brushless etek problem

Hey guys,

I made a brushless etek pocket bike some time ago with the intention of racing it against the gassers. i did end up finishing it but it had a few problems relating to gearing and just general build quality! but the real death of it was the terrible performance.

the specs where

-Brushless etek
-Sevcon millipak controller
-80x cp2400 nicad cells so 24v at about 9ah or so
-total weight of the entire setup was about 70kg myself included.

now the top speed was great once you get it wound up (it didnt have much gearing on its side as it was geared to about 150km/h) but the low speed performance was terrible. to the degree it would barely move and just shudder violently. i thought a factor of the gearing.. so i changed the gearing for a top speed of about 50km/h and the accelleration was still quite terrible. if you slowly apply throttle it will take off just fine BUT if you crack it open from closed it will just shudder violently like taking off from a dead stop in your car while in 4th gear.

a good friend of mine also put the same setup (with 36v SLA batteries) into a 250cc road bike and experienced the same thing with very mediocre performance.

im just dont believe that a motor reported to be at least 15hp can perform so terribly. on the calibrator the current at take off is almost 300 amps so it *SEEMS* to get full current but what its doing with it i know not...

has anyone else had problems with the sevcon millpak? personally i think its the controller screwing the works up and i will be getting rid of it none the less probably for a kelly controller as i cant stand the vast looms of wiring you need to run the sevcon and its pretty large for a pocket bike. just a bit dissappointing to spend $800 on a setup that i could ride faster then on a pushbike ive seen plenty of vids of normal eteks and they seem to go like i want.

is there some specific setting in the controller maybe ive elluded? ive changed everything basically one at a time from stock settings and everything in between to no real avail..

heres some piccies anyway to exemplify dodgyness of the prototype!





heres some specs on the new version -

- 36v A123 cells (probably 9ah) with integrated charge ports to get it charged nice and quick
- Kellycontroller
- much smaller chain drive and gearing to 100km/h
- custom alloy frame

heres the new frame - all 100% made from CNC machined aluminium. was gunna be a new gasser but the EV takes priority!! :lol:



look foward to being a member of the forum too. thanks guys!

wow those are some big freakin wires, i dunno about the shuddering deal, sounds like the controller, or hall sensors or something.

Yeah, shuddering and poor accel are typically symptoms of either dead or mismatched hall sensors.

Also 90mph on a pocketbike = ROFLCOPTER! :lol:

mmm ill test the hall effect sensors then, i found a good article on that here.

hahah and yes 90mph was somewhat optimistic but that particular sprocket was $30 cheaper then the next size up :lol: i race the euro gas ones and they run up to 55mph or so. which isnt too bad.

wiring was also a tad overkill. i figured on 300 amps for 5 minutes. perhaps abit pessimistic there ahhah

90 isn't too far from the mark, actually. 90mph (~150kph) should take about 14kW. You'd need either a higher voltage system or more amperage to get there, but the motor should be able to take it for limited periods of time.

Is that 2AWG? That should handle 300A no problem. I think I knew of a guy who was putting 1,000A through them during acceleration.

What are those batteries rated for, anyway?

cool stuff, im gunna test the halls tonight but the batteries are fairly old school techonlogy, i just had them laying about from previous projects.

http://www.master-instruments.com.au/Data%20Sheets/NiCd%20Industrial/Sanyo%20CP-2400SCR.pdf

there quite tired but apparently still dumped over 300 amps pretty happily i seem to recall 100+ amps per cell somewhere.

but im sorely tempted to just get rid of the entire setup and get the etek-R or -RT just to simplify it and get something i know works. being at the mercy of a non servicable (or hackable!) controller just sucks.

Well I may have a solution for you (and forget sevcon tech help from what I've seen):

You need a calibrator to re-program (modify a bit really) your controller... I had to change menu elements 2.07 and 2.08 (PID speed settings) to get better response. I made the P gain very high (like 200 instead of 64), and the I gain to 100 (instead of 32).

You may also want to adjust menu items 2.01 thru 2.04 - they are acceleration delays. put them low, like 1.5 secs for most.

sweet thanks for the response! will give it a go

I think the voltage should be boosted to around 48v to really be seeing the peak amperage from your controller. Heck you might not be see more than 100 amps at that voltage.

Phase resistance on a brushless Etek is 10mΩ. 36V / 0.01 = it'll draw 3,600A.

id love to run more voltage but im really capped to 36v on account of the sevcon. i believe they have a 48v version but they clamp that down to 200amps so id doubt it to be any more beneficial. and size is a big constraight, i just cannot fit anymore batteries in until i swap to LiFe batteries which leads me to ask,

where can i get some info on the prismatic A123 (or whatever brand LiFe cells are in that form) cells? for my other projects ive been ripping apart dewalt packs for $140 a pop + the time to resolder them all by hand. something 36v and 10ah would probably be the ticked in a big prismatic block if such an animal exists!

thanks for the quick replys i appreciate it

Glen,

You may also want to play with menu items 1.01 through 1.04 (PLUG). These control how fast the controller actively brakes the motor in different situations. I think it would also control regen, but I have not tried this because I have a freewheel on my bike.

Also, menu item 0.01 is your current limit, check that it is not at 50Amps or something! You may also want to try changing menu element 12.03 from speed to torque mode control.

For a professional product, sevcon have not really impressed me. They haven't even tested their controller with anything other than the etek, and don't seem to even want to anytime soon. I had to find out everything for myself pretty much, but managed to get my 48V 5kW ecycle motor working in the end (and it seems to work even better than the etek, and is half the weight and quite abit smaller too). No cooling fan though...