Burnt 244110ST-MII-LYEN Edition - Alternatives?

[Ilja]

Hi!
My motor controller 244110ST-MII-LYEN Edition burned out, can it still be bought? if not which alternatives do I have?, I want some controller 5-10kw with CA-DP (Cycle Analysts V2 connector) but did not found it.
I am new in this theme and dont know so much

 

[Ilja]

Firmware would also help me because I think that main chip is broken, may be I can buy one from china and flash it?
Controller: 244110ST-MII-LYEN
Chip: GPM8F3108A

 

[amberwolf]

The firmware is not available, AFAIK. You could check with Lyen to see if he had custom firmware flashed to them and if he has that file to send you, or could flash it for you if you send the controller to him. I don't know who actually made his controllers for him, but they could do it too, if you could find out--if you can tell them what FW version to use (not sure how you can find out, other than from Lyen, because you can't read the MCU's firmware out even if the MCU is working; they usually set the protection up inside them to prevent that).

What makes you think the MCU is broken? And since they dont' typically just fail, what have you found that caused it to fail?

 

[Ilja]

It happened after I set 90 Amps on Cycle Analysts V2, I think it was too much. My bike just stopped working after I wanted to accelerate from zero. 1. Should not the controller itself monitor the current and limit it if its too much? (like I set 90A on Cycle Analysts V2 but controller limit it like 50A because its max)

I checked handle, speed switch, then I opened motor wheel and there was 3 hall sensor, 2 of they did not worked ( It was no response to magnet ) changed it and tested, while I spin the wheel all 3 wires from hall sensors changing levels so I think its works.

Then I opened the Motor Controller, all looks right inside, measured levels from sensors on chip pins - while holding handle full throttle its coming 3.5V on chip, coming signal from hall sensor, and its exist 5v and 15v, chip sending on signals on all pwm pins to close mosfets (its inverted) and nothing more happens.

So I think that something happened on 5v supply. Maybe the voltage has risen some reason and killed hall sensors and chip.

I hope you understand my english)

 

[amberwolf]

It happened after I set 90 Amps on Cycle Analysts V2, I think it was too much. My bike just stopped working after I wanted to accelerate from zero. 1. Should not the controller itself monitor the current and limit it if its too much? (like I set 90A on Cycle Analysts V2 but controller limit it like 50A because its max)

Yes, the controller should limit, but sometimes they're not well designed or built and the actual parts used by the builder can't handle the limit chosen by the designer. Or the limit is chosen for a certain situation (high current full speed) and installation type (well-ventilated, open air with airflow), so that in other situations (high current low speed, etc) or other instalation types (air blocked, or in a bag or box, etc) it will overheat and parts will fail.

Usually that is the FETs. When they fail under load, that often blows up the gate drivers, but rarely damages the MCU or other parts. If the design is poor, and the overload the right kind, it could damage other things, but it's not that common.

The CA shouldn't have to do the limiting...but if the system works with the CA doing the limiting but not the controller, then the controller's limiting is insufficient for the purpose, and the CA's highly tunable limiting can simply be used instead.

I checked handle, speed switch, then I opened motor wheel and there was 3 hall sensor, 2 of they did not worked ( It was no response to magnet ) changed it and tested, while I spin the wheel all 3 wires from hall sensors changing levels so I think its works.

If the hall sensors failed, there's usually a reason. That can be overheating of the motor, or it can be wire damage in the motor cable, or if the controller did fail catastrophically beyond just the FETs, it could have damaged the hall sensors with improper voltages on the signal lines (if it were on the 5v supply line, it would probably have damaged all three FETs and every other thing on the bike running on that 5v).

If it's motor cable damage, it's usually at the exit of the motor (axle) and usually has some visual sign, even if you can't actually see the wire damage. It may not be a hard short, but rather intermittent only when the conditions are right (usually vibration or bending of the cable in the right way).

Then I opened the Motor Controller, all looks right inside, measured levels from sensors on chip pins - while holding handle full throttle its coming 3.5V on chip, coming signal from hall sensor, and its exist 5v and 15v, chip sending on signals on all pwm pins to close mosfets (its inverted) and nothing more happens.

So I think that something happened on 5v supply. Maybe the voltage has risen some reason and killed hall sensors and chip.

If something happened to 5v like that, it would probably also have killed your throttle, PAS sensor, and anything else that runs on 5v. Since the 5v is presently working and nromal, and the MCU is sending signals out in response to throttle, then damage to those parts is unlikely.

It's more likely that the FETs failed from the overcurrent or from a short in the motor cable (which if it shorted phases to hall signals could also have killed the halls)

 

[Ilja]

Thank you for trying to help me.
I think you are right, temperature killed sensors because I had some not connected and they also was broken, and few wires was brown in motor(white cable insulation). I have tested coils and it seems that they have same resistance like 0.4 ohms (Do controller gonna try to spin the wheel if few of them is shorted?, I think it gonna try and I gonna hear like vibrations). Soldered out all 24 N-Mosfets, one was shorted on the ground! . I don't really understand how it happened, did this fet burned out because chip got wrong signal from hall sensor and like for example opened opened mosfets too early?? Can chip have protection for burned fets? Will it work if I solder all back but with out one fet?

 

[amberwolf]

If hte windings were shorted, the wheel would be hard(er) to turn even if disconnected from all wires.

Consistent resistance across each phase pair is encouraging, though without a special meter for very low resistance testing it's tough to get an accurate measurement (I use a DE5000 and it doesn't go accurately low enough to measure shunts or really low resistance motors; I'd need something significantly better and more expensive for that).

The FETs could die from overcurrent, or from overheating, or voltage spikes, or even from ESD damage from handling at any point in their existence from where they were manufactured to where they were installed in the controller. Overheating could kill just one FET either because it happened to have the highest resistance in that phase bridge, or because it's thermal connection to the heatsink was poorer than the others, by just enough to cause more load in it than the others. If it had been slightly better, then the failure could've taken enough longer to happen to also out more parallel FETs rather than just the one, assuming this was the issue.

If the MCU had a wrong signal and sent the wrong drive commands, it would be sent to all the FETs in that phase and whichever other phase was seriesed with it at that instant, and so it's likely that more FETs would have failed.

If you don't replace that FET, you cut down the phase current handling ability of the entire controller proportionally. It's a 24FET, then that means 4 FETs in each half of each of 3 phase bridges, so you cut the ability of the controller down to at best 3/4 of what it had been, so you would need to adjust it's settings to match that lower capability, or risk blowing up the other three FETs in that phase during any particular moment of operation.

If it's not adjustable, then you would need to replace the FET with an identical one to have it perform like it used to; if ti's not repaced and the ocntroller still tries to do what it could do before...you'll probably see a failure quickly enough when sufficient current is demanded by the motor. If the FEts aren't identical, or very very close in characteristics, then they won't share current well, and whichever one has the lowest resistance will have to ahndle more current, and be more likely to fail, and then others in parallell may fail at the same time if current demand is high enough at that moment.

Controllers are not usually designed to protect individual parts, just all or nothing. If your controller has individual phase current sensors *and* it's software is well-written, it may at least detect a current flow problem and be able to shut down, but if it is only battery current it has no idea what is happening to the FETs or motor phases.

SInce it doesnt' know the condition of the FETs, it doesn't do anything different if one or more or blown--it will still try to do whatever it's commanded to do, until enough parts fail for it to be unable to continue.