OK I am stumped!

[cal3thousand]

you should never lock the bike in fixed/stall position so that the motor cannot turn and then apply full throttle. it will kill the hiside mosfets.

Yeah, I thought that was a strange testing routine too. Just dumping heat straight into the windings and stressing the FETs.

 

[dnmun]

no, it is because the hiside mosfet needs the charge pump to force voltage above the source leg of the hiside mosfet so it can turn on the hiside mosfet.

if the motor does not turn then there are no cycles for the charge pump to capture charge above the B+ voltage so the gate/source voltage will drop and as the mosfet drops into the transition region while a large current flows through the channel of the mosfet then the mosfet will overheat and blow up quickly.

this is not a problem for the loside mosfets which can obtain the gate voltage directly from the 12V rail.

so this is why the hiside mosfets are the ones that fail when the bike goes into stall and the motor is not turning.

 

[999zip999]

Thanks dnmun.
As I read that I had a cartoon picture of tge Inside of a controller and a mosfet working.

 

[docnjoj]

Thanks Dennis and Cal for the voices of reason.

Well it proved my point and didn't make the controller any worse since I only did it for 20 sec intervals. It showed 23 amps and then shut down. When I took the bike out for a test and it worked fine, that was long after the stupid "torture" test! However this was out of frustration and I defer to you folks and wont do it again. The bike ran fine. I brought it in when the rain started and when it stopped raining I turned on the controller rode the bike out of the house and it then didn't work. Anyways could all 5 controllers be bad? The other 4 never had this this done, just wheel in the air stuff. So I am back to square one. I am beginning to think that even though the tester shows halls OK and Phases OK that perhaps the motor could be the problem. I wish I had another working motor to test with.
otherDoc

I may have one. I just took my 2 speed apart and it has another failure of the sun gear for the low speed side. The high gear side is fine. I think I'll put it back together (minus the broken gear debris) and try it with only the high side working. At least it should go. We shall see. It has it's own controller, though so it won't prove anything but it will make the trike go.......maybe?

 

[whatever]

sensorless controller worked, but then didn't work: either a problem with phase wires, or controller failure.
Do you have a tester you can put on the controllers? see if the fets are turning on/off and 5v supply is ok?
You mentioned phase wires ran hot at one stage, I'm wondering if there is intermittent short in phase wires either inside motor or somewhere else? Shame we cant all come over to your place and sort it out!

 

[docnjoj]

I'm thinking motor problem, but I have enough problems for an entire team to come over and help!
otherDoc

 

[e-beach]

I don't know about anybody else ,but I don't think I will make it by today.... :lol:



But if it is raining and you have time before a new motor comes, well, it sounds like you might have time to do a diagnostic test on the motor.

Something like this: (and please stop me if you have already done this...)

Disconnect all connectors and visually examine them.
Using DDM check for shorts in phase wires - controller disconnected, DDM set for continuity - Neg probe to axle - Pos probe to connectors (one at a time) - with wheel suspended in air so you can test it at a stand still and spun by hand. See if the phase wires are shorted to the axle.

If that shows no sign of short, repeat the process but bend the wiring in a z pattern every 3 inches or so and wiggle the wires around to see if you have an intermittent short in your wiring.

 

[Jonathan in Hiram]

If that shows no sign of short, repeat the process but bend the wiring in a z pattern every 3 inches or so and wiggle the wires around to see if you have an intermittent short in your wiring.

In my experience an open is considerably more likely than a short in most types of wiring including ebikes. Shorts tend to cause catastrophic damage that's fairly easy to spot, opens on the other hand usually just don't work. I've never had a short on a ebike that I didn't cause myself with immediate obvious consequences (KFF to be exact) but I've had opens several times with no other damaged caused.

 

[whatever]

I have one ebike which has some sort of connection problem just inside the axle, a bit of a wiggle and bike is good to go,
annoying, but since its inside the axle, cant be bothered taking motor apart!

 

[docnjoj]

Thanks again, gentlemen. Lucky me, it is raining again today! Whoopie! At least I have time for more tests. Thanks.
otherDoc

 

[docnjoj]

I don't know about anybody else ,but I don't think I will make it by today.... :lol:



Aw shucks ebeach, that is just an afternoons ride! We have lots of extra bedrooms, so come on along!
otherDoc

 

[docnjoj]

Curiouser and curiouser? Now the Tester shows one bad Hall. I don't see why this should effect the sensorless controllers? The phases all show no resistance when the meter is checked from the phase wire to the axle even on the lowest "buzzer" resistance setting. So I guess this means that the axle is electrically separated from the wires, and that is good?
otherDoc

 

[dnmun]

zero resistance is shorted. can you post up a picture of the measurement you are making so we can see what you have the ohmeter connected to?

 

[docnjoj]

Sorry dnmun, my brain had a Senior moment. The meter read OL on all resistance settings. I guess that actually means "over limit" or infinite resistance. This was with the red probe on the phase Anderson and the black probe on the axle for all 3 wires and all resistance settings.
Brain Dead
otherDoc

 

[e-beach]

Hmmm OK, looks acceptable, I would still repeat the test while wiggling the wires just to be sure. Although if your tester says you have a blown hall sensor then that would make sense for your description of the oroblem.

To double check your tester, perhaps you should check the halls the Gren Tech way. It's a very simple procedure.

http://www.ebikes.ca/documents/HallSensorTestingFinal.pdf

 

[docnjoj]

Thanks ebeach
That I will do. Still don't know why a bad Hall should affect a sensorless controller ( actually 2 of them) both never used before though?

Luckily spokes are coming today for my Cute 100 and new gears for the XiongDa should be here this week. Both motors use different controllers than any I have attempted to use on the 9C that is on the trike now.
otherDoc

 

[e-beach]

Still don't know why a bad Hall should affect a sensorless controller ( actually 2 of them) both never used before though?

Perhaps 2 things at once, like a bad hall sensor and a bad connection, or connector with intermittent connection or break inside of wire insulation.

Hence moving the wires around, a few inches at a time while testing to see if a reproduction of the problem can be achieved.

 

[docnjoj]

Roger that, ebeach. I am giving up for the day after my 2 cross Cute 100 wheel ended up looking like a bad sculpture. To the bike shop with it tomorrow, then back to the wicked 9C and the tale of the five bad controllers!
otherDoc

 

[999zip999]

Don't know why these problems take so many days but sometimes they do. Good luck Doc. Wish you the best.

 

[d8veh]

I just had a similar situation with that Majic pie that I just rebuilt. It had a bad hall and a lot of rust inside due to salty water getting in. After fixing it, I tried three different controllers that could all work sensorless, but none of them would make it turn with or without the sensors connected. All I could get was clicking and the occasional kick. Finally, in desperation I tried an old KU120, which worked first time.

 

[docnjoj]

I just had a similar situation with that Majic pie that I just rebuilt. It had a bad hall and a lot of rust inside due to salty water getting in. After fixing it, I tried three different controllers that could all work sensorless, but none of them would make it turn with or without the sensors connected. All I could get was clicking and the occasional kick. Finally, in desperation I tried an old KU120, which worked first time.

Well I am one worse off from you. The last controller that I tried was sensorless, but before that I used an old Infineon that used to work with this motor. It did not. Once I get one of the geared systems going I'm gunna take the motor apart and see if I did something wrong since none of the 5 controllers worked well or at all.
otherDoc

 

[TheBeastie]

. This is the behavior of all 5 controllers that will not power the trike up our little 10 foot hill to the street, at least the ones that worked at all. The geared 2 speed did it fine till the gears broke. The geared Cute 100 did it fine till the wheel failed. They used different controllers from the ones talked about here.
otherDoc

Interesting.
I am still on my original S12S controller but I have had an occasional problem of when I just first start riding it just cuts out for about 25mins before it just magically decides to come back alive.
This mystery cut out has always happened when leaving my underground car park and going up the very steep slope drive way to the main road.
It basically cuts out right at the end when I have made it up and am just entering the main road.
The LCD display is still alive and everything looks OK just no power to the motor.. hitting the walk 6km/h mode does nothing either.

I decided I should update this as I worked out what made my ebike cut out when going up my steep driveway which was the magnetic sensor slipping around on my ebike lever which I failed to secure properly.
More details about it here.. http://www.endless-sphere.com/forums/viewtopic.php?f=3&t=69159&p=1044105#p1044105

 

[docnjoj]

I too have an update. The sensorless controller worked with my newly rebuilt XiongDa 2 speed. I didnot take it on the road since I could not shift gears with it, but it did work with the stand. The Xiongda controller arrived from China and the trike is once again mobile. Since I now have a "known good" controller I can resume testing the 9C to see if it works correctly. At least I have a working trike again, for the moment.
otherDoc