Kelly Controller has no torque, jerks motor

Get all your technical information about electric bikes here.
User avatar
izeman   100 GW

100 GW
Posts: 5104
Joined: Jun 21 2011 8:25am
Location: Vienna, Austria

Re: Kelly Controller has no torque, jerks motor

Post by izeman » Oct 21 2019 9:44am

You need a HOT AIR GUN. Heat the pcb and pins and the FETs will just fall out. Imho that's the only proper way to desolder them.
Then use a solder pump to clear the pcb holes, put in new FETs, solder them and you're done.

User avatar
amberwolf   100 GW

100 GW
Posts: 30787
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group

Re: Kelly Controller has no torque, jerks motor

Post by amberwolf » Oct 21 2019 1:39pm

A big fat tip soldering iron, commonly called a hammerhead or chisel tip, will help a lot in these cases, because the FET pins (except the gate pin) are on huge traces that act as heatsinks and make it very hard to desolder or solder with a small tip iron.

It doesn't have to be a huge wattage iron, but that can help too--I use an 80w Weller with a chisel tip about the thickness of my pinky, for stuff like this, and heavy gauge wires, and it just works quick and easy like the smaller tips do on regular components and traces.

FWIW, you probably can't properly solder the FET legs to the existing leads sticking out of the board, because you first have to cut the legs off the new ones to the same length as the leads sticking out, whcih leaves very little space between the new ones and the board, to be able to reach with the iron tip.

(if you don't have them the same length, they won't be able to bolt to the heatsink or spreader bar, cuz the mounting tab hole will be at the wrong height. Also, another possible problem is if the legs arent' shaped to overlap the existing legs in the right way, the new FETs wont' be vertically parallel to the heatsink once mounted, and will not be able to shed their heat correctly)

The iron tip itself may not be able to heat them quickly enough to do the soldering for the same reason you can't heat the existing leads up enough to get them off, and if you heat them a long time you can damage the new FETs.

I use a solder sucker on some things, but solder wick works better for cleaning up holes after removing components. Sometimes to get the solder out with a sucker, you have ot first *add* solder to fill the hole completely to make the sucker work. Don't have to do that with wick.

bigchief   100 mW

100 mW
Posts: 45
Joined: Jun 20 2009 12:35pm

Re: Kelly Controller has no torque, jerks motor

Post by bigchief » Oct 25 2019 1:50am

Thanks for the advice, I got a hot air gun on Amazon and it was great for desoldering the pegs. The only problem was when clearing out the PCB holes with my plunger-type solder sucker, the plunger "jerked" when I hit the button and the shock knocked an SMD resistor off the board that was a few millimeters from the PCB hole. Lesson learned not to ever do it again from the component side, it took me a while to find that little speck on my solder mat but I soldered it back on the spot.

It took me a while but I properly replaced every one of the mosfets and put the controller back together. I screwed them all to the two heatsinks first so they were all lined up and I made sure the whole row slid into the PCB holes evenly, then I soldered.

Now the controller still powers up, but when I turn the thottle the motor does an extremely faint jerk and then the RED fault LED starts blinking 3 then 2, which means "Internal reset: May be caused by some transient fault condition like a temporary over-current, momentarily high or low battery voltage. This can happen during normal operation." according to the manual.

The battery voltage is correct, 38 volts (lvc is 33 and hvc is 60). The controller still holds it's parameters correctly when I connect it to the PC, and it throws the correct error code when I intentionally don't plug the hall wires in, so the controller isn't totally busted or anything.

I also took it out of the case to see if it behaves any different (heatsinks are still on, just not attached to the outside case... since someone here said the new mosfets dissipate 1/8th the heat it should be fine to run outside the case for a few seconds to test). But it behaves the same outside the case.

I'm 95% sure I soldered that resistor back on correctly, so I'm stumped what else could cause that fault. These new mosfets should be compatible according to y'all. I don't see any blown capacitors or anything else weird. Maybe I'll take a look tomorrow with my infrared camera, it might reveal some other faulty component.

Post Reply