how do you manage heat of 5kW into 3540

izeman

izeman

1 GW
Joined
Jun 21, 2011
Messages
5,132
Location
Vienna, Austria
out of curiousity: i had a thread opened regarding performance of a xclyte @2500w. many of you mentioned that you have to feed it load of watts to make it fun. someone mentioned 6kW etc ...
i assembled my bike today, and did some shunt modding in the controller and didn't recalibrate the shunt value of my CA-V3. so it was reading way to low AMP values, and cutting the power @40A did mean 65A in reality. which was 60A, as this was the maximum the controller was programmed to. with 80V this means almost 5kW of power.
i have a temp sensor installed. just next to where the phase windings enter the stator. about 1cm "after" the soldering to the phase wire which leave the motor.

i don't have a picture after install, but imagine it sitting near the center of the green circle, glued with temp resistant repair metal to the windings.

temp_sensor.jpg


so now to my question: pushing 5kW to the motor in a 20" rim temp raised faster then you could follow it. showing 16°C/60°F ambient when i turned the switch, it raised to 120°C/250°F (programmed high temp cut off of the CA) within a minute.
i have upgraded to 12awg phase wire, and they are barely warm where they connect to the controller.
the motor case is cold to the touch, but after coming down to a stop after some more minutes (with disconneted temp gauge) the case is still at 40°C/100°F after half an hour.

HOW ARE YOU ABLE TO RUN THIS MOTOR AT 6KW+ w/o instantly melting it's inner parts???

btw: even with the temp problem i can't get rid of that ebike grin :) this bike performs REALLY nice. hitting 68km/h / 42mph on the flat and pulling like a tractor *ggg*
 
Have you drilled the side covers?

6 KW continuous is a lot of power, even for a X5.

Unless you are going to liquid cool it, drilling the side covers to allow airflow is about all you can do re cooling the motor. But as I said, 6KW is a lot. I am sure others will disagree, but personally I have found that with my X5s, over 3-4KW you aren't getting much more performance, just creating more heat.
 
i agree philistine. i DON'T WANT to do 6kW. this was a rethorical question. or a question out of interest. using ebikes.ca's calculator shows that performance does not get better in the same degree power raises. only heat produced.
so my answer to your question: the motor is stock except upgraded phase wires and added heat probe.
i found 3kW to be the optimum. if you can't climb that mountain with 3kW - just don't do it :) trying it with 5kW MAY work, but it also most probably melt the motor.
 
izeman said:
i agree philistine. i DON'T WANT to do 6kW. this was a rethorical question. or a question out of interest. using ebikes.ca's calculator shows that performance does not get better in the same degree power raises. only heat produced.
so my answer to your question: the motor is stock except upgraded phase wires and added heat probe.
i found 3kW to be the optimum. if you can't climb that mountain with 3kW - just don't do it :) trying it with 5kW MAY work, but it also most probably melt the motor.
Click to expand...

if you let the stock pahse wire there sure you can melt them if you plan to go over 60A , the drilling hole is a great idea because i have that on my previous set up and never have any problem with my 9c motor running at 24S and 100A of current , the key is to let the motor cool down a bit when you do hard acceleration you must run the bike at lower speed but not doing a dead stop . the barbecue sensor must be a place cooler because the epoxy crack at high temperature so i glue it to the middle of the hub with piece of stainless string so if the glue fail the sensor still not moving .
 
nice info about what glue to use, i used epoxy :( ill have to keep an eye on that.. anyway heres some great info on cooling a hub..
http://endless-sphere.com/forums/viewtopic.php?f=2&t=48753
Justin tried diferent combos of holes and basically the best cooling effect can be had by having large holes on the perimiter of one plate and holes around the axle of the other side plate, causing a draft through the stator. fewer larger perimiter holes work better than lots of small ones (i thought it would be the other way around) . but if you dont want little rocks n sh!t getting in your motor then you'll have to go as large as you think is safe for you, 10mm (1/2 inch) is normal, some go 1 inch / 25mm.
 
pendragon8000 said:
nice info about what glue to use, i used epoxy :(
Click to expand...
there is some eopxy that can withstand higher temps than 150°C/300°F, but only very few.
i used (also some kind of epoxy) that is used to repair metal. it comes as a stick with within a stick, and you cut as much as you need, than knead it with your fingers until it gets warm, and it will become hard within 3min. after that you can file it, or drill it.
it can easily withstand temp above 300°C/600°C.
 
You manage the heat by stopping in time. I don't know those motors, but on 9c, anything over 3000w works for shit. Oh yeah you get more, but at the cost of crazy heat. Surely 5kw is similarly too much for the clytes.

Do the cooling holes of course, but mostly just don't carry enough battery to melt er down. Make yourself stop in time by running out of capacity. Definitely a cool down run at 500w is best compared to just stopping.
 
i just discovered that it was even 95A (or ~6800W) i fed into the 3540. a wonder that the controller did do it (12fet lyen).
and no wonder that temp rose very fast. a friend of mine that did a test drive was really impressed ("... a lot better than my 250cc scooter").
letting it run at 40A max (3000w) is ok. as long as you don't do it for very long up very speed hills (~15%)
 
Yes, keeping it down to 3000w is much more manageable. Obviously the motor is unlikely to pull all 3000w continuously. But it does take vigilance on hills.

Start looking at cromotors or something big if you want to really run big wattage in a hubmotor. Personally though, I don't like that heavy a wheel in a dirt ride.
 
i set the controller to 45A now. which is around 3000-3500W. runs nice, hits 65km/h (40mph+) on flat ground, but i miss this instant response to the throttle still which i have with my mac 10t @1500W. but i won't complain in this thread anymore, as i already have another one open about this exact "problem". it's just a nice reasonable fast bike, but not the untamable monster i was expecting :(
with 3000w i can run it all day long i guess. temp raises - yes. but quite slowly, and declines after the hill climb. also the limiting of the CA-V3 works really nice. i set it to start limiting at 80°C and cut the power at 120°C. i very nice gradual decline in power until it's cut to zero. well done justin ;)
 
good setup this CA system. wish i would have had that on my first CA!!. IMHO the 3540 cant do much over 12fet(45amps) at 60v(15s lipo) . 20s lipo with 12 fet will kill it on flat ground . as in burn the windings. the HT35 is MUCH more able to take more volts and make less heat. But a hot motor can be fun for one ride, till your windings fry. remember you cant monitor the temp inside the winding bundle, deep in the motor, where it will fail during a burst of power in a hot motor being driven to saturation. size your motor for your task and you wont make heat.

possibly the only way to cool this properly would be oil cooling
 
hydro-one said:
remember you cant monitor the temp inside the winding bundle, deep in the motor, where it will fail during a burst of power in a hot motor being driven to saturation.
Click to expand...
that sounds a bit worrying, shall we say if you have a 3kw limit and power ramping temp sensor it will be ok?
IMO yes it will. but also Hydro-one makes a good point about heat in the centre taking time to make its way to the sensor (I think thats what your saying)
 
I installed the temp probe directly in contact to the phase wire and winding connection. I guess this is where heat will be highest in shortest time. So before anything else gets hot, the probe will. I personally think that 120 is a good maximum temp. And 3kW will hold you pit of trouble if you watch your temp and don't overdo it (e.g. climbing step hills for half an hour with disconnected temp probe)
 
With a lot of caution, temperature probe, cooling holes and cold weather. Cooling holes doesn't stop it from getting hot but they help a lot cooling it down when you stop. You can see the difference easily with the probe. Also, might wanna put some kind of rust protection in snow and water with cooling holes if you don't want to end up like this. :oops:
 

Attachments

  • whole bike.jpg
    whole bike.jpg
    158.1 KB · Views: 1,342
  • rust.jpg
    rust.jpg
    182.7 KB · Views: 1,342
I should change the topic from 5kW to 3kW. Did my first test run over 20km and was quite happy with the bike. If, yes if, I could get all its power for the whole ride.
I started with a quote hefty climb 12% for 1.5km/1mi. After that temp was already at 100oC. I was easy at the throttle to let it cool down, but even after some km decline with some power added it didn't cool down but stayed at 100oC. Maybe 90 sometimes and then rose again to 120. I must not say that CA limited power to lower the temp. But temp was still creeping from the inside to the outside of the case.
Going home was a climb again, and I was home after 15 km. Power used was 24Wh/km. Average speed was 29 km m/h.
Temp inside was still 50oC as after 45min.
Any ideas what to do?
Justin's calculator says 7min, but the motor was at its limit after 2-3 min.
 
The simulator is good for a flat surface. Climbing a hill or with the wind facing you it's different. I always found that my motor was heating up faster then what the simulator was saying going up hill. But it's kind of hard to juge with precision the slope of a hill .
 
Wasn't there another thread from Doc from Canada filling the hub 1/3 full of transmission fluid to cool it? Or am I dreaming???
 
No. You are right. Oil cooling is quite simple. Take care that the side cases are sealed and fill some oil inside. Stop below the height of the axle or the oil will drip out where the cables leave the axle :)
This is a good idea to spread heat evenly and remove hot spots from the windings, but it will only help for some time. Over time the whole motor will be hot and heat has nowhere to go. There are no cooling fins or stuff, so the cooling surface is quite small compared to the mass of it.
It's also a very good way to avoid rust inside.
 
You must log in or register to reply here.

Similar threads

fja2
Bafang BBSHD How I repaired my controller and how you might do it!
Replies
8
Views
6,961
NiallDarwin
N
markz
What do the "Guru's" say?
2
Replies
31
Views
4,447
thunderstorm80
T
H
How to run Sunstar mid motors with Cycle Analyst V3
Replies
0
Views
1,201
handbike_electric
H
oldswamm
E-bike control system w/BMS, lights, controllers, chgr, etc.
Replies
7
Views
2,820
The Speaker Guy
T
Z
Do EB3XX controllers limit phase current? - Looks like it!
Replies
7
Views
1,452
gwhy!1
gwhy!1
Back
Top