Purchased a completed E bike project (with failure :( )

Joined
Jun 18, 2016
Messages
78
Hi all

I recently purchased Mike's Fazer 600 conversion (see here https://endless-sphere.com/forums/viewtopic.php?t=87495)

The overall specs are:

Enertrac 602 Liquid Cooled Hub Motor
Kelly KLS8080 400 Amp Controller
Kelly HWZ 325w DCDC
Kelly Charger
13 Leaf modules, Bulk Charging to 107.9v

I was having great fun on it, for 2 days or so, until the controller released the magic smoke.
I am fairly good with electrics etc, but nothing like jumping in at the deep end!!

The failure scenario was as follows:

Power up the bike by powering the actuating a large safety switch in series with the battery pack and the controller. This allows the power to come into the controller via the precharge resistor across the contactor. I wait until I see pack voltage on the ammeter (powered at the controller terminals). I then pull out of a car park onto a 60mph road, so give it full acceleration. I reach approximately 40mph and see roughly 290amps on the meter, before I get an abrupt cutout of voltage and and very jerky stop.

I immediately hit the safey kill switch and asses the situation. I go to power the controller up again, which results in jerky movements of the powered wheel and lots of smoke from the controller.

Get the bike home and find that the controller is dead with no signs of life (even when fed by its 12v control circuitry only). If I feed it pack voltage, I just get motor jerks and smoke again!

I pull the controller and open it up to find this:



That got rather hot. I was hoping to attempt to repair it, but I don't think that will happen.....

At this point Kelly Controllers cannot really tell me why it failed. It would be nice to know the thoughts of other people so I can prevent it from happening in the future...

Thanks all!
 
Seems this has happened to this bike before:

Re: Fazer Conversion Completed
by mcress » May 15 2017 7:33am

So - major problem this weekend.

My Kelly KLS 80801 controller released the magic smoke. Just pulled away from a junction, and at about 20mph there was a series of loud pops followed by much smoke. I hit the emergency cutoff, but the controller is toast.

I have an email in to Fany at Kelly to see what we do now - only had 4 rides out since I finished the bike. Confidence not high in the Kelly controllers right now.

I'm hoping nothing has been damaged in my Enertrac hub motor, need to find out how to test the windings / halls if anyone has any info?

Cheers,

Mike.
 
Yes indeed, Mike mentioned that he believed the first failure to be related to a precharge failure, or not waiting long enough for the caps to come up to voltage.

Kelly replaced the controller under warranty (admission of fault?) but they will not do the same with me, as its now 2 years old.

Peak amps is roughly 300 amps, and you only see it for approx 20 seconds, with this controller being rated at 400amps for 30 seconds. Max continuous current measured is around 100amps, with this being rated at 160, so I don't think its being taxed beyond its specs.
 
There is no BMS between the battery and the controller.
Or do you mean between the controller and the motor (IE a short between phases of the motor wires)

It goes

Battery > Safety switch > Contactor > Controller.

I have not tested the motor for damage yet, I would like to know how, if possible. I think I need to measure resistance between the phases but do not know the values to expect.
 
At least two things that can kill controllers like that, besides overvoltage spikes on the FETs, which is probably not the case here:

--motor phase shorts.
These can be intermittent, usually where insulation is damaged on either the phase wires or on the windings. This can result in problems under vibration, if the wires can move around. If they are secure, then they can result in problems as the wires heat up and are pressed together by expansion. They can also short under high voltage, if it's the windings. It may not be enough to directly destroy FETs, but it may cause glitches that the controller attempts to correct, which then causes spikes that do damage FETs. (or just overheatings on each power burst, eventually heating the FETs to avalanche point)

To test for shorts or phase-to-phase resistance, you have to disconnect the WYE point inside the motor, where all three windings are tied together. (if it's a delta motor, there are three interconnect points, where the phase wires connect to two phases at once.

--position sensor errors.
These can also be intermittent, but are usually constant, cuased by a design or manufacturing problem, especially with hall sensors. The slots in the stator laminations the sensors are usually placed in may be cut incorrectly, so they are offset relative ot each other or to the stator teeth, causing timing issues (the effect of which can be different depending on speed, power level, and which controller is being used, as some are more tolerant than others), or too deep, so the sensors could be unable to reliably read the magnets as they pass, so some phases may be triggered incorrectly or not at all. Other sensor types are less likely to have the same types of issues, and less likely to lead to catastrophic FET failures.
 
Your experience mirrors my experiences with Kelly controllers.

Be leery of any controller that you can make a motor run with no tables of PID current control loop tuning.
 
amberwolf said:
At least two things that can kill controllers like that, besides overvoltage spikes on the FETs, which is probably not the case here:

--motor phase shorts.


--position sensor errors.

Many thanks, I have emailed Enertrac, asking for the best way to test the windings in their motor, and to show me where the separation points are to check the wiring. Will see what they say. I will check all the windings up to the motor in detail as well, but from the initial check, the wiring seems OK.
 
liveforphysics said:
Your experience mirrors my experiences with Kelly controllers.

Be leery of any controller that you can make a motor run with no tables of PID current control loop tuning.

Any recommendations for an alternative controller, I am open to changing brand.
 
Back
Top