Death of a Controller (the biggest thing to avoid)

John in CR

100 TW
Joined
May 19, 2008
Messages
14,954
Location
Paradise
Last night I blew a controller that's been proven for 5 years at the same settings. With my newest rig I swapped over the dual Zombiess 34fet high voltage controller and the ventilated HubMonster to my new rig. Though the bike weighs more than my old SuperV, I've lost enough weight to compensate for about the same all up load. The new controller location gets more airflow, so the controllers should run a bit cooler despite the 6% larger wheel and 3% higher pack voltage.

The first week of daily use showed that the controllers were happier, ie not as hot as before. I even took my 100lb son for some long fun rides that included my daily 8-10% grades. Last night we get the bright idea to head up to the wind turbines on the mountain peaks near my house to look at the stars away from city lights. The climb is several kilometers of continuous 20% grade that they built for the trucks to get the turbine parts up there. I've done the climb alone on hot days, so I thought the drive system would be fine. Well it might have been, but it's been so long since I've popped a controller that my number 1 rule slipped my mind.

That is avoid low percentage throttle positions for sustained period with heavy loads, especially steep hills. I've got a box of blown controllers from 6=10 years ago blown from running in the lower third of the speed range during hill climbs. I've even blown a controller at under 5mph on level but very bumpy road due to the same issue, which is short spikes in phase current. While easing up a hill can be easier on a motor, it can mean a quick death of a controller. At low percentage throttle, phase currents spike and overshoot the set phase current limits.

When I made the climb before it was daytime, and I really attacked the climb accelerating hard out of each switchback...accelerating to 80kph on the short somewhat straight stretches. Last night I didn't attack the mountain and now I'm paying the price. Thank goodness I had the foresight to order a spare when I bought the controllers from Zombiess 6 years ago. Hopefully last night was the last time I forget to attack a climb.

One good thing to come of it is to speed up my soon to come off-road beast. I'll run nearly the same power, 25-30kw, into a ventilated HubMonster, but it will be geared down from a top speed of 160-180kph to only about 100kph. That gear reduction will make it so there's no stress on the drive system no matter what speed I ride up hills
 
Motor temp was only 71°C , but I didn't think to feel the controllers immediately. It was Saturday night, so there was slow moving traffic and pedestrians that kept me pretty slow for over 2km of uphill before I got to the meat of the climb. I've been caught going slower than I like on ascents in the 5-10% grade range numerous times in recent years, and I became complacent about the low throttle climb risk to controllers. 5 years ago when I was paranoid low speed during climbs, I would have pulled over and checked everything and let it cool before the really steep stuff. With a bike so friggin fast it's hard to avoid those low % throttle occasions, and the complacency cost me.

Once I explore and document the true top speed, I'll probably decrease the voltage, since I simply don't use those upper speeds. That will help. I'll also go back to the smaller tire size, since I can feel the difference in acceleration, and that will help even more.

Another option I've used with good results in the past is to add fans to force ventilation of controllers. I almost put fans 5 years ago when they consistently ran quite warm, but they just kept on going. I need to fab a rainy season worthy belly cover for this bike first though. Cooling a controller from the inside works great, but I can't be sucking in water or magnetic dust, which has been in abundance for the the past year with a local volcano going more active. That means I can't go with just intake holes and must add a filter. Something that really surprised me was the lack of any buildup of magnetic dust inside the motor after 5 years of running ventilated with blades. Wait till you see the smoke test video once I get my new cameras in.

Then of course there's the other solution I've been wanting to try, and that's to double battery capacity and add another motor. 8) I was saving that for a bigger and heavier registered moto build, but this bike rides far smoother with a passenger, so the extra battery weight would be a benefit. If I run the front motor at just 10 or 15kw peak input, then system stress on the rear drive would go way down, and I wouldn't even need to ventilate the front at that power at this voltage (115V nominal). The rear motor already stays cool, with the absolute highest temp I've seen in nearly 2 weeks of riding being 81°C a minute or so after shutdown, so I'm not worried about the motor at all. Controllers have always been my limitation. At some point I'll have to bite the bullet and get spendy on controllers.
 
So we take a simple low cost controller, and hop it up, put better FETS in, beef up the traces, rescale the current monitoring, and put more current through it than it was ever designed for. If it has enough capacity and isn't overstressed too much it lasts awhile, then one day the demands are too high, but it doesn't know that it should shut down to protect itself, so it fails, and leaves us possibly in a dangerous situation, perhaps shorting out and skidding the wheel, without power in a jam, etc.

This is not a good engineering approach.

Better to get a controller that has the sensors to monitor motor current, was designed from the start to handle the currents desired, and has temperature monitoring for both itself and the motor.

It won't be quite as cheap, it may not be as small, but it will be reliable.

Otherwise we're gambling every time we stress it.
 
Back
Top