2 burnt kelly controllers , anyone seen something like this

[poko]

Need some urgent help !!

i just burnt down 2 kelly controllers and i have no clue why.

1st try -
20s LIPO with Cromotor. all wiring seems totally fine , controller and motor worked fine . than after around 3-4 cycles on turn on and starting throttle a "poof" sound and controller is dead and motor is stuck.
controller sent to kelly and they claim it was a short circuit and a new controller was sent.

2nd try -
everything re-wired , now a 25s Setup , same motor with new controller.
worked fine for 3 cycles including high speeds.
than on start up and small throttle again "poof" sound (short circuit) and motor is stuck while in contact with controller (free when controller is disconnected).

opened up motor wiring loom and cover , no sign whatsoever of short circuits or bare wires. looks 100% ok.

controllers are KBL/96V (rated up to 120V).

i am clueless. anyone can help with an idea of what can it be ?

 

[peters]

Normally the controller should protect itself against short-circuit at the motor outputs.
Can you draw an interconnection diagram including all parts?
Kellys are configurable from a PC, did you check the settings?
When the motor is disconnected, the controller is still alive? There can be an error code displayed on the LEDs.

 

[dnmun]

full charge would be 105V DC so maybe it was too much for the mosfets or some other parts. if it is open can you check the mosfets to see if they shorted? that would help stick the motor.

 

[poko]

the setup is 100% fine. the first controllers lasted around 700km before burning up and the second for 30km or so.
this is around the 20th ebike i am building so i am quite confident in my wiring .
in any case after the first instance i went over all the wiring and re-built the pack , so should be fine .

as for voltage the controller is rated for 120v , so 105v should be within limits.

both instances were on startup . controller turned on , gentle throttle and bang , no controller.
the controllers and motor were cold when this happened .

any guesses ?

 

[riba2233]

Did you have the right kind of throttle set up?

 

[methods]

Take a picture of your setup that shows the wiring including controller, motor, throttle, etc. + any other components in the system.

Link back to your build thread.

Include more pertinent data... like a picture of your wiring diagram, which Kelly instruction set you followed, how you deviated, etc.

-methods

 

[Bluefang]

What voltage are you running for the power supply of the controller. I think for the 96V controllers you have to supply the 12-30V on the supply.

 

[methods]

I was thinking the same thing since it used to be that way back in the day...
Pulled up the datasheets for the model he was using and it was murky - unclear from the KBL zip I downloaded
The timing certainly points at that.

-methods

 

[poko]

Ok ,
I am attaching some photos to make all clear.
the wiring is really simple , a pack of 25S lipo connected to the + and - on the Kelly through a shunt for the display.
motor phases all correct .
lighting etc are through an external 3s Lipo
throttle is a magura 0-5Kohm and set correctly in the kelly software.


View attachment 1


you can see also a video from the time it was running . very easy to hear/see all is fine

http://www.youtube.com/watch?v=jgoS3z6SBjs

 

[methods]

Video wont show in this country.
Can you post it elsewhere? (I suppose I could spoof my IP but... got other stuff to do)

How loud was the "cogging sound" coming from the motor?
How hot did the motor get after a ride?
What is the freewheel current with the rear tire off the ground at full throttle? (answer this question)
- follow on - how was the CA calibrated?

How do you *know* the wiring is correct when there is no correlation between the Kelly phases, Cro phases, Kelly halls, and Cro Halls
(Unless Kelly controllers now "learn" combos?)

-methods

 

[poko]

Methods , For your questions :

1) for the video , i didn't upload it to youtube and i dont have the file. weird it's giving you location barring

2) as for cogging sound , let me explain better the situation -
after motor is cool and bike is lying around over night .
turn on the ignition (controller)
slight throttle.
a "short circuit" like sound (very short) from the motor side or controller side , not sure.
controller is dead.
for the first controller we opened it up before sending it to kelly , all A phase is melted down.

3) the motor never got really hot and on those two instances it was after several hours of non use in any case.
we also opened up the motor to see if there is no short circuit between phase wires , all look new and no apparent short circuits.

4) Freewheel current i can't remember exactly but it was few A , i dont remember if 2-3 or 5-7 at top speed, but those are roughly the numbers.

as for "knowing" the wiring is correct , it all seemed fine
low consumption on freewheel and on road driving (3-4 miles / AH on slow driving)
motor "sounded" correct phase wise
Speed matched suggested KV less traction (around 10% less than Freewheel) and matching ~9KV

so it all looked ok.

the question if it can be some hidden motor problem ?

 

[John in CR]

Sounds like both controllers came out of the same production batch and had the same problem.

 

[flua73]

What voltage are you running for the power supply of the controller. I think for the 96V controllers you have to supply the 12-30V on the supply.

Is it a kelly KBL 96 251 controller? if so, is the supply voltage range, PWR, 18V-120V. My own 96 251 controller is powered by 100 volts,

 

[hydro-one]

I run the same controller and motor . good to 119v. has pretty good error detection (at least with throttle shorts).the only times ive blown fets(not in the kelly thank the lord ) has been - , phases shorted to ground or each other in wiring, (axle entry ) or windings rubbing the cover. or overcurrent event caused by very cold system(-35c) and poor current control(infineon). also , dodgy phase connection that caused some kind of resonant pulse that killed a fet or two. also a bad hall or connection could possibly cause a mistiming and fet short (unlikely)

you dont need to supply the controller with 12 or 24v unless the drive voltage is less than 24v...... do you use a precharge?

have had nothing butgreat times with my kelly although i did just take both knees out due to power wheelie when throttle got salt water in it an hour ago lol

 

[Bluefang]

Do you have a precharge setup? Do you get a big spark when connecting the kelly to the battery.