Programming Broke my Lyen Controller

[themadhatter106]

So, I've had my Lyen 12 FET controller for a long time and just recently modded it. I beefed up the traces, soldered the shunt and added an additional shunt. I also added a cooling fan and vent holes to get some airflow inside. The controller worked fine right after I modded it but the high current was causing my BMS to cut out so I reprogrammed the controller for lower current so I could ride around and figure out the new Rshunt value with an amp meter, and everything was fine.

However, when I programmed again for higher current and twisted the throttle there was a small noise from my motor and then nothing. No response from the throttle, checked the connectors and there was no problem. It seemed like every time I would reprogram the controller it would work for a split second and then do nothing afterward. One time when I was very easy on the throttle I got it to work. I was able to have the motor move the bike at a crawl speed but then when I gave it some throttle the controller stopped working and there was no response from the throttle.

I am programming using the profile 124110_Profile.asv which uses the EB318 board type.

Its so frustrating to go through all of the hours of work modding the controller, have it work and then stop working when I change the settings again! Can anyone help?

 

[amberwolf]

It doesn't sound like your thread title is accurate, as it sounds like your controller still works, as long as you have it programmed for the lower current. I might be misunderstanding your sequence of events, however.


That said, if it doesn't work at a higher current, but does at a lower one, and the shunt was modified, perhaps the modification is causing it to fault from overcurrent? (I know the mod is usually to prevent that)

What happens if you return the shunt to it's original value/state, and then try the higher current?


I presume you've already eliminated the BMS cutting out, as well, which could also give a symptom like you describe.

 

[themadhatter106]

Oh no it dosent work at all now at any current now. I've tried programming in a range of different current values.The BMS is definetly not cutting out now, the controller isent drawing any current besides quiescent current.

 

[MadRhino]

Upgrading controllers to feed more power is a risky business. I have had more than my share of fried fets and shorted boards. To have one 12 fet 4110 that is reliable to 10 Kw, or a 18 X 4110 to 15Kw, does require to buy and mod and repair many of them. They are not born equal, once you found one good it may survive a long time of hard abuse.

Then, never make one pulling more that your battery is capable. Never place it off the air flow, IMO a fan is not a replacement for adequate exposure of the controller casing to air flow. Big, quality wires and robust low resistance connectors are a must. Resistance must be equal on all 3 phase complete circuits. The board has to be insulated from the casing better than a sheet of paper, use bakelite or plastic matte. After that, it is a matter of luck.

Test yours for resistance, or short between the phases. The fist thing to know is the mosfets are good or not.

 

[Drunkskunk]

I wouldn't blame the programming just yet. There should have been no reason to add another shunt. the 0.001333 ohm (1.333 mOhm) shunt that was in place would pass more power than your motor could hope to survive, and the controller needs the shunt to remain a constant resistance in order for the programming to do what it was suppose to. Not enough resistance on the shunt would cause false readings on the controller, allowing the power to go much higher that it's suppose to.

Assuming you can fix the rest of the burned up components on the controller, you're going to need to be able to measure low resistance with micro ohm accuracy and recalibrate the controller to work with that new resistance. Otherwise, the programming will be worthless.

 

[John in CR]

I had a modded controller work kinda/sorta like that, and then quickly stop working altogether. It had nothing to do with programming. I had an electronics repair shop swap out the mosfets for me as an intended upgrade. They weren't familiar with controllers and some mosfets touched each other that shouldn't have and shorted on the first try. The motor still spun up no-load, but after less than 50ft of the first ride, the rest of the fets blew.

I'm not saying you created a short during mods, but maybe the high current setting just blew some of the fets at first and it only seemed to work before enough popped to make it not work at all.

 

[ferret]

Perhaps when you reprogrammed it you inadvertently changed the low voltage cutoff setting?

If the battery voltage is very close to the cut off value, the controller might start the motor turning and then the battery sags past the limit and the controller stops the motor.

Avner.

 

[themadhatter106]

Well I found the problem. It was just a coincidence that my controller stopped working right after I programmed it.

Testing the phase wires from the outside the resistance was fine so i decided to open up the controller and take a closer look. I tested all of the mosfets and none of them were shorted.

The shunt that I installed had started to melt and had put soot onto the surrounding PCB area.



I had a few 5mΩ 2W shunts from a motor controller I built a couple years ago. I wasn't expecting it to melt though because I didn't plan to draw any more than 100A battery peak and the power disipated at 100A by 5mΩ shunt, in parallel with the original 1.3mΩ (actually even less after soldering) shunt is (100*(0.0013/(0.005+0.0013)))^2*(0.005) = 2.12W at 90A it drops to 1.7W, and even then high currents would be experienced for a very short amount of time. I guess the first time when I was cutting out my BMS could have been very high current which caused it to melt. I found out my BMS cuts out in overcurrent protection at 240±40A. Anyway, I just snipped it off. That wouldn't have caused the problem though because the original shunt was still fine.

I also found a shaving of aluminum in a precarious place.





It's possible it got shifted and shorted something. I removed it.



After I did these two things my controller was working again. Found my new shunt to be 0.7mΩ from current comparison with a watt meter. I am pulling 100A peak on the battery.
I'm not going any higher than that with 12 FET.




I rode it around a lot and so far so good.

 

[amberwolf]

FWIW, it looks more like a little solder blob or similar might've gotten between things, shorted ground (at the shunt) to a B+ trace (if one is near there), then vaporized in plasma, leaving the soot (which would be metallic) and heating stuff in the area up.

I am not sure how a shunt could get hot enough to melt any other way (other than hooking power up backwards, causing a similar kind of dead short; I've seen the results of that in a controller sent to me)--but it is certainly easy enough to get *solder* to melt and flow under the right (wrong) circumstances.