oh the agony!

dirty_d

dirty_d

10 kW
Joined
Jun 16, 2007
Messages
985
Location
Massachusetts
i was out taking my new bike for a test ride today, while i was going around a corner i hear a popping noise and then i had no power. i looked it over and couldn't figure out what was wrong. i did some ninja type shit and held the brake while I connected the motor straight to two 12V batteries in series to get me home at a startling rate of 15 mph. when i got home i noticed part of my black wire conduit stuff was melted, i opened it up and saw that a connection had disconnected while i was taking that corner and the arc melted the plastic. i reattached everything and reattached all the wires. I was standing to the side of the bike and just touched the throttle too see if it worked now, the bike went full speed ahead and wouldn't stop with me holding it, right into a fence, i quickly disconnected the battery terminals. there was smoke shooting out of the holes in the controller! ouch. i took the controller apart and it looks like one of the FETs melted down somehow, two of the terminals on it are now non-existent and the solder around it is gone and on the bottom of the controller box. what the hell caused this? was it the motor wire disconnecting under power and the HV spike from the motor shutting off frying something? any way to fix this? im not sure of other things are broken besides the FET.
 
:arrow: Sounds painful...

I'm sure you'll get it figured out.

Those pictures suggest that your controller is screwed up. The easiest thing is to just buy a new one. I had a controller fail and just replaced it. Then in your spare time you can tinker with the old one and maybe modify it somehow to trick it out. (maybe create an MCL version or something or put in higher amp FET's)
 
Sure looks like a shorted FET to generate all of that heat. Do you have a fuse between the batteries and the controller?

What kind of connector were you using that came apart? A coil discharge from the motor could have created a pretty high inverse voltage spike.
 
fuse, what is a fuse? i guess you learn from your mistakes :oops:, the FETs are F1010Es and on the other side there are some STPS20S100CTs i don't know what that is, i looked it up it said some kind of rectifier, i guess they act like the diode in a buck converter, they all look fine. i looked up the F1010Es too it said they are 60V 84A 13 mOhm on resistance, i should replace those shouldn't i? maybe with a higher voltage and lower resistance? maybe ill throw another 12V battery in after that.
 
Geebee said:
On the plus side if it had been brushless you would have had to get home under your own power.
And brushed controller tend to be cheap.
Click to expand...

On the minus side, if it was brushless it would have just died instead of taking off on its own.
 
I'd be suprised if just a fast disconnect would melt your raceway. I'll wager the connecton was loose and heating way up before it came apart.

Which side of the controller did the connection come apart: batt or motor?
 
the motor connection came apart. i cleaned up the board with alcohol to see past the black, besides the gaping hole where the FET used to be(the board really just turned to black power underneath the FET) everything else looks fine, with the short removed i should be able to test the controller at low amperage, since the FETs are just all in parallel.
 
Something's goofy.

I suspect it has brain-damage from intermittant connecting to the motor under power.

It should have limited current before frying the FET, plus if you had a runaway condition after restoring connections, more than a bad FET is going on.

:?
 
i tried testing it, as soon as i connect the wires to the battery the motor runs, its not the throttle, i have all the hall wires disconnected.
 
Once the FETs decided to blow (and release their aroma), they usually short out, which will put your motor on full throttle, stuck on.

This can be dangerous.

It's hard to say what caused it to blow, but loose wiring anywhere can cause failures.

Brushed controllers are pretty cheap at TNC scooters and similar places. You might be better off just getting a new one. If you really liked the old one, you could replace the FETs (with better ones) and try to fix whatever else failed. Sometimes this can be a pain, if the drivers blew too. If you have an oscilloscope, you could see if the drivers are still intact after removing the fried silicon.

I was seriously injured by a blown FET on a brushed motor system. With a brushless motor, when a FET blows, the motor just stops or goes really slow and starts smoking.

Having some kind of safety cutout is a good idea. A kill switch and a big relay. With my precharging circuit, you can't activate the relay if you have a blown FET.
 
how hard would it be to make a new controller? they make PWM driver chips dont they, it should be a lot cheaper than buying a new controller right?
 
It would be hard to buy all the part to make your own controller for less than you can get one of those TNC controllers for.

If you're really cheap (like me), then I would look into fixing and upgrading the old controller. You need new FETs for sure, and probably the gate driver resistors and driver transistors (or chip).

You can measure the continuity of the driver transistors with the diode check function on a mulitmeter to see if they're blown.
 
I have a couple of IRF3205 fets that have 110A 55V 8mohm.
I think they come from a 4840 controller

PM me if you are interested.

Doc
 
well, i removed the two remaining FETs and checked them with my multimeter, one was bad and the other was good. i soldered the good one back into place. i hooked the controller up to a little electric motor like the one in a toy rc car, i touched the throttle and it worked, i decided to see how high that little thing can rev and then it turned into a ball of plasma, great. i checked if everything was still ok by hooking a 120V light bulb to the motor leads, it stayed on with no throttle, i removed the FET and it was fried. hopefully nothing else was damaged AGAIN, and all i need are new FETs.
 
By the time, i learned that a fet can be partially damadged

So if two fets are in parallel and one is short, the next one should be short or partially affected.. and will short few time later...

if you suspect that one or more fets are partially damadged but that they work, REPLACE IT ANYWAY.. that will avoid you to spend more money on a second blow up when gate drivers, pwn ic, reguilator blow too because of the last blow up... :wink:

Doc
 
dirty_d said:
i think 55V is too low, on a freshly charged pack its probably 56V.
Click to expand...

I think Knoxie has pushed the rating a few volts w/ success. You could also add an R-C pre-connect to kill the initial spike. Sag could bring the V down into good range fast enough after that.
 
i don't mind buying new FETs but i don't know what kind i need, like gate voltage and whatnot id like them to be able to handle at least 72V in case i want to add more batteries in the future, lower resistance would be nice too. is there a way i can protect them from being fried again if something disconnects?
 
i think im going to get some of the STB120NF10T4 i wanted the 4110s but i cant find them anywhere for less than $6 each, still the STB120NF10T4 is better in every respect than the stock FETs on my controller. anyone disagree and think i should use something else?
 
xyster said:
Geebee said:
On the plus side if it had been brushless you would have had to get home under your own power.
And brushed controller tend to be cheap.
Click to expand...

On the minus side, if it was brushless it would have just died instead of taking off on its own.
Click to expand...

I'd rather kick-scoot the bike than have unintended acceleration.
 
anyone have any input on the replacement FETs, mainly im worried about the difference in capacitance from the stock ones?
 
Lowell said:
xyster said:
Geebee said:
On the plus side if it had been brushless you would have had to get home under your own power.
And brushed controller tend to be cheap.
Click to expand...

On the minus side, if it was brushless it would have just died instead of taking off on its own.
Click to expand...

I'd rather kick-scoot the bike than have unintended acceleration.
Click to expand...

Depends on the wattage 250w is pretty easy to control and a kill switch within reach is always good.
A bit late for the original poster but first runs (after repairs etc.) on all my bikes are done with the drive wheel off the ground or chain off, I knew there was a reason for my paranoia.
 
You must log in or register to reply here.

Similar threads

C
Smoke and sparks coming from controller. Trying to learn what happened so it doesn't happen again.
Replies
4
Views
1,320
fechter
fechter
C
Chimera”... Antique style E bike build, circa 1915...why and how
Replies
6
Views
975
classicalgas
C
bombadero
  • Question
Short/arc fault in cold weather on XT-90
2
Replies
27
Views
2,334
amberwolf
amberwolf
electroneg
New to LiFePo4 and New Member - Battery, charging and BMS Questions.
Replies
24
Views
2,444
john61ct
J
Ianhill
Kuberg freerider owners look here if you don't like fire.
Replies
18
Views
1,681
Ianhill
Ianhill
Back
Top