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

[Rational]

The miev only has 88 cells, so the 117 makes no sense.

Thanks. I didn't realize I'd chosen the miev van profile originally. Definite oversight.

I've added the following note above to the posts with that 117 cell screenshot and will post more correct screenshots later. I wanted to keep the mistakes posted so others could read and learn in order to avoid them as well. Thanks for you and Mickey pointing that out.

Later Edit: My i-miev does NOT have 117 cells. I had the wrong profile chosen in hobdrive. I had somehow chosen the miev van titanate profile which is INCORRECT. It should be the i-miev LI-ION profile which will show the correct 88 battery cells. So this screenshot for hobdrive below is not reflecting reality correctly within this post. I will post correct screenshots again in a later post below.

i know you drained your pack from Full, but does your car charge properly with no DTC?

Yes, I drained it down now to about 61-63% while it is sitting idle. But I've not tried to charge it up yet. In fact, I'm glad you asked as I was going to open up the MCU this morning, but will first try to charge it to confirm if this is working. I was assuming the charging was all working as it had a full 100% charge when I got it, so thought the prior owner had tried to charge it to see if that would fix it, but that didn't help. I'll test charging this morning before taking opening up the MCU.

The cover of the MCU can be removed without disconnect under the seat, if care is taken to ensure there is no HV from the pack. The only reason i mention is that many folks have caused an Airbag system fault when doing this, and that code can't be cleared easily. If the correct sequence is followed then the DTC won't occur, but more times than not folks get caught by it.

Oh, I was forgetting about that airbag fault I'd read others cause. I think people get caught up by this because in other forums I'd read that the seat needs to be pulled out completely. But read elsewhere here on this forum I think it was that you can just slide the seat all the way back and still get access to the cover. So that is a lot less work unbolting and risking an error from disconnecting that sensor. This is why I hang out here. You guys save me so much frustration and risk. Thank you.

You never answered my question about the fasteners on the OBDLink LX, but i'm happy to sit back and watch.

That question I must have missed. I'm usually pretty good about reading everything carefully but don't recall that one. Apologies. The screws on it? I'll try to find your question and reply.

Thanks again for this. It all helps.

Sadly it will be 3 days before the auction ends on the MCU we're bidding on, so I will lose some days here.

But I have ordered the best FULL CHIP version of the Lexia 3 cable as recommended by that all French language forum site. I'd done further reading and it really matters it turns out to get the best one as lesser quality versions of this cable don't always write properly nor work with all versions of Diagbox like the FULL CHIP version does. I'll share more on this later.

Thanks guys.

 

[Rational]

I once repaired a car with stuck contacts in the high-voltage battery. The HV did not turn off. It is necessary to check.

Okay. Good to know. Thanks.

I'll definitely be using the multimeter on the incoming battery leads to the MCU before continuing to work in there after pulling the main fuse. Always best to check and be sure. Thank you. Will do.

 

[Rational]

Oh, so to clean that back motor compartment of dust. You'd posted earlier I can just vacuum it and then use soapy water on it?

Safe to get lots of water in there and hose it all out? Or don't over do it with the water?

I wouldn't mind getting one good high pressure hose-down in there but don't want further problems cause by doing so should that be a thing. I've read posts about water getting into some pump motor was it?

Actually, I may just leave the deep cleaning until after it's all running and confirmed good.

That way I can move it and park it in an area where I can use lots of water and it can all run off down the drain easier then. Right now it's in the garage and I can't really do that there anyway...

 

[Rational]

Seems that car scanner says the OBD-II protocol used to talk to my 2011 JDM miev is ISO 15765-4 CAN. 11 bit ID, 500 kbaud.

So are these identifier numbers used for anything?

asdf

asdf

 

[Rational]

Okay. Correct readings now with the correct hobdrive profile chosen: Miev -Li-Ion:

88 cell volt BMU page

66 cell temp page

66 Temp Diagram page

MCU page

EV-ECU page

CHARGE page

BMU page


CMU6 page. So why does page 6 and 12 show only 4 cells vs 8 cells for all the other modules?


CMU1 page for comparison

 

[kiev]

Have you read thru the technical information manual? It is online at the same site as the workshop manual. There are 3 FSM selections: body, technical and workshop

88 cells and 12 modules..hmm, how is that done...if only there was a layout diagram somewhere

 

[vik013]

88 cells and 12 modules..hmm, how is that done...if only there was a layout diagram somewhere

10 modules of 8 cells and 2 of 4 cells

 

[Rational]

Charging works! I'll post some pics later.

Okay, I'm opening up the MCU cover now. Got the main fuse removed.

But this thing is glued tight! Any videos showing this? Can't find much anywhere yet.

All the videos online don't show how they got the cover unstuck from the sealant. But it's not that difficult now that I've done it.

I'll post some pictures and video soon on how to do it easily.

Now the black MEIDEN cover over the IGBT's is being a bear to pull off. Need to figure this out next...



He states that more is holding on that black cover perhaps underneath?

I found the relevant page in the online manual for removal: http://mmc-manuals.ru/manuals/i-miev/online/Service_Manual/2017/index_M1.htm

Seems I have to drain the coolant as well to get this out...

 

[Rational]

Have you read thru the technical information manual? It is online at the same site as the workshop manual. There are 3 FSM selections: body, technical and workshop

88 cells and 12 modules..hmm, how is that done...if only there was a layout diagram somewhere

Haha. You forgot the /s at the end. lol. But I got it. Okay. Thanks. I'll take the hint. I do need to read more. RTFM FTW!

 

[Rational]

Status update.

I was able to pull out the entire MCU today. Now I understand how it's all connected and put together. Turns out you don't have to take it out to swap just the IGBTs if you don't want to. It's a lot of work having to disconnect everything and drain the coolant.

If you just have to change an IGBT it's easiest to just keep it in place and remove the top cover. Then take out the correct bolts holding the black MEIDEN cover (Edit: It's not a cover! It's a high voltage capacitor! It stores at least 185v even after the main breaker is pulled so handle with gloves!) in place, then taking out a few more bolts holding the IGBT and copper bus bars blocking it.

This way you don't have to reprogram any VIN/chassis number as the bottom PCB board is not switched out.

All the HV power stuff is located in the top part. All the logic board stuff in the bottom. They are two separate compartments with a plate that has to be removed for access from either the top or bottom depending on what you need to work on.

Getting the top plate off is not as hard as it first seemed. I've made a quick summary video here for others in the future.

 

[Rational]

Alright. Update #2.

Here is a longer version of the MCU removal process.

I'd not seen a comprehensive video online yet for this for the i-miev MCU removal. So here is a general overview of how to remove the cover and all the cables as well as drain the coolant hose.

FYI! You DON"T have to take out the entire MCU in order to just change an IGBT. I was wrong about that in this video.

You can leave the MCU in place and remove the correct bolts in order to get the MEIDEN black cover off and swap out the IGBTs underneath.



Quick overview of the MEIDEN Shizuki N71P49691 800uF 450VDC 2490031 (capacitor?) first removal.


***Note: I just realized when posting this and finally looking at the part number on it that it may be a HUGE capacitor!? Whoops. That could have been a big shock. I need to be more careful next time. Some other guy called it a 'cover'. It's not just a cover. I think it's a huge capacitor. Need to test the voltage on it to see if it's not really drained yet...

Also, IGBT #1 I believe is the middle one after looking at it closely. My middle IGBT was the one that was burned. I'll share in the next video the details of the failed IGBT.



Here's the IGBT taken out. Showing how to take out an individual IGBT and looking at the allen bolts and thermal paste. More closeups of the damage:


Hope this helps the next guy.

 

[Rational]

https://myimiev.com/threads/mcu-20a-450vdc-fuse-investigation.4310/

Just found this thread when searching for "N71P49691 450VDC". It was the only search hit on "MEIDEN Shizuki N71P49691 800uF 450VDC 2490031".

So yeah, it's a capacitor. Kiev shared this in that thread.



And there I was just tossing it onto the metal MCU cover like it's nothin'. Not smart.

I have to guess it was discharged as I followed the manual's recommendation to wait 5 minutes after disconnecting the aux battery negative terminal before removing the main traction battery fuse. Hopefully there was enough time to deplete this capacitor before I got in there an hour or two later.

In other news, I did confirm continuity of the main MCU fuse as soon as I got the cover off. It's good.

 

[Rational]

@kiev Finally time to share the pics I'd promised of the OBDLink LX.

I know you wanted to see the fasteners on this unit. Here you go. Look like the real mccoy?

I couldn't find a small enough star nut wrench to open it up yet to get pics of the pcb and chips. Will do though.

 

[MickeyS70]

I have to guess it was discharged as I followed the manual's recommendation to wait 5 minutes after disconnecting the aux battery negative terminal before removing the main traction battery fuse. Hopefully there was enough time to deplete this capacitor before I got in there an hour or two later.

There is a discharge resistor that takes care of residual voltage, however the only way to be safe is to check any terminal before removing components inside the inverter.

Never assume/guess anything when working in any HV environment!!

 

[kiev]

i'm glad to hear that charging works okay, one less thing to worry about.

That 2011 MCU looks like the same as the 2012 units.

i didn't realize that you were going to dig in to the MCU or could have saved you some trouble with tips, but now you will be an expert with first hand experience.

Good use of the search function--it can be handy to find details on technical stuff that has already been discussed and posted. i use it all the time to "remember" old posts and threads.

There is a threaded hole for a jack screw to remove the lid of the MCU, located on the right hand side up near the access cover. Just thread one of the screws into that hole and it will push the lid up against the fat flange on the housing.

i started a troubleshooting and repair thread for the MCU but it got overtaken by events. My intent was to collect all the relevant circuit board info there, but the P1A15 troubleshooting thread has most of the details for that main DTC issue.

The OBDLink LX looks good, no need to take it apart unless you just want to. There have been counterfeits sold that mimic the color and labels, etc.

p.s. i'm not sure what your reasoning is for removing/swapping the bottom control board? The damaged IGBT is likely the only thing needed to be replaced. The damage was likely due to towing and the motor got stalled by the heavy load.

There is no "neutral" for the 3-phase motor. There is A,B,C or U, V, W or whatever scheme for labelling the 3 motor windings.

The control board has the switching circuits for commanding the gates of the IGBT, hopefully it was not affected, i didn't notice any blown parts damage in the video.

Also the eeprom is located on the control board, so swapping one out will fuss up the VIN matching with the EV-ECU, i.e. car won't go to READY.

 

[Rational]

Never assume/guess anything when working in any HV environment!!

Yep. I'm raising my level starting today. Can't ever have a mistake with HV. No assumptions from here on out.

 

[Rational]

Good use of the search function--it can be handy to find details on technical stuff that has already been discussed and posted. i use it all the time to "remember" old posts and threads.

Thanks. Same here. I learned about the power of search when using discord. Most people don't use it or realize it exists and end up asking the same questions over and over and over again. Not me. Search speeds up how fast I get answers and lessens the work on others.

There is a threaded hole for a jack screw to remove the lid of the MCU, located on the right hand side up near the access cover. Just thread one of the screws into that hole and it will push the lid up against the fat flange on the housing.

Oh. Didn't realize! I'll try to find it for next time. Thanks.

i started a troubleshooting and repair thread for the MCU but it got overtaken by events. My intent was to collect all the relevant circuit board info there, but the P1A15 troubleshooting thread has most of the details for that main DTC issue.

I think it would be great to have a dedicated MCU thread. I think I'll start one and add in my refined posts and videos for others to benefit in a single thread.

i'm not sure what your reasoning is for removing/swapping the bottom control board?

My reasoning is that I've already removed my whole MCU. If it were still in the car, just doing the IGBT is less work. But since it is out already I think replacing just the communications board on the underside is actually less work. That way all the power stuff is left completely untouched and pristine as well. No dust or gunk getting into the IGBTs. Not that it really matters anyway, but just cleaner and less work I think doing it this way.

I am curious if with these JDM chassis number mievs if they might not be VIN encoding the boards? If so, should I try just doing a complete swap with the replacement MCU? Or does that potentially mess something up with the CAN bus that is hard to reset? I'm curious and would like to know if the JDM stuff can just be hot swapped. Worth a test? Might learn something for everyone's benefit.

Then if it does throw an error, is there a way to pull CAN logs or eeprom data somehow to further learn and research while I have it in the car?

I could then swap in my original controller board and gather logs again for further comparison? Now's the chance. Just need to know the steps to gather logs and eeprom data if someone knows how.

There is no "neutral" for the 3-phase motor. There is A,B,C or U, V, W or whatever scheme for labelling the 3 motor windings.

Oh. I was conflating AC and DC terminology? I see. Yes, makes sense now that you mention it. These EV motors are just like on my CNC machine. U,V,W with a VFD speed controller (MCU) basically.

Also the eeprom is located on the control board, so swapping one out will fuss up the VIN matching with the EV-ECU, i.e. car won't go to READY.

But is this a temporary error that goes away if I put the correct controller board back in the car? Still require a reset with hobdrive or car scanner?

 

[kiev]

Knowing that if a VIN didn't match and was an issue that could potentially brick the car, well it might be a good idea to avoid doing anything to create such a condition if at all possible.

Nobody knows the code for the operating system of these cars; nobody knows the code for the MUT tool or what are it's capabilities and limitations.

We do have one hint, found in the FSM for swapping out various ECUs and the battery pack: the MUT tool is used to read certain data, save it, and then write it back into the new/replacement unit.

What's odd is that we have never seen a Japanese version of the FSM; surely something like that exists, but is it identical to the rest of the world version?

 

[kiev]

Also the eeprom is located on the control board, so swapping one out will fuss up the VIN matching with the EV-ECU, i.e. car won't go to READY.

As mentioned earlier, The VIN is stored in the eeprom on the control board (bottom plenum) of the MCU.

 

[MickeyS70]

If I were you I’d try and get the car up and running before doing anything else. You’re right, the quickest way forward is to change the faulty IGBT (maybe all 3 for good measure), there is a good chance this will fix the issue.

Based on a recent case of temporarily swapping coded EV-ECUs between cars, chances are slim that a VIN conflict will brick it permanently, you’ll get a relevant DTC and the car won’t (?) go Ready.

MCU faults seem rare, the two instances with blown IGBTs seem to be caused by overloading the circuit.

My motto generally is ‘never to mess with a working system’ but if you’re willing to investigate further, wait until you have your Lexia/Diagbox setup ready and then try out the ‘cloning’ function of ECUs as @kiev mentioned.