The Troubleshooting and Repair for On-board Charger (OBC) Thread

[piev]

Fast charging still has the following symptoms:

"

Before the spark the charger was displaying an error which said "Charging stopped by the vehicle" (some clicking)

After the spark it said "No connection to the vehicle, check cable".
(Complete silence)

"

Did you inspect the charge port and the EVSE connector. Ensure there is no corrosion or debris in the connector? Maybe try another EVSE?

 

[kiev]

Troubleshooting Chademo issues[Starting a troubleshooting thread for Chademo; moving this discussion over there]

Looking back at your original post:
So the spark occurred while in the process of doing a Chademo charging session--the Level 3 connector was in the port when the pliers shorted?

The car starts to READY and drives okay?
The car charges using the OBC with the Level 1 or 2 EVSE port?

If the answer is YES to all of these, then the damage is likely not to the OBC but to the Chademo circuits.

 

[Rass]

Did you inspect the charge port and the EVSE connector. Ensure there is no corrosion or debris in the connector? Maybe try another EVSE?

Yes, yes. Still the same.

Looking back at your original post:
So the spark occurred while in the process of doing a Chademo charging session--the Level 3 connector was in the port when the pliers shorted?

The car starts to READY and drives okay?
The car charges using the OBC with the Level 1 or 2 EVSE port?

If the answer is YES to all of these, then the damage is likely not to the OBC but to the Chademo circuits.

Yes, during starting of the charge. I pressed start, went over, fumbled, spark. Might have been during the switching. So ... First few seconds.

Yes yes, the error is in the chademo circuit.

should I move this to chademo charging troubleshooting thread?

The reason OBC was intermittent was because of the EVSE plug latch not grabbing fully...

 

[stoatwblr]

The ZNRs and the SA look ok to me, no visual signs of damage
Hires 3

I am a bit reluctant to change lots of stuff not knowing if it causes any issues but it is probably the only way forward dealing with intermittent failures. I should have checked the error codes before starting the repair. Tried it but couldn’t read it. Found out only now that the full version of hobDrive can read those. But putting everything together now is too much work.

Anyone have the part number of good snubber caps? Maybe buy those and a relay anyway if I decide to get into the desoldering of WP I will have the parts already. What do you use instead of the black rubber stuff to secure the components in place and protect from vibration? Hot glue is probably not good for caps. Silicon? Or some other special stuff I need to order?

 

[stoatwblr]

"but it is probably the only way forward dealing with intermittent failures."

Putting my "seasoned technician" hat on...

In my experience virtually all intermittent faults are down to bad connectors(*), hairline cracks on traces or dry joints

I've only encountered genuinely intermittent devices a handful of times and generally they were being heavily thermally cycled (which leads to metal fatigue and internal/external connection fractures). Most of the time the actual fault has been in the "how those devices are connected to the rest of the circuit" part of the equation. A magnifying glass is handy at times, as is metering individual circuit traces with a decent tester

(*) Corrosion on pins, poor mechanical contact or bad crimps (which are exceedingly rare on factory harnesses)

WRT the snubbers: Hot glue works fine on electronics but in this instance if it gets hot you'll lose the mechanical support if it softens. There are non-acid cure silicon sealants suitable for use with electronics (look for electronics silicon potting compound)

 

[Piki]

"but it is probably the only way forward dealing with intermittent failures."

Putting my "seasoned technician" hat on...

In my experience virtually all intermittent faults are down to bad connectors(*), hairline cracks on traces or dry joints

Thanks for the advice. I did check for dry solder and bad joints but will check it again. As I said, I opened the OBC just to check if some obvious failure was there. Though blown X2 capacitors should in theory do not lead to the charger not working, it doesn’t feel right to just close it and look somewhere else. Maybe it does cause some current leak through the caps when the charger warms up. Maybe the wall charger or the OBC detects the high EMI and shuts off. The wall charger is a beefy original one with lots of electronics inside, it has to do something else except of closing the relay and do some simple communication. Maybe it was just that case corrosion from the humidity entering through the plastic plug that made some semi-conductive path. There was some deposit on the board which I cleaned, maybe it was causing the issue. I will probably just resolder the X2 caps and give it a try. Will have a chance to check for error codes so I don’t search blind.

 

[ceskimiev]

Good Morning...

Thanks Kenny for post #1,215, forced me to check all pins in the waffle board connections.

I'm happy and sad i believe I've found the issue. I got a short between pins 25 and 26 plus from there to T4A from the schematic on post #1,197 (thank you LDTVN, left a like). Since I'm not getting shorts on any base-emitter junctions for any of the other bridge IGBTs, I think is safe to say something is shorting on that quadrant of the bridge. May someone please confirm before I go buy a desoldering station and possibly the IGBTs? I haven't found yet a replacement charger and I definitely do not want to pay +3k$ for a new one.

Thanks!

So I've dismantled the charger with the intent of removing the waffle plate after getting my desoldering station. This is a b... to remove! So far no success. I've regretfully seen something else that concerns me: one of the IGBT exploded, taking the epoxy out and it looks like the epoxy took out some resistors and (hopefully not) traces out of the PCB. The whole came out as a chunk that was rattling between the board and the waffle plate. Doesn't look too promising.

 

[kiev]

That would certainly be interesting to see.

Desoldering the WP does take quite a bit of heat to the pins. i alternate between sides in order to not put too much heat locally, but it may not matter. Just get the joint liquid and let the vacuum get it off. i stand the board up on the bench or even slightly leaned over to let gravity assist

Another member successfully replace parts inside a waffle plate; circuit traces can be repaired/replaced using 30 AWG wrapping wire.

 

[edsonedson]

I come from the my nissan leaf forum hoping to find some information on what other people have done to fix their waffle plate. Apprently the OBC on the 2011/2012 Leaf is similar to the one in the imiev(not sure what years). I opened up my OBC and found a blown capacitor and what looks like some pieces that came off the waffle plate. I have yet to unsolder the waffle plate from the PCB, will attempt to do this next week. Has anyone here fixed a waffle plate before? how do you remove the epoxy without damaging the internal components?

 

[kiev]

Howdy edson,

There have been a few repairs of the waffle plate, the hardest part is cutting thru the black epoxy.
Here is a post with a schematic of the leaf WP:
#1,197

There is another post with pictures of a successful machining of the epoxy but i'm not finding it just now. It seems like he was using dental tool bits. The epoxy doesn't "cut" like normal materials and quickly dulls the edge of cutting tools.

 

[coulomb]

There is another post with pictures of a successful machining of the epoxy but i'm not finding it just now.

I thought I'd find it for you... no dice. I can't find it with Google or with archive.org. This could get annoying.

I don't have time to chase this up just now, but something has to be done.

 

[coulomb]

@kiev, I found it, one of the first links (search for "Aug 2021" without the quotes). Perhaps add some more text to the link to say that it's about repairing a capacitor in the epoxy of a mini-Waffle plate (2010 model). So this repair was not to the large Waffle Plate, but I imagine that the techniques are the same:

https://myimiev.com/threads/trouble...-on-board-charger-obc.4079/page-40#post-43621
Alas, he never answered your (kiev's) question about what magic chemicals or techniques were used.

Also, I came across this post with the gorgeous photo of the IGBT's part numbers; that should be findable from the index:

https://myimiev.com/threads/troubleshooting-and-repair-for-on-board-charger-obc.4079/post-44938
It seems that the links from the index do work OK; the confusing thing is that the new forum software numbers the posts sequentially in each thread, but the old forum software gave every post a unique number, so adjacent posts in one thread could be tens or hundreds apart in "post number", depending on how many posts were made to other threads. But the anchors in the posts are recorded correctly with the old forum's numbering. Just don't expect the post number given by the new forum to have any relationship to the #post-nnnnn parts of URLs.

Re the links from the index: the pages with a lot of images jump around a lot and don't stop at the right place, at least on my laptop running Chrome browser for Windows. But you can just copy the link and open it in another tab, and once the images have loaded into cache, it does jump to the right spot. So maybe the index isn't as big a nightmare as I first thought; perhaps it just needs a few more well-chosen keywords.

 

[kiev]

Good on ya to find that, it was a @philsuth repair.

In hindsight using page numbers for reference wasn't such a good idea--probably using the post # would have been better, but it seems like they weren't as readily available in the old forum format, or maybe i'm mis-remembering...

 

[ceskimiev]

That would certainly be interesting to see.

Desoldering the WP does take quite a bit of heat to the pins. i alternate between sides in order to not put too much heat locally, but it may not matter. Just get the joint liquid and let the vacuum get it off. i stand the board up on the bench or even slightly leaned over to let gravity assist

Another member successfully replace parts inside a waffle plate; circuit traces can be repaired/replaced using 30 AWG wrapping wire.

c

That would certainly be interesting to see.

Desoldering the WP does take quite a bit of heat to the pins. i alternate between sides in order to not put too much heat locally, but it may not matter. Just get the joint liquid and let the vacuum get it off. i stand the board up on the bench or even slightly leaned over to let gravity assist

Another member successfully replace parts inside a waffle plate; circuit traces can be repaired/replaced using 30 AWG wrapping wire.

You wanted to see... Here's how I found my waffle plate. Two blown IGBTs, out of the epoxy. Silver lining: i have less to dig. There is more damage, as even with these not there, I still have the same shorts. Let the digging begin!

 

[edsonedson]

Thanks for links, so far I've only removed the flux from pins, desoldering them will be a pain! I'm going to get a Hakko FR301 to make it a little easier. Although I noticed there's solder under the PCB as well, not sure if the FR301 will be able to suck that up.

 

[ceskimiev]

[Deleted]

 

[coulomb]

It seems that the links from the index do work OK;

I typed too soon. I tried to find the snubber caps specifications; the link would not work. I think maybe some links near the top have been changed to the new forum's format?

Here is the new link for "optional caps": https://myimiev.com/threads/trouble...-on-board-charger-obc.4079/page-12#post-37542

If you happen to know the right page, you can search for the anchor in the page source (^U). The anchor is the part of the URL after the final #. Then find some unique looking plain text after the anchor, go back to the page and search for that unique text. Tedious, but automatable I think. But you have to know the right page first, or try several pages.

I thought maybe Google search could find the anchor even though it's not part of the visible text of the page, but alas it could not.

I guess the solution is to scrape a copy of all the html into a giant text file, and write a Python or other script to print out a list of links and their new addresses. One day when I have more time.

 

[ceskimiev]

Thanks for links, so far I've only removed the flux from pins, desoldering them will be a pain! I'm going to get a Hakko FR301 to make it a little easier. Although I noticed there's solder under the PCB as well, not sure if the FR301 will be able to suck that up.

I just finished going through this exercise and yes, it's really hard to unsolder. Whatever you do, do not pull, since it can remove traces and/or damage the board.

I followed the schematic on post #1,197 as to see which points could need more time: single signal points would require less time, while power carrying points (two or more pins with same signal together) will require more time (single points took about 5 seconds at 400 Deg, others took up to 15 seconds, must monitor to avoid burning the board). Select a nozzle that is just larger than the size of the pins. I also used rosin and resoldered more than once with a lower temperature melting wire (with lead, lead free requires higher temperature), kept the station tip and the area clean. Once the hole is clear of solder, I heated just the pins with the solder to dislodge them from the walls of the holes. Only once all the pins were free the waffle plate will come out.

I found out that it's best if you can get all the solder at the first go, partial would be left if you attempt extracting when not all the solder is molten, so patience is needed. I learned that too late in the game.

 

[kiev]

If the hole doesn't clear then add some 63/37 eutectic solder to wet and draw the solder up the pin from the bottom of the board to the top--put the heat of the soldering iron at the top of the pin and the solder will flow toward the heat.

Another option for stubborn joints or for big blobs on the top side is to use solder wick .

When desoldering i hold the board vertical or upside down somewhat such that gravity will assist getting the solder out of the joint.

Empty the suction filter frequently as lots of solder will block the flow path.

 

[ceskimiev]

Giving up on the waffle plate. Cannot do anything with the epoxy. Tried boiling, applying heat, hammering, acetone, nothing clears this beast. Plus, on further troubleshooting it looks like all the IGBTs of the H Bridge are defective, which is an extensive repair, would be much easier without epoxy.

For the more gifted, I think making these boards is a business opportunity, or perhaps making an off the shelf charger talk to the ECU of these cars. Saw these: https://www.evcomponents.com/elcon-tc-hk-lf-6600w-charger.html.

The dealer has no stock but gave me a price of $CAD 5900.00 for the 9499A171... hard pass.

I will put everything together and attempt to sell the car for parts, while keeping my ear on the ground for a replacement charger or a crashed I-Miev near me. It's a pity, everything else seems to work.

Thank you all for your great help!