Controller dead? 72V 40A 3000W 36$ shipped

420b

100 W
Joined
Aug 13, 2018
Messages
126
Update: Finally got this one, it’s really amazing! https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.com%2Fulk%2Fitm%2F302033173933 18 FETS with 100v caps. It’s called 72V 1000W, but it easily does 72v 40A. Hits 30 mph according to the Strava, probably capped at 28 mph. Runs in sensorless mode or sensored mode! Runs cool, the FETs have no problem.



So I posted this in general discussion, but i think it should go here.

Basically, my motor controller broke when I was riding on some off-road bumps.

My motor is a DD hall sensored hub motor. The motor spins freely when the controller is off but not when it is on. That is said to be a common symptom of a shorted FET, but I can’t seem to find any shorted FET. I checked 12 V and 5 V supply, along with hall sensors in the wheel and throttle, all of which are good. I noticed the brake wire and regen brake wire are low on voltage (1.3 V and .6 V). I tried to boost their voltage to 5 V to possibly disable electronic braking, but it didn’t fix the issue. So what do you guys think? Dead controller? Any other tests I can do to help you guys help me?

A quick update, I bought a new controller on eBay. Really excited for this one. https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.com%2Fulk%2Fitm%2F302033173933 (WOOHOO, out for delivery TODAY!!)
 
markz said:
Testing for Blown Mosfets
http://www.ebikes.ca/learn/troubleshooting.html



420b said:
Ill try it once again. I didn’t find any shorts last time. Also, why are there like 12 mosfets in mine? Seems like there are more than just 6 things that need to be measured. Right?
 
You don’t have to test each mosfet. Just test continuity between each phase wire out of the contoller (blue, green, yellow), and red input wire to the controller. That is 3 tests. Now repeat with black input wire, each of the 3 phase wires. 6 tests in all, no matter the mosfet count in your controller. If you have a short, you can replace the blown mosfets and then only, you will have to test each of them.
 
Thanks. The new controller just came in. 18 ru1h140r MOSFETS, this thing looks like it kicks ass.
 
What the frock is going on?!! I wire it all up, taking great care not to damage anything and to make it look nice. Then I plug in the battery and it explodes in my frocking face... Whip out the multimeter and theres a fcking short between negative and positive?? WHYY!!!
 
So looks like something to do with mosfets again... great. There is black char which looks to have originated from the left pin of the mosfet. Complete bullshit. So the mosfet is shorting the gate which is negative to the source which is postive terminal I’m guessing. img image hosting
 
420b said:
The motor spins freely when the controller is off but not when it is on. That is said to be a common symptom of a shorted FET

Important distinction... A shorted FET, or when it's contacts so to speak are damaged closed. Will always cause the motor resistance. Not just when the power is 'ON'. But also when it's 'off'. So I think you were on the right track below by looking for the controlling aspect of the FETs.

I noticed the brake wire and regen brake wire are low on voltage (1.3 V and .6 V). I tried to boost their voltage to 5 V to possibly disable electronic braking, but it didn’t fix the issue.

Controllers can use high on, low on or even analog variable inputs to enable brake and/or regen. Did you verify which your controller and wiring uses?
As a data point, my MP5 internal controller puts out a 3 vdc output on the brake sensor wire which when shorted to ground. (low state)
Enables both the motor cut-out function and regeneration. Although regen on the Pie doesn't work under 6 mph.
 
TommyCat said:
420b said:
The motor spins freely when the controller is off but not when it is on. That is said to be a common symptom of a shorted FET

Important distinction... A shorted FET, or when it's contacts so to speak are damaged closed. Will always cause the motor resistance. Not just when the power is 'ON'. But also when it's 'off'. So I think you were on the right track below by looking for the controlling aspect of the FETs.

I noticed the brake wire and regen brake wire are low on voltage (1.3 V and .6 V). I tried to boost their voltage to 5 V to possibly disable electronic braking, but it didn’t fix the issue.

Controllers can use high on, low on or even analog variable inputs to enable brake and/or regen. Did you verify which your controller and wiring uses?
As a data point, my MP5 internal controller puts out a 3 vdc output on the brake sensor wire which when shorted to ground. (low state)
Enables both the motor cut-out function and regeneration. Although regen on the Pie doesn't work under 6 mph.

Thanks tommy. Yes, I think some other element of my board is dead. That’s why I bought a new controller. Unfortunately the new controller just went boom
 
markz said:
Get some alcohol and wipe the black marks off.
I think there is solder shorting the traces
I thought so too. I think I’ll desolder the FET to check.
 
420b said:
markz said:
Get some alcohol and wipe the black marks off.
I think there is solder shorting the traces
I thought so too. I think I’ll desolder the FET to check.

Actually I think there is some solder dripped on there. Damnit, should have checked that before........... now I probably have to wait another 3 weeks for a FET. Thanks for the tip though mark.
 
You could also do a continuity check with your digital multi-meter between that top bus bar, and each of the through hole components. Buy some drug store isopropyl alcohol and a kitchen scrubber or use an old paint brush and be sure whats there.

You can see how the 2nd through hole from the left, almost did the same thing.

And where my orange circle is at 5 and 4 o'clock along with inside the orange circle where I believe it may have shorted.

Sloppy basement sweat shop assembly line Chinese worker solder job. What most likely happens is the basement dwellers buy the pcb's and stock up on componentry and solder away, or they buy the pcb's partly fullfilled and install their genuine or generic parts to suit their profit margins.



Clipboard01.jpg
 
markz said:
You could also do a continuity check with your digital multi-meter between that top bus bar, and each of the through hole components. Buy some drug store isopropyl alcohol and a kitchen scrubber or use an old paint brush and be sure whats there.

You can see how the 2nd through hole from the left, almost did the same thing.

And where my orange circle is at 5 and 4 o'clock along with inside the orange circle where I believe it may have shorted.

Sloppy basement sweat shop assembly line Chinese worker solder job. What most likely happens is the basement dwellers buy the pcb's and stock up on componentry and solder away, or they buy the pcb's partly fullfilled and install their genuine or generic parts to suit their profit margins.



Clipboard01.jpg

Yes mark, there was continuity between all three pins. I just pulled out the other 6 and like 3 of them are ruined due to whatever happened there. One of the three (the one with the sloppy solder) has short between all three pins, and the other 2 have drain source short. I’m going to try for a refund, I feel truely that I deserve one in this case.
 
Good luck with that, I wish the best outcome for you.
Take pictures of how it is, burnt up.
Take pictures of it cleaned up.
And take pictures of your meter readings.
Hopefully the basement dwellers will realize their bad workmanship and give you a complete refund.

I believe no one has had bad QC on ebay seller hksunwin.
Plus there is an entire long thread on that sellers controllers.
My controller has really been beaten up and yet it is still going strong!
My next controller will be a sinewave controller for ultimate quietness.


420b said:
I’m going to try for a refund, I feel truely that I deserve one in this case.
 
markz said:
Good luck with that, I wish the best outcome for you.
Take pictures of how it is, burnt up.
Take pictures of it cleaned up.
And take pictures of your meter readings.
Hopefully the basement dwellers will realize their bad workmanship and give you a complete refund.

I believe no one has had bad QC on ebay seller hksunwin.
Plus there is an entire long thread on that sellers controllers.
My controller has really been beaten up and yet it is still going strong!
My next controller will be a sinewave controller for ultimate quietness.


420b said:
I’m going to try for a refund, I feel truely that I deserve one in this case.
Thanks. I’ve been looking at sunwin. Just witnessed a 25.99$ 72V 1000W controller (15 FET) end auction without selling. I saw the thread and the soldering looks really messy, so I decided to pass.

Anyways, it seems like the controller I’ve got would be pretty sick if it had just not had that solder on the FETs. It’ an 18 FET beast with some nice copper bars and stuff. Looking to get a replacement.
 
So I think I'll buy some new 4110 FETs from here: https://www.ebay.com/itm/20Pcs-IRFB...959999?hash=item3ae1a7973f:g:KcAAAOSw7UZbPh9N
They have the same Vgs as the old RU1h140R FETs, but a little less than half of the Rds (ON) which is good.

Seems like the only thing that blew was a little SMD diode called Y2, which I can replace from my other dead controller. I ran a test with only 2 of 3 phase power, and it seems that SOMETHING is happening in the motor (electricity is doing it's thing), but it doesn't spin and I won't know for sure until we get the new FETs.
 
MadRhino said:
If you install 4110 mosfets, change them all.
Good advice, I was still unsure if I could maybe replace one in each group.

Currently, attempting repair by using the remaining 15mosfets with one missing from each phase. 15FET controllers do this right? How do 15FET controllers utilise an uneven amount of FETs per phase power wire?
 
Here are some images of my controller. If anyone knows which FET would be preferable to remove let me know, otherwise I’m just going to choose a random one of the six per phase.









 
dustNbone said:
I've never heard of a 15 FET controller. 6, 12, 18, 24, etc. All even numbered multiples of 3.
One would think multiples of three makes sense. But look here: https://endless-sphere.com/forums/viewtopic.php?f=2&t=71128

Anyways I’m getting the feeling my repair isn’t going to work. There are a bit more dead SMDs that are hindering me.
 
Yeah no dice on the repair. And the seller isn’t responding to my inquiry about a refund. Damn... I want to ride!
 
Actually come to think of it I've seen lots of 9 FET controllers so why the hell not 15.

Anyways, yeah controllers blow up sometimes, it's a good thing to keep a spare of on hand. I carry a little 18A one and a throttle in my saddle bag. Never needed it so far but I'll feel very clever the day I do.
 
dustNbone said:
I've never heard of a 15 FET controller. 6, 12, 18, 24, etc. All even numbered multiples of 3.

I have 2 of them 15 fets controllers, that I had bought 2 years ago. Sleeping on the shelves, for I never got a chance to upgrade them to high power. They were cheap, I believe both for 120$ shipped, so I decided why not giving them a try. But I have many 18 fet controllers already upgraded, so I never felt the need to bother with those.
 
Back
Top