[JustintheFalcon]

For the last five years I've been running a pair of 1239-8501 Curtis controllers on a HPEVS AC35-2 at 144V's with no troubles. Now the primary controller is giving me an error 38 which is "Main Contactor Welded". Just to be thorough, i replaced both the secondary and the primary main contactor relays. Replaced both precharge relays too. Still same error. I even swapped the primary and secondary main relays with each other(you never know even with newly purchased stuff). Still only the primary gives the error and secondary boots without errors. I know the primary main contactor is not welded shut because simply measuring no DC voltage across the controller inputs proves that. I sent the controller to FSIP... an approved Curtis repair house and for a mere $500 they replaced all the FETs on one of the three AC legs as they said they were all blown. They did nothing else. OK fine, but when it's returned to me I get the same exact error.

The controller is designed to close the precharge relay(with 50ohm resister) for a moment before closing the main relay. This is good because we don't want the main relay to arc when it closes. Fine simple enough. However with my primary controller error state it won't keep the main contact closed. My question is, how crazy is it to bypass the controller? If I were to add switches to manually flip ON the precharge relay, manually flip ON the main contactor relay and then manually flip OFF the precharge. Too crazy? Should i sent the controller back again for another go with FSIP? Any other suggestions?

Couple pics of my setup...

today(rusty/dusty)

five years ago(shiny and new)

 

[morvolts]

rather than physically moving the contactors can you trick the controller by moving just the wires that sense voltage at the contactor>
Maybe problem has to do with this and not the physical contactor.
Garbage in=garbage out

 

[JustintheFalcon]

rather than physically moving the contactors can you trick the controller by moving just the wires that sense voltage at the contactor>
Maybe problem has to do with this and not the physical contactor.
Garbage in=garbage out

I think i know where you are going with that. However I assume the sensing is internal to the controller at the main lug nuts for DC input. I don't think there is any wiring to/from the controller other than the "do it... close the main contact switch" and the heavily cabled +B side power line. Does make me wonder if that "do it" is burnt out though... maybe it's not got enough juice from the controller to close the relay quickly enough... but that wouldn't be a "welded" error scenario but rather an open. Hmmm.

 

[kennybobby]

For $500 repair, i would expect that they should have fully tested it to verify proper operation of ALL the functions, not just replace some FETs. What caused the FETs to fail?--could be the same thing that is causing the False welded error message. Seems that they didn't do a complete repair and a call back should be on them to finish the job, but i doubt they will volunteer for a free rework although they should.

Don't know the algorithm, but it is likely measuring the voltage at the HV input terminal before and after the "precharge begin" signal is sent. If the signal conditioning circuit is crapped out then it may be reading a "Hi" signal all the time, and this triggers the welded error (if the HV is present before the precharge command, then the contactor is welded).

There is likely an op amp or comparator being used to determine the voltage, and that is the circuit that needs to be verified. i don't know if there are data fields available to the user to view and monitor, but maybe it would show up in that data as showing HV present before the command.

i wouldn't recommend to try to bypass the controller, you don't know what the control logic is for running and operation with a fault or if it is even possible (not likely).

You may have 2 issues, to me the failed FETs is a bigger concern to find the root cause.

Good luck let us know what you find out about this.

 

[JustintheFalcon]

For $500 repair, i would expect that they should have fully tested it to verify proper operation of ALL the functions, not just replace some FETs. What caused the FETs to fail?--could be the same thing that is causing the False welded error message. Seems that they didn't do a complete repair and a call back should be on them to finish the job, but i doubt they will volunteer for a free rework although they should.

Don't know the algorithm, but it is likely measuring the voltage at the HV input terminal before and after the "precharge begin" signal is sent. If the signal conditioning circuit is crapped out then it may be reading a "Hi" signal all the time, and this triggers the welded error (if the HV is present before the precharge command, then the contactor is welded).

There is likely an op amp or comparator being used to determine the voltage, and that is the circuit that needs to be verified. i don't know if there are data fields available to the user to view and monitor, but maybe it would show up in that data as showing HV present before the command.

i wouldn't recommend to try to bypass the controller, you don't know what the control logic is for running and operation with a fault or if it is even possible (not likely).

You may have 2 issues, to me the failed FETs is a bigger concern to find the root cause.

Good luck let us know what you find out about this.

Yeah I'm going to send the unit back. They are the recommended/qualified Curtis controller repair house. It's best not to start bypassing the stuff in place meant to protect us. The issue occurred when my daughter left the vehicle "on" and charging overnight. This means the contactors we're engaged for +8hours... instead of what normally would only be an hour or so at a time. The accelerator obviously wasn't on but the motor could have been sinking some (~1amp?) current thru one leg of the controller for this extended time. I'm only guessing here. The secondary controller on this dual motor setup comes up with no errors. Fingers crossed i get a better report on the primary controller. Thanks for your input. I'll post when I know more.

 

[JustintheFalcon]

Soo.... 2nd time I sent the 1239-8501 Curtis primary controller unit back it they said...

"your reinstallation likely did not cause any damage"
"the solder joints were not as secure as they should’ve been due to the damage to the boards, so while it appeared that the components were secure, a decent bump probably broke them off"
"replaced all three of the IC chips that drive the controllers phases"

... and this time when I re-reinstalled the unit in my EV in Dec2020 the original errors were gone and things were working fine. Oddly the motor temperature read something like 896 degrees C on the master spi glass. Ran fine for several months until a few weeks ago I had to get towed because this time both controllers stopped working (last time i could limp home on the secondary controller). Secondary says it's welded shut... this is a brand new TE Tyco Kilovac LEV100 12V 100A-Cont. Sealed Contactor. Is my secondary precharge not working or could the main controller cause excessive current on the secondary? Tapping in place didn't free the weld. When I removed the relay and pounded it on the table it did break free.

 

[morvolts]

I wouldnt trust a relay or contactor that was welded closed and then opened by banging on it.
Only in the case of rebuildable ones that the contacts can be properly dressed or replaced.
I just has a 100A one do the same and pitched it and put in a 400A.

 

[JustintheFalcon]

I hear you. So freaking annoyed at replacing these things. I'm not pleased with electricmotorsport.com. Where did you get your 400A relay from? Also do you have a respectable high amp kill switch?

 

[morvolts]

I hear you. So freaking annoyed at replacing these things. I'm not pleased with electricmotorsport.com. Where did you get your 400A relay from? Also do you have a respectable high amp kill switch?

I got my 400A one off ebay look for the brown/tan one.
I have also had good luck with Newark and the name brand stuff.