly regular 31-MPH on flat land, get a Direct-Drive (DD) hub in a smaller wheel (24-inch/20-inch?), which has more copper mass.

For significant hills, consider a mid-drive, maybe an upgraded GNG from LightningRods.

If you plan to beat on the 8T MAC until it dies, maybe dis-assemble it and ventilate the side-plates with holes to let some of the heat out.

spinningmagnets said:

...maybe dis-assemble it and ventilate the side-plates with holes to let some of the heat out.

Click to expand...

I would not do this - lets heat out but also lets grit in that will be captured in the grease, will wear the gears, and get into the sprag clutch and clutch bearing (open - no seals)...

I should mention I live in FL and the entire route is quite flat. There is a very mild bridge I go across for around a minute but I bet most of you guys have small hills that are more of a grade than this bridge....

Anyways I do have the cycle analyst setup to limit amperage if it goes over 90C and is set to off at 110C.

If it really is safe to operate up to 140C then I will change that but can anyone verify that?

It really stinks because the last 5 minutes of my journey to work are on reduced power because of the temps so I can't keep pace with traffic (25-30mph streets!) so I end up pulling off onto the sidewalk.

I'm already looking at torque raptors and cromotors so as my first bike this will be a stepping stone as with most of my hobbies (don't even get me started on R/C helis... ).

lets heat out but also lets grit in

Click to expand...

I forgot to mention this, you are right. This is a vital consideration that must be weighed out before cutting metal.

The Cro has a 145mm/155mm drop-out spacing, and has been proven multiple times to pull 50-MPH regularly at 72V / 83A = 6,000W (8-HP).

A step that's between the two (48V MAC vs 72V Cro) might be a Crystalyte HT35XX @ 72V X 42A = 3,000W (4-HP)?

Justin's simulator is set up to consider 'overheated' to be 150degC which considering the source, I would take on faith to be a good threshold. The simulator uses a simple thermal model and assumes still air so in real life things should be a bit rosier when you are wailing down the road with cool breezes washing over the hub.

Looking at BMC V2S data in the simulator (similar to a MAC), it appears that the motor will not begin to overheat (i.e. begin to accumulate heat faster than it can be dissipated - so it will eventually reach 150degC) until it is generating around 340W of waste heat.

• 350W of heat takes about 35min to overheat,
• 385W takes 15min,
• 550W takes 5min.

The eZee V2 is about the same - it overheats with slightly less waste heat - the MAC is likely in the same neighborhood as these motors.

Looking at the factory data shows that running a MAC 8T WOT should give about 31.5 mph estimated speed and be using 1300W of battery power. It will be dumping about 250W of that as heat. So, on the flat, it should pretty much run forever without overheating.

But - with a small 3mph headwind, the speed drops to 29.5mph and the waste heat runs up to ~310W - just on the threshold of 'heading to overheating'. A small 3% grade (no headwind) drops the speed to 28.7mph and the waste heat goes up to about 350W. I think this matches your experience fairly closely.

So - it looks like you are right on the edge of 'beginning to overheat' the motor, but frankly, your waste heat production is very low and the motor will shed heat faster as the temp differential to ambient temperature increases. I think you need to raise your V3 temp limits to something like 100/135 or 95/140. This will allow the motor to heat up more, shed heat a bit faster, and still stay away from the 150degC limit. With a bit of luck, this will let the motor temp reach equilibrium or at least only creep upward slowly so you can complete your trip at the speeds you wish.

High ambient temps can foul up this plan, but things look like you are just being a bit too conservative and a little leeway will get the job done.

If you have reprogrammed the controller, be sure you are not running with speed settings over 99% - these are going to expend a bunch of power as heat in exchange for little if any speed increase. Also, running at 30.5mph with a 26" wheel, your motor is doing just about 400rpm - pretty much the max before the speed pushes you into inefficient running, so monkeying with the Infineon 'overdrive' speeds (>100%) can only hurt your heat situation.

So - this advice is based on snooping the data, not real experience, but it looks promising - might be worth a try....

If things don't work out - throw a second 8T on the bike and split the load and heat dissipation across two motors. I run the equivalent of two MAC 8Ts at 64v and the bike flashes by 30mph without trying...

Ah, I also forgot about the 2 hub option running 2WD. Teklektik has experience doing this. If running one geared hub, a MAC is my favorite suggestion (nothing wrong with eZee/BMC, both excellent products). But for the front fork, you could save a few bucks by considering the Bafang-BPM as an option.

The stator/magnets are about 25% narrower on the BPM compared to MAC/eZee/BMC (17mm vs 22mm), but if you only need to add 75% more copper mass to your system to put you back into a very safe heat zone for the amount of watts you like to use, the BPM is an affordable experiment. Definitely put the weaker hub on the front and use a strong torque-arm...

you might give Paul a email. Ask what he suggests as a max temp.

Where abouts in Florida? What's the temps lately?
We just went above freezing, two days in a row

Dan

Oups did not realise that.

I have a MAC 7T running on a 12S 15Amp/hour lipo pack. http://endless-sphere.com/forums/viewtopic.php?f=6&t=55647
I ride about 17 miles to get to work and a lot of the commute is at WOT. A few times the casing of my motor must have been at 125 degrees... I guess I'm lucky it did not smoke!!!

How ever I did notice that, at WOT now, after keeping it there for about 20-30 seconds, it gets more whinny. Then ,if I release the throttle and WOT again, I'm good for another say 30 seconds without the louder whine. If I remember, this started happening after runs when it got hot???

I'm getting a CA V3 so the max temp is 150 then??

Thanks

spinningmagnets said:

Ah, I also forgot about the 2 hub option running 2WD. Teklektik has experience doing this. If running one geared hub, a MAC is my favorite suggestion (nothing wrong with eZee/BMC, both excellent products). But for the front fork, you could save a few bucks by considering the Bafang-BPM as an option.

Click to expand...

Ya - but the advantage to running two identical motors is a simplified throttle/CA setup and exactly matched loading on the two motors. This simplification is (IMHO) more than worth the savings of getting a less expensive but dissimilar motor.

With a V3, this lets you monitor just one and 'assume' the other is roughly the same temp (CA can only monitor one thermistor). Using an external shunt, you can run both motors with the same V3 using a single throttle and any throttle mode (current, power, passthru,...) and the CA won't know it is running two motors since the shunt will show the total current and the two identical motors with each draw about half the CA-regulated current... Things get trickier when the two motors have different Kv and power/rpm ranges - too much dynamic twiddling to share the load well. I have my stock CA V2 45A shunt clamped to an aluminum plate and run 60+A with no problem - the shunt is small and off-the-shelf - no trickery needed.

You should be able to dump 1.5kW through the fork w/o difficulty. Of course, the trick is that since the motors are only seeing half the load, they don't need as much power - to go the required 30mph in this case means each motor will only need to provide half the power instead of one MAC pushing it all. With the same efficiency at lower power levels, the combined battery power remains roughly the original 1300W. So in this case the fork would be bearing a very modest power load (equivalent to about 650W battery power - which at 80% efficiency is ~525W motor power to the ground). Of course, fun-filled jackrabbit starts at full power will certainly press it harder....

RC_guy said:

I'm getting a CA V3 so the max temp is 150 then??

Click to expand...

The max junction temp for silicon is around 175degC, so 150degC seems a safe margin without baking the halls - which seems the real issue - not melting the varnish. I'm sure there are anecdotes about running to 200C but component longevity seems a question.

As I mentioned above - the 150 figure came from the simulator documentation and appears to make sense, but this is not an area where I have expertise...

m52 power! said:

I should mention I live in FL and the entire route is quite flat. There is a very mild bridge I go across for around a minute but I bet most of you guys have small hills that are more of a grade than this bridge....

Click to expand...

I wonder why you need 1300W on a flat area to maintain a speed of "only" 31mph. To me this sounds like bad air resistance and/or bad rolling resistance for a "bike". Maybe some pedaling would also help.

Why not improve the bike part and lower the resistance?

Cephalotus said:

I wonder why you need 1300W on a flat area to maintain a speed of "only" 31mph. To me this sounds like bad air resistance and/or bad rolling resistance for a "bike". Maybe some pedaling would also help.

Why not improve the bike part and lower the resistance?

Click to expand...

This is a reasonable figure for a MTN bike (or any bike with a slight 'lean forward' posture), semi-fat (ebike) tires (Cr of .008), a combined bike+rider weight around 235lbs, in still air on 0% grade, with an 80% efficient motor. The actual expected power to push the bike is about 1000W - the difference goes to heat in the motor. Unless he wants to get into a tuck position, the aero drag is not going to change much - which is the biggest loss (CdA of around 0.58).

Since he's on the edge of overheating, swapping out the crankset so pedaling is more useful at 30mph would help but he can probably only contribute 185W or so if he works at it - so if it's a commute to work in FL and he doesn't want to arrive at work dripping, the rider contribution may have to be somewhat more modest.

Remember that the CA parameters bracket the 0% and 100% rollback settings, so adjusting the upper limit to 140degC means the motor will be getting no power at 140. The idea is to allow the motor to reach a more workable equilibrium of heat generation, heat dissipation, and CA rollback so maybe the temp gets no higher than 125degC or so.

teklektik said:

This is a reasonable figure for a MTN bike (or any bike with a slight 'lean forward' posture), semi-fat (ebike) tires (Cr of .008), a combined bike+rider weight around 235lbs, in still air on 0% grade, with an 80% efficient motor. The actual expected power to push the bike is about 1000W - the difference goes to heat in the motor. Unless he wants to get into a tuck position, the aero drag is not going to change much - which is the biggest loss (CdA of around 0.58).

Click to expand...

Even clothes can make a huge differnce at that speed. I don't have a picture of the bike, maybe even minor modifications will help a lot. But lt's only assume 100W less resistace,

Also different tires with higher air pressure could save you 100W vs. MTB tires with low pressure (which is of no benefit on termac roads)

Add another lousy 100W from you own pedal power and you end with 1000W-100W-100W-100W = 700W. At 80% efficiency (the efficiency will also increase in that case) this makes just 875W from the motor and 875W vs the former 1200-1300W can make ALL the difference that is needed. No chances in the motor-drive train, just more efficient biking. Also only 2/3rds of the battery consumption, so maybe 2 times longer battery lifetime (less consumption plus you can take more degradation before a replacemnt is needed)

I don't have a ebike that is able to travel 31mph = 50km/h. My BionX bikes with 48V batteries are able to do 45km/h (if the speed limit is off) with some own pedaling. (maybe 150-200W?) Battery consumption is around 500W at that speed, but I pedal and I use a modestly efficient position and my bikes have low rolling resistance, because sometimes I drive them without a motor, so I can feel it, if you lose power somewhere.

With an aerodynamic raiding position on a road bike and some significant own effort 45km/h should be doable with only 200W motor power, 50km/h maybe with 300-400W motor power, but this is an entirely different riding experience and needs an entire different bike which not everyone likes.

It's also just a suggestion to also have a look at the resistance and not only at the motor if you want to go fast without overheating. Minor changes can make a huge difference.

after doing the mid motor conversion for my mac i must say i'm surprised that these motor can handle even 1500w that good. as spinningmagents said: there is almost no real thermal path from the windings to the case/cover. all heat is transfered through the air and through the small bearings. totally different from a DD motor - which eg a hs3540 are easy to smoke too.
i think a temp sensor is a must for these motors and i would not change CA's default values. you are just pushing the end of your drive for some seconds. once the motor really heats up it takes very long to cool down again. so not letting it get too hot at all may the best way.
florida is great for any electric vehicle anyway. like the netherlands. no hills that kill the motors. and totally different from austria with the alps - the reason why i went for mid drive for my next ride now.