I'm buying a non-runner MIEV here in Japan!

[kiev]

The reason to only jack up one side is to lock up the differential in the gearbox so that the rotation of the wheel will primarily back drive the motor rather than spin the wheel on the other side.

The spin direction doesn't really matter, just whichever is easier. Either side is fine too, i just figured the right rear would be easier to spin. The hard/easy sections of spinning may be due to brake drag if the rotors are rusty. Remove the right wheel and use a screwdriver to pry the pads away from the rotor (manually spread the pads)

The DTC check with the motor leads disconnected seems to indicate that maybe there is a shorted winding in the motor, or that one of the current sensors is defective, or the control board interpretation of the current sensor data is flawed. This was seen in one of your previous data captures showing current at 2000 Amps !? Was that the new or old MCU ?

Were you able to measure resistance between the motor phases, and between phases and chassis ground/motor housing? You can refer to them by color if that is easier, R, W, B.

 

[kiev]

Did you try turning a wheel while 2 phases were connected together (shorted phase test), and also with all 3 connected together?

 

[Rational]

The reason to only jack up one side is to lock up the differential in the gearbox so that the rotation of the wheel will primarily back drive the motor rather than spin the wheel on the other side.

Thanks for that clarification detail. The wheel on the other side doesn't spin when I spin the other side.

So I guess this means the test is still valid with both wheels off the ground?

Remove the right wheel and use a screwdriver to pry the pads away from the rotor (manually spread the pads)

Will do next.

This was seen in one of your previous data captures showing current at 2000 Amps !? Was that the new or old MCU ?

It was? Where did I post that showing current at 2000 amps? I guess I didn't know to even look for it to notice.

This is the new MCU but with the old/original controller board. I did this so the chassis number would match.

Were you able to measure resistance between the motor phases, and between phases and chassis ground/motor housing? You can refer to them by color if that is easier, R, W, B.

I will do that now. Do I just set it to ohms on the multimeter and see what it shows?

R,W,B is easier. Will do it that way from here on out.

 

[Rational]

Okay. Here is the i-miev motor UVW resistance test results video.

 

[kiev]

Just a post listing the resistance would be fine. The utube videos are not playing with message "Sign in to confirm you're not a bot"

There is a reason why i asked that you only turn the key to ON after clearing the codes, and then check which codes reappear without starting (looking at the U1111 CAN buss code). But finish all the phase testing first before putting it back together. i'm not really sure what could be learned or proven by starting without the motor connected? Putting the new control board back into the new MCU with the motor connected would be more useful to prove the IGBT code goes away, even if it throws an eeprom chassis error, that would be progress. Then the old eeprom could be desoldered and installed on the new board,

i'm hoping that spreading the brake pads on one side will free up the spinning wheel and you can capture some winding voltages while spinning. Maybe use a peak hold function on the meter while in AC volts (the output is a sine wave with + and - sections). If you could spin the wheel using a lug wrench or some tool to get maybe 1 rev per second, or per 2 or 3 seconds, however fast you can spin it. Just catch the peak hold at a certain speed, then write it down and do the other phases. If you can spin them all about the same, then compare the voltage readings--just looking if they are all about the same, or if one phase causes a lower reading. On the one video i could watch, the phone was moving too much to read the screen, that's why i suggest using peak hold feature, or min max, if your meter has such.

 

[Rational]

There is a reason why i asked that you only turn the key to ON after clearing the codes, and then check which codes reappear without starting

When you say "ON" you mean ACC? Which I believe is the half position before full start up? I just want to make sure we define our terms before I misspeak.

So far when the ignition is in the half position before full start position it does not read DTC codes nor talk to the car. No codes can be obtained unless I have the car fully turned on and showing START on the dash.

i'm not really sure what could be learned or proven by starting without the motor connected?

I thought you or Mickey had said this would tell us something if no P1A2A codes appeared again with the motor disconnected.

Putting the new control board back into the new MCU with the motor connected would be more useful to prove the IGBT code goes away, even if it throws an eeprom chassis error, that would be progress.

II will do that once I take the right rear wheel off and spread the break pad away from the rotor and try to get further resistance tests with the multimeter set to peak hold if I can figure that out.

while in AC volts

Okay. Will do.

the phone was moving too much to read the screen

I see. I will up my precision by using a stand for future videos.

I think logging into youtube if you have an account will ensure they play. But I understand if you don't feel the need to comply with their BS.

By the way, I just received 3 more OBD readers from aliexpress just this minute. I ordered several types to use on this and future car repairs.

Just got a GODIAG GD101, a Multidiag v3.0 and one Lexia 3 that the french site said was satisfactory, so we'll see if this one works today..

https://www.aliexpress.com/item/100....order_list.order_list_main.65.307b585a49eoUF

https://www.aliexpress.com/item/100...der_detail.order_detail_item.4.51d6f19cNzbYau

https://www.aliexpress.com/item/100....order_list.order_list_main.77.307b585a49eoUF



Still no FULL CHIP Lexia 3 cable yet though.

 

[MickeyS70]

Here are the results of the tests between the UVW leads.

This is an AC motor, need to set your DMM accordingly

I wonder if I can start the car in N and see if it won't error and actually spin the rear wheels?

Don’t think so, has to be in P to start.

Don’t forget to do below and while you’re at it measure resistance between U/V/W:

kiev said:
If the motor has a shorted winding then spinning the wheel may be difficult as if the brake is holding. You can get a feel for this by connecting 2 motor leads together and try to spin the wheel, then connect all 3 leads together and try to spin the wheel. A shorted winding will also have a distorted or lower amplitude sine wave, or no voltage output at all.

 

[Rational]

This is an AC motor, need to set your DMM accordingly

Oh, I wasn't quite realizing this is an AC motor. Thank you for nailing that down in my mind firmly from here on out. Got it. Will do.

I was also forgetting about connecting the motor leads together to see if it increases turn resistance. Will do.

I've got the tire off now and opening up the drum brake now to see if it's rubbing.

 

[MickeyS70]

Okay. Here is the i-miev motor UVW resistance test results video

Measuring coil resistance is a static test, need to set the DMM to it’s highest sensitivity (mOhm)

 

[kiev]

Resistance is a static measurement from phase to phase. No wheel spin for that.

It appears that your meter was reading 000.0 Ohms, which is basically zero. The resistance is likely in the milli-ohm region which your meter may not be able to measure.

What about measuring resistance from motor phase to chassis? Do you get any reading or is it open?

====
The back-emf voltage reading is measured while spinning a wheel. When back driving the motor is acts like a generator, where the voltage is proportional to speed. The faster the spin, the greater the voltage. The main thing trying to do is just to determine if, while spinning the wheel at about the same speed, then the voltage is about the same for all 3 cases. If one of the phases were shorted then the results involving that phase would be lower.

Hopefully you found the brake to be dragging and it can be relieved to spin freely.

 

[Rational]

Okay. I got the drum cover off finally. Had to find two 12mm head bolts to use in the two holes to separate the drum from the car.

I'll post two pics of before and after.

Hopefully you found the brake to be dragging and it can be relieved to spin freely.

Yep. You're right again as always.

I can confirm after removal that all the resistance seen and felt before was entirely from the drum brakes. Right now I am able to spin the axle with a lot less effort by hand.

 

[Rational]

Great progress, i'm gonna have to drop off for now and let Mickey carry on. i'll check back later in the morning...

Thank you so much for all your help again today.

Okay. I've shorted out R,W,B to each other in pairs separately and all together. It becomes very hard to turn the axle when they are shorted together in pairs and all three together.

I will try to understand and get the resistance to chassis ground next.

I have another DMM that has a good peak hold and is able to read about 4-5v when I spin the axle for all three pairs tested individually. No real difference between them that I can tell.

 

[Rational]

Resistance is a static measurement from phase to phase. No wheel spin for that.

It appears that your meter was reading 000.0 Ohms, which is basically zero. The resistance is likely in the milli-ohm region which your meter may not be able to measure.

What about measuring resistance from motor phase to chassis? Do you get any reading or is it open?

I also tested the the static resistance measurement with both DMMs and found the resistance to be 0.000 in k-ohms on my green DMM for all three pairs of R,W,B. This DMM allows all three ranges to be manually selected (k-ohms, m-ohms and ohms).

My other DMM auto detects and gives 000.1 ohms for all three pairs.

Both DMMs show 0.00 OL for R,W,B to chassis.

 

[MickeyS70]

Let’s summarise then:

Motor unit:
The turning resistance was caused by the ‘sticky’ disk brakes, AC voltage measured between phases U,V,W when turning the wheel manually seem identical, which indicates similar coil resistance and no short to chassis. Connecting two or more phases together increases the spinning resistance of the wheels.

IMHO this all points to the motor/gearbox being OK?

MCU:
One of the original IGBTs looked blown, therefore the inverter was replaced completely except for the control board with was switched back from the original car to overcome VIN/start angle issues
Disconnecting the motor and starting clears P1A2A even when in D or R. As this DTC indicates IGBT overcurrent this behaviour is expected as no current flows (unless there is a short inside the inverter).

P1A2A is either ‘real’ (caused by the ‘sticky’ brakes?), due to a faulty current reading (control board?) or caused by something else entirely (what’s left?)

I would disable the rear brakes then reconnect the motor cables, leave it jacked up, start the car and put it into drive.

If the same DTCs are set, see can you read out the ‘initial phase angle’ with your Lexia cable, then change the MCU control board and try again (don’t expect anything to move just check for a change in error codes).

 

[kiev]

Mickey and i are on the same channel on this, i was trying to write up a similar response just now too.

All the testing seems to have ruled out the motor as culprit: no short to chassis, or phase to phase, and bemf response is as expected.

Hopefully the old control board did not command anything that might blow one of the "new" IGBTs

The new MCU has "new" current sensors, so those can be ruled out assuming that the new igbts are good.

The 2000 Amp reading seems to indicate bad data from the control board. So its likely the old control board is defective in addition to the old igbt, a new failure mode for this P-code.

Other notes:
If you run the MCU without coolant, do only very short durations.

Your key positions may be different, but the CAN should be active when in ON position. OFF, ACC, ON, START (spring-loaded returns to ON when released)

 

[MickeyS70]

Anothper possibility (or is it?) is that the IGBTs blew twice, initially because of towing a 2 ton truck and again just now on the replacement inverter because of the sticky brakes (similar to what happened to @Neo5)?

Might be worth having a visual inspection of the current IGBTs just in case…

 

[kiev]

Anothper possibility (or is it?) is that the IGBTs blew twice,

That's what i was worried about, if the gate drivers were blown and shorted out, it might cause both the top and bottom gates to come ON at the same time--that could cause a 2000 Amp spike, yikes!

Probably a good idea to inspect that the devices are OK. Make some measurements (diode drop, resistance, continuity, etc) to check and verify.

@Rational said: I thought you or Mickey had said this would tell us something if no P1A2A codes appeared again with the motor disconnected.

It wasn't me

 

[MickeyS70]

That's what i was worried about, if the gate drivers were blown and shorted out, it might cause both the top and bottom gates to come ON at the same time--that could cause a 2000 Amp spike, yikes!

Not sure this was real, HobDrive only showed 1A

It wasn't me

Guilty as charged, as corrected earlier, this would only be a ‘smoking gun’ if the DTC remained with the motor disconnected..

 

[Rational]

Okay. Thanks guys. Excellent write ups!

It's Monday here now so I won't be able to do much today. I got kinda tired late yesterday as I had other honey-do tasks to get done so had to stop there. But I'll get back on it again tonight for at least an hour or two if at all possible.

I'll likely take off the left rear tire and disc drum cover to eliminate any further binding. I also want to adjust these to have as little rolling resistance as possible when I put them back on. That way the car gets optimum range. I'll check the fronts too once I get to that point to as well.

I got a camera mount/holder and will record the resistance measurements for posterity and to double check them while I have things disconnected. Just to be 100% sure and allow you both to see the motor condition with your own eyes to be absolutely sure that is ruled out.

As mentioned I have another MCU still inbound. I'll swap in that one untouched when it arrives to see what we get for codes. Hopefully just the VIN mismatch error and the P1A2A will go away.

I'll also double check my ignition behavior to see if it is as you described and I was just not realizing it had four positions. OFF, ACC, ON, START. From memory it feels like it has OFF, then two more, and the fourth is a temp over reach position that feels like traditional ignitions to start the car but springs back to the ON position. So it is likely the same as you describe and I wasn't realizing the nuances here between ACC, ON and start.
It almost felt like I didn't have to go all the way to the forth temp position and the car would still go to START when just set to ON. But I'll test this theory to be sure. I was always going all the way through all four positions to the max turn radius of the ignition switch and then letting it spring back to ON. I'll try just going to ON and not going to the fourth position to see if that's any different.

If you run the MCU without coolant, do only very short durations.

Okay. I'll limit tests to about 5 minutes or less and give it 5 minutes to rest between tests while in testing.

Couple of questions though.

Where did I post a screenshot or data indicating that 2000 amp spike? Just so I know what you're talking about and how to look for it again?

assuming that the new igbts are good.

This is one of the problems. I'm assuming the replacement MCUs are coming with good IGBTs. Likey they are but I've not opened them up to verify. Not optimal testing strategy to assume.

Might be worth having a visual inspection of the current IGBTs just in case…

Yes. I think so. I will pull the current MCU out and visually inspect the IGBTs in it while we're awaiting the next MCU unit.

But does that mean I shouldn't just put in the third MCU when it arrives as is? Should I open it up to inspect the IGBTs first to be sure they're okay?

That's what i was worried about, if the gate drivers were blown and shorted out, it might cause both the top and bottom gates to come ON at the same time--that could cause a 2000 Amp spike, yikes!

I basically follow, but didn't know until this statement that there are 'top and bottom gates'. Can you share a bit more info on what these are and how they work? Are they the IGBTs you're talking about? So the three IGBTs are actually 6 gates? 3 top and 3 bottom?

Probably a good idea to inspect that the devices are OK. Make some measurements (diode drop, resistance, continuity, etc) to check and verify.

Just so I'm sure exactly what devices and how to check them, are you talking about the IGBTs being tested each in isolation? So taking them out and testing them somehow? I'd need some specifics but am glad to perform them. I'd learn some things and could share with the group for future users.

If the same DTCs are set, see can you read out the ‘initial phase angle’ with your Lexia cable, then change the MCU control board and try again (don’t expect anything to move just check for a change in error codes).

Glad to do this one as well. Do you happen to know where in the software this is retrieved? And is it in MUT-3? Diagbox?

I'll have to spend some time getting the software working too. E.g. WIn 7 32bit or 64bit, or Win 10 64 bit depending on what it requires. I've found all these Chinese cables to be quite finicky about initial OS setup in order to work. Might take some time to get it talking.

Thanks guys. Excellent learning for me thanks to you both. This is a real education in EV repair.

I look at it as an excellent opportunity to truly learn how these things work and how to repair them in a high quality manner so I can do so on future cars and for others later. So this is actually fortunate this car happened to land in my lap. It's all good to help me gain expertise.

 

[kiev]

Several years ago i bought 2 OBC and 2 MCU thru a Japanese auction site The price was very low, but the shipping was a nightmare to arrange and the cost was very high. Then, when i finally got them-- they were broken.

So from my experience i am skeptical of used parts from Japan, and would definitely recommend an inspection before installation.