MadRhino
100 GW
Yep. Try to use copper wire size to match shunt size.
Repairing a shunt on a mainboard
- Thread starter ebike11
- Start date
Solder has higher resistance than wire, especially copper, so it would actually work *better* across a broken shunt than copper, for doing the job the shunt would do. 
But even a zero resistance shunt (pure copper) would still allow the controller to "work", but it would have no way of knowing it was putting far too much current thru the FETs, and could easily destroy itself under loaded conditions.
It's also possible that the circuit design allowed the high voltage to be available on the MCU input pin itself, and then the MCU would be damaged and the controller essentially toast.
But even a zero resistance shunt (pure copper) would still allow the controller to "work", but it would have no way of knowing it was putting far too much current thru the FETs, and could easily destroy itself under loaded conditions.
^^This^^ I didn't think of that, but that makes perfect sense, and is probably the failure. Whatever failure mode the LM358 has may create an output that is whatever the controller thinks is too high a current, so it's refusing to operate.fechter said:
It's also possible that the circuit design allowed the high voltage to be available on the MCU input pin itself, and then the MCU would be damaged and the controller essentially toast.
amberwolf said:Solder has higher resistance than wire, especially copper, so it would actually work *better* across a broken shunt than copper, for doing the job the shunt would do.
But even a zero resistance shunt (pure copper) would still allow the controller to "work", but it would have no way of knowing it was putting far too much current thru the FETs, and could easily destroy itself under loaded conditions.
^^This^^ I didn't think of that, but that makes perfect sense, and is probably the failure. Whatever failure mode the LM358 has may create an output that is whatever the controller thinks is too high a current, so it's refusing to operate.fechter said:
It's also possible that the circuit design allowed the high voltage to be available on the MCU input pin itself, and then the MCU would be damaged and the controller essentially toast.
If it is in fact the LM358 that is the problem here, would I still get normal voltage and 5V power? because the controller turns on and is outputting the 72V and also throttle/halls is reading properly at 5V
All that part does is translate the shunt voltage into what the MCU needs so it can understand how much current is flowing.
It isn't part of the LVPSU (12v & 5v) regulation system.
BTW, the controller doesn't output 72v--that's just your battery.
It isn't part of the LVPSU (12v & 5v) regulation system.
BTW, the controller doesn't output 72v--that's just your battery.
amberwolf said:
ok thx..ah i meant that the controller is showing 72v when using my dmm on the mainboard while being connected to the battery
i thought id check if the voltage is properly flowing through the controller and that nothing is blown
but as you said, other componants are part of the system
Can you post a good picture of the controller board? You may be able to trace out the circuit and find a spot to test the current amplifier.
fechter said:
Hi there! Here are a few pics i took. You can see the shunt i damaged in the center of the board. I can take more if needed. I tried to upload through the forum but the size limit is only 512k
https://ibb.co/JtW9mPb
https://ibb.co/8jdSfv1
https://ibb.co/QQH87yv
https://ibb.co/0Y0rnWG
https://ibb.co/wdbtH6j
DogDipstick
100 kW
ebike11 said:
What is the numbers on the top right of this board? Is that a 618 - or a 318 -?
I can't see all the pics, but there's one that does show the three shunts.ebike11 said:
None of them are connected to their other end.
Until there is a connection across the two ends of the shunt, nothing can work, becuase the FETs (and possibly other stuff) has no battery negative connection.
What looks to be the current sensing line is marked in the pic below. I can't see where it goes under the wires. Look to see where that trace goes. In this model of controller, possibly the current sensing amplifier is integrated into the main MCU chip. There may be a resistor in series between the sensing line and the MCU that should be checked. If there are no obviously failed parts outside the MCU, it will probably be hopeless as replacing the MCU is not likely to be possible.
docw009
1 MW
How about you just make a new shunt with copper wire. See if the motor turns. If not, you cannot fix it. Throw the controller out or save it to steal spare parts like capacitors.
Very few of us have the skills, tools, and eyesight to replace a surface mount component. Twenty years ago, I could do the simpler ones with a heat gun and solder pencil, but not now,
Very few of us have the skills, tools, and eyesight to replace a surface mount component. Twenty years ago, I could do the simpler ones with a heat gun and solder pencil, but not now,
ebike11 said:
That looks to be jumper resistor. Beyond that, there are a couple more resistors that lead to the main chip. I'm not seeing a separate amplifier or any obviously fried parts. My guess would be the main chip is toast.
ebike11 said:
Probably not. It's just a piece of wire inside a resistor package. The next one in line going to the main chip could blow and would be worth checking.
If a FET blew, you'd have other problems and they are generally obvious. This is pretty easy to check by disconnecting the motor and measuring resistance between each phase wire and battery negative, and each phase wire to battery positive. None of these combinations should look like a short (zero ohms). Each combination should measure like a diode.
Similar threads
