Controller overheat - advice please

[bdteakle]

I'm new to this forum, but have learnt a lot from reading through the threads. You're all a great resource!

I recently bought a road-legal front hub motor kit - 200w here in Australia - and installed it on my Dahon Boardwalk folder. The kit is an EVO electric bike conversion kit, retailed by an Australian company. It uses a 36V 200w Suzhou Bafang geared brushless hub, an Ananda Drive controller, a 36V, 10Ah Lithium Phylion brand battery, and a generic thumb throttle with 3 battery charge indicator lights. I have no measurement equipment installed, and no idea of currents etc.. The hub motor is in a 20" wheel, which I chose to achieve higher torque and efficiency for hill climbing. The 20" wheel also means that the motor spins out at 25km/h, meaning I mostly travel by pedal power on the flat, and only power on when I drop to 25km/h or lower. This gives me a long range, and keeps me well exercised, while avoiding the extreme efforts on steep hills, and greatly increasing my uphill speed and average speed. It all installed easily and has gone well for over 300km - until yesterday.

My reason for buying an electric bike kit is to make it more easy and practical to cycle in the mountains where I live. Travelling from our nearest town and train station back home up the mountain is a trip of 27km, climbing 650m altitude, including a few km of 15% gradient. This is all doable, but no fun, by pedal power alone, especially with any load. I'm very happy with 200w (or thereabouts) of power assistance, along with my leg-power, and want to remain road legal, and to maintain a long range.

All has gone well, including one very successful trip up the mountain with the luxury of electric assistance, until yesterday when I was climbing the mountain for the second time. Just as I reached the steepest section, the power went off and lights went out. Motor was luke-warm, but controller was too hot to keep a hand on. After 10 minutes and switching the power off and on, the controller reset and would provide power, but cut out again after a few seconds of powering. I had a steep, hard, disappointing push home. It appears to me that the motor was going fine, but the controller couldn't handle the sustained current.

I haven't found the supplier to be helpful so far, so I'd like your advice.

I do realise I'm aiming for something different to many of you on this forum, who are going up for high power and speed, but I expect your knowledge will apply. I just want my system to be able to maintain its power for longer.

- Is it common for controllers to overheat on long hills?

- What can I do about it?

- A higher powered controller (can these be set so the motor doesn't use more power?)?

- Better heat sinking for the existing controller?

- What do you advise?

 

[Russell]

If the controller is in that rear bag get it out of there and mount it out in the open!

-R

 

[icecube57]

Yeah do what Russ said. My controller runs ice cold. I mount it under my seatpost rack. Some controllers are just heatmizers. I noticed that with my GM regen controller. Its always warm even when its not doing anything. Just my 2 cents.

 

[philf]

The choice of transistors used in the controller (and how many) can make a HUGE difference to how much heat is generated under load. I alway run my controllers "in the bag", and even after long ascents in the middle of summer have found they get merely "warm". Any idea what's inside that controller, FET-wise, and how many there are?

 

[dogman dan]

Yes, mount that thing where it can heatsink into the bikes frame some, and preferably where it gets some wind. Then you should be ok. The motor might be a bit hotter than you think inside, which increases the load on the contoller if it is real hot.

 

[bdteakle]

Thanks for the advice. As guessed, the controller was in the black battery bag on the back. It had a gauze window on one side, but overall was poorly ventilated.

I've followed your instructions, taken the controller out of the bag and clamped it to the rack with some pieces of aluminium extrusion, which I hope will help draw heat. Photo attached. I'll at least be easily able to feel how hot it is.

I've also photographed the insides of the Ananda controller, so that Philf and others can have a look. I presume the 6 components along one side with screws holding them to the box are the FETs? If so, I don't think they can be identified.

I hope to ride up the mountain again in the next couple of weeks, and hope the new arrangement gets me all the way home.

Are there cooler controllers which could do this?

Thanks from Bruce.

 

[Drunkskunk]

Getting that controller out in the air should make a big diffrence, although the way you have it clamped might actualy be insulating it more than letting air get to it. Aluminum is funny that way. Still, its better than in the bag.


Running a controller using 4110 fets would also run cooler, but you're looks extreamly impracticle to modify.

I solved a heating problem by running a Cycle Analyst compatable controller twice the size I needed, then using the Cycle analyst to limit it to the current I wanted.

 

[bdteakle]

For the record, I want to post a follow-up note.

It seems the battery was causing the trouble, and the controller never was the problem.

After the first couple of "lights out" breakdowns, the retailers sent me another controller. I mounted it outside the bag and it now it never gets more than lukewarm. However the bike became more and more prone to sudden failure, often when freshly charged.

I sent the battery back to the retailer, who "tested" it and informed me that I had permanently damaged the battery by going up hills which were too steep and drawing too high currents. They considered this damage to be beyond the scope of the warranty.

Before returning the battery I did some testing with a shunt and multimeter, and found that peak current was 14.8A at 10km/h. FYI table attached.

I was unhappy with the retailer's position: this was an unaltered kit installed as instructed, 14.8A for 10 minutes (my peak hill-climbing load) seems very modest for a 10Ah lithium battery, there had been no warning not to go up big hills, and I'd travelled only 400km on this rig. None of the kit components came with any specifications, despite my repeated requests. I told them all so and asked for my money back on the whole kit.

Finally they replaced the battery with a new one. The new battery is the same size box, but 36V 14Ah.

I strongly suspect that the old battery has a dud cell, which fails under load and trips the BMS. I haven't opened it up to find out yet, and I'm not sure if I could buy replacement Phylion cells. Any advice?

The bike is going perfectly now. I love it. But I will be a much better informed consumer when I set up the next bike.

 

[nineball]

Get another battery, and parallel it, get a CA to monitor and control it, get a higher amp controller from e-crazyman/keywin, or one of the guys you mods them here.

 

[dogman dan]

Yeah, a 20 amp controller on a 10 ah battery is a mismatch when you start climbing steep hills. The amp spikes would be well above 20 amps, and even 20 is 2c, which requires a better cell than what you had I bet.

One thing people don't allways consider too, is that the gearmotors can be really hot inside and still seem cool on the outer covers. Feel or take an infared reading of the axle to get a better idea of motor heat. When a motor gets hot climbing, It will draw more amps, increasing the strain on the battery. I bet you were hitting 3c on that thing, and the way you were using it, all the use was at 3c or more.

That'll kill a battery. A ca to limit amps, a smaller controller, or a bigger or higher c rate battery would be good. 10 ah headways? Bosh fatpacks?

 

[Russell]

Yeah, a 20 amp controller on a 10 ah battery is a mismatch when you start climbing steep hills. The amp spikes would be well above 20 amps, and even 20 is 2c, which requires a better cell than what you had I bet.

One thing people don't allways consider too, is that the gearmotors can be really hot inside and still seem cool on the outer covers. Feel or take an infared reading of the axle to get a better idea of motor heat. When a motor gets hot climbing, It will draw more amps, increasing the strain on the battery. I bet you were hitting 3c on that thing, and the way you were using it, all the use was at 3c or more.

That'll kill a battery. A ca to limit amps, a smaller controller, or a bigger or higher c rate battery would be good. 10 ah headways? Bosh fatpacks?

When my controllers hit their limits that's it they won't allow any more current from the battery.

-R

 

[dnmun]

do you know if your initial battery ever charged up completely? did you ever measure the voltage on the charger while it was finishing the charge?

ever measure the individual cells to see if it was balanced?

 

[bdteakle]

I never took the battery case off, so never checked individual cells. I didn't want warranty trouble by messing around with it. I also didn't measure charge voltage. I presumed the charger and BMS were regulating properly when the green light was coming on after a charging period. I never measured anything until I started having trouble. I'd paid a premium for a locally-retailed kit in the hope of having something already sorted-out which I could just plug and play. Now I understand that's not the way it works - now I'm buying a Cycle Analyst, buying pieces from China and asking questions on this (very useful) forum. I shall be taking the battery apart when I get a chance and testing cells.

Re discharge current, from my measurements with a simple shunt and meter, the controller is quite strict about limiting currents to 15A. If you put the brake on and throttle it at stall, you get 15.7A for 4s. Maximum current when motoring was 14.8A (at 10km/h), which is 1.5C for the 10Ah battery. At 15km/h (hill climbing speed for a much less steep hill than my big 15%, 2km hill) current is 14.5A (1.45C), and at 20 km/h (speed for a flat hill) it's 13A (1.3C). The motor spins out at about 25km/h. I have no idea what discharge current rating the Phylion battery has, but if 1.5C is more than this battery can sustain, I would consider the system to be badly specified.

Bruce.

 

[Russell]

I have no idea what discharge current rating the Phylion battery has, but if 1.5C is more than this battery can sustain, I would consider the system to be badly specified.

Bruce.

The Phylion battery most often included with E-bikes, including eZee bikes, is a 1C continuous / 2C peak battery.

http://www.xingheng.com.cn/en/product.asp?classid=1&pageno=1

This is a curve I pulled from their website some time ago when I was considering a Phylion from AmpedBikes.



The plot is for a 24V pack but the curve is whats important. You can see it isn't very happy at delivering a 2C rate.

-R

 

[bdteakle]

Thanks Russell, it's great to have some real information about the Phylion battery.

It appears that 2C is within the capabilities of the battery, but that it will deliver reduced Ah capacity, and faster V drop at 2C than .5C or 1C. Presumably 1.5C will produce a curve in between 1C and 2C, but there is no reason to think that this rate would result in death of the battery after 20 charges.

How do other Li batteries compare? I see some claims that Headway LiFePO4 are rated at 5C continuous, 10C intermittent (http://www.zeva.com.au/tech/headway/). DoctorBass's report elsewhere on this site states 10C continuous and 20C intermittent. I presume they will suffer similar losses of capacity and voltage at higher currents, but these are very promising compared to the Phylion.

I see that Headway are available in Australia from EMTB for a lower price than Phylion from E-vehicles.

Bruce.

 

[317537]

I found my rear hub axel nuts lose again yesterday. Damn!

I did them up real good too. This was wasting plenty of current with the rim rubbing up against the brake pads. Was wondering why it was not playing properly on the hills.

Installed a torque arm yesterday and see if that stops it moving around. I think this is whats happening.

Just a suggestion check your brakes and stuff if they are off centre or too tight this can create unessesary heat and drain the batts. This and a hill would be a real bad mix.