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

Mitsubishi i-MiEV Forum

Help Support Mitsubishi i-MiEV Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
More progress. With a scapel and wooden scraper I managed to clear out a neat rectangular area to remove the old capacitors. I cut them off their own legs part way up which gave me about 2mm of leg left in the board to work with.

The caps I ordered have a slightly wider leg spacing than the originals but that actually worked out perfectly as I was able to wrap them around the pins in a U shape to create a good solder fillet joint:

uc


It's not until seeing them side by side that I realised just how much beefier these caps are than the original - they're 6kV rated instead of the 2kV of the originals, only slightly larger in diameter, but about 3 times as thick! :twisted:

I was able to solder onto the leg stumps at the top without melting or causing a dry joint on the soldered joints under the board - there is quite a large gap between the waffle plate and the PCB so you can shine a torch into the gap and see the original soldered joints to make sure they're not disturbed.

The gap is so large that I reckon if there were no stumps to solder to you could possibly insert the component right through the hole, apply the soldering iron to the component leg on the top side (with the caps sitting a little higher than normal) and apply solder to the underside of the board in the gap to the waffle plate... I considered doing it that way but decided to try soldering to the old stumps first.

I have some black high voltage high temperature neutral cure electrical RTV silicone that should be a pretty good match with the original - firm but still rubbery in case anyone ever has to go back in there again, however I'm not going to put it in until I've tested the relay.

So now I need to test the resistors and relay and components on the waffle board...
 
Those are great looking snubber caps and a super find to get 6kV rating. Awesome solder job too.

This isn't very pretty but has the measured diode drop voltages on a good waffle plate, with no caps or inductors connected with the fastons.

VbRyfIb.jpg
 
kiev said:
Those are great looking snubber caps and a super find to get 6kV rating. Awesome solder job too.
Thanks. :)

These are the caps I ordered, however I notice that the picture shown is NOT accurate as the ones I received were orange, not yellow, and a lot thicker:

https://uk.rs-online.com/web/p/ceramic-single-layer-capacitors/8313240/

I suspect the picture is from a lower rated capacitor in the same series.
This isn't very pretty but has the measured diode drop voltages on a good waffle plate, with no caps or inductors connected with the fastons.
When you say with no capacitors connected do you mean the electrolytics were removed ?
 
kiev said:
No just the flying leads from anything on a faston tab, (the leaf charger has a big cap on the fastons and uses the same waffle plate)
Ah OK. The board is right out with nothing connected to it, so I'm off to measure it...

Interesting that the Leaf has the same waffle plate - since the i-Miev launched nearly 2 years before the Leaf, does that mean the Leaf copied the i-Miev here ? :)
 
kiev said:
This isn't very pretty but has the measured diode drop voltages on a good waffle plate, with no caps or inductors connected with the fastons.

VbRyfIb.jpg
Thanks for that, I've been through and measured all the diodes on your schematic and they all measure healthy. :)

I've also checked the two 4.7 ohm resistors and relay using skyloggers method - the neutral path measures 9.6 ohms with the relay off and <0.1 ohms with the relay active. The relay has a satisfactory click and I clicked it on and off many times to make sure the reading always went to near zero ohms and it did.

The black mark on the 4.7 ohm resistor corner I mentioned before turns out to be a very thin film of silicone resin - it peels right off if I slip a blade under it with clean white resistor underneath so I can only assume that it is factory "over spray" when the resin was sprayed in, and therefore a red herring. There is no sign of heating of the resistor or nearby capacitor.

I've repotted the capacitors using the black high voltage silicone - quite difficult to smooth it out nicely because it's so incredibly sticky, also it needs one to two days to fully cure, however it should do the job nicely.

So the board is ready to go back in the car. The only thing holding it up now is I don't have the thermal paste (should arrive on Monday) and I don't have the MCU fuse yet either - which will probably arrive on Tuesday.

By the way earlier in the thread you said that your waffle plate had lots of thermal compound on it - mine was quite the opposite, with a very thin stingy application. So much so that significant chunks of the waffle plate didn't have any compound on it at all!
 
i get a old credit card and use it like a putty knife to spread the thermal paste as thin as possible and into any voids in the surface. same on the plate. Metal to metal is the best transfer, the paste is just to fill tiny voids where metal is not touching metal.

if you need to drive the car then you can make a temporary jumper to replace the fuse--just a short length of 26 to 20 AWG would work since it only carries about 10 Amps when charging.
 
The thermal paste arrived today so I have fully re-assembled the OBC unit including sticking the lid down with new silicone. (I ended up just using the same neutral cure stuff I used in the doghouse around the capacitors) So now I'm just waiting for the fuse which I hope will be arriving tomorrow. I got an email to say that it had arrived in the UK at customs this morning from Japan, so fingers crossed!

I think my advice to anyone attempting this repair is don't bother removing the OBC box from the car - it's a lot of extra work, especially if all the mountings are rusted, and means draining and refilling the coolant, disturbing all the external connectors etc and there really isn't any need to do this unless the actual DC/DC converter on the under side was faulty which is almost never the case.

Just unscrew the top lid in the car, (no need to disconnect the cables to the filter box on top) carefully prise it up (there is a small bead of silicon sticking the lid down so it will be reluctant to let go at first) remove the top board, then remove the main PCB. Make sure you label all the spade connectors as you remove them and take photos so you get them overlaying each other in the right configuration so they fit comfortably under the lid of the box. Be careful with the white ribbon cable on reassembly. You'll need thermal paste to reapply between the waffle plate and cooling channels on reassembly and be careful not to drop anything like screws down the holes in the corners otherwise you will have to remove the box to retreive them from the DC/DC compartment...

Taking the PCB out of the box and leaving the box in the car makes for a MUCH easier job IMO than either taking the whole box out, or trying to work on the top side of the PCB while it's still in the box.
 
Simon and Kenny, thank you both for such wonderfully detailed information about fixing the OBC and thus making it easier for those of us who will have to face this in the future. :ugeek: Wishing you all the best as the moment of truth arrives today when you fire up the car :!: ....
 
JoeS said:
Simon and Kenny, thank you both for such wonderfully detailed information about fixing the OBC and thus making it easier for those of us who will have to face this in the future. :ugeek: Wishing you all the best as the moment of truth arrives today when you fire up the car :!: ....
My fuse (I actually ordered two just in case) got caught in customs today and there are customs charges due before it will be released which will add a day or two delay. Argh!! Waiting for a part for my Petrol car as well which has turned out not to be the write-off that I first thought, but it still leaves us without any working car for the next few days... :(
 
Looks like the fuse is finally arriving tomorrow after paying customs charges... in the meantime here is a picture of the potted caps that I meant to post earlier:

uc


This is neutral cure high voltage high temperature electrical silicone which sets to a similar rubbery consistency to the original stuff. I dare those caps to move with vibration :D

When you're soldering onto the old existing legs (which are quite skinny) I think it's important the caps are re-potted for physical stability as they do wobble around a bit otherwise and would eventually fatigue and break the original leads. It wouldn't be as wobbly if you could remove the old legs completely and solder the new capacitor legs through hole properly but that would mean removing the waffle plate with the inherent risk of causing further damage.
 
Good news. :) The fuse finally arrived today and the rest was a bit of an anti-climax!

I fitted the fuse, put the cover back on the MCU, refitted the traction battery safety link, reconnected the 12v battery, plugged the car into charge listening for a bang but all I got was charging. :D I checked the 12v battery and it was charging at 14.5 volts so the DC/DC is working.

So back in with the seat over the link plug, reconnect the airbag and then reset all the fault codes with Diagbox. I then let it charge to 100%, ran some errands to the tune of about 40 miles and on the way home used a rapid charger just to double check the Chademo port was working which it was.

So it's all fixed. :mrgreen: Thanks to everyone in this thread - without those who have trailblazed this issue I don't think I would have been able to solve it by myself, at least not nearly as quickly! Most of the delay in getting the car fixed has been waiting for parts and materials and good weather! The actual repair itself could be done in a day, allowing a bit of time for the silicone to cure properly.
 
DBMandrake said:
Good news. :) The fuse finally arrived today and the rest was a bit of an anti-climax!

I fitted the fuse, put the cover back on the MCU, refitted the traction battery safety link, reconnected the 12v battery, plugged the car into charge listening for a bang but all I got was charging. :D I checked the 12v battery and it was charging at 14.5 volts so the DC/DC is working.

So back in with the seat over the link plug, reconnect the airbag and then reset all the fault codes with Diagbox. I then let it charge to 100%, ran some errands to the tune of about 40 miles and on the way home used a rapid charger just to double check the Chademo port was working which it was.

So it's all fixed. :mrgreen: Thanks to everyone in this thread - without those who have trailblazed this issue I don't think I would have been able to solve it by myself, at least not nearly as quickly! Most of the delay in getting the car fixed has been waiting for parts and materials and good weather! The actual repair itself could be done in a day, allowing a bit of time for the silicone to cure properly.


Hi DBMandrake, Glad to hear you're up and running again, Can you tell me what potting compound you used or even better provide a link for it? Think my board is a bit worse off than your one but again with the help of the great guys in this forum it should be sorted soon,
 
Easky15 said:
Hi DBMandrake, Glad to hear you're up and running again, Can you tell me what potting compound you used or even better provide a link for it? Think my board is a bit worse off than your one but again with the help of the great guys in this forum it should be sorted soon,
Sure no problem.

It's not a traditional "potting compound" as such - as they usually set very hard like varnish, It's basically just a neutral cure silicone rubber with a high breakdown voltage rating and reasonably high temperature rating. It has a very similar consistency to the original stuff - eg rubbery and easy to cut with a sharp blade or dig out with a small scraper.

I ordered it from here:

https://uk.farnell.com/acc-silicones/as1701-75ml/sealant-silicone-neutral-cure/dp/8497516

From the description:

The AS1701-75ML is a 75ml neutral flame-retardant thixotropic Adhesive Sealant is a specially formulated neutral cure silicone sealant designed for use with sensitive electronic assemblies. It is described as an alkoxy 1-part room temperature vulcanising (RTV) silicone sealant. The alkoxy cure system produces a silicone sealant with excellent adhesion to most common substrates. Black paste appearance with 3 minute tack free time.
As it's a cartridge (and an uncommon small 75ml size) I had to order a gun for it as well:

https://uk.farnell.com/acc-silicones/480002010/cartridge-gun/dp/4625183

A bit pricey but I was in a real hurry to get it so I could get the job finished and get the car back on the road. I'm sure something similar could be found in a normal squeeze tube dispenser at a cheaper price. Just make sure it's neutral cure and rated for high voltages. (Eg non-conductive!)

I also used some of the leftover for the new gasket on the top lid - I just scraped the bead of old sealant off the lid with an old credit card as well as the top edge of the box and ran a bead of the silicone around the box then tightened the lid down and let it set.
 
Hello. I have the same problem I've seen in this forum. The car has stopped charging. As I unscrewed it, I noticed that the R106 resistor is damaged. This is a resistor to limit the charging current of three 680 uF capacitors. After replacement for a working one, nothing has changed. After connecting the car to 230V, the fan cooling battery turns on for a moment, 350V appears on power factor improvement and in a second a yellow exclamation mark appears and immediately turns off Power factor improvement and the voltage begins to drop on 680 uF capacitors. There are no control signals on the T501 T502 transformers. Error P1A12 appears. I checked all the components that are in the big black cube (transistors and bridge rectifier.) I checked the relay. I noticed that there is a circuit that checks the voltage drop across resistors. All components are functional (D312 D311 D315 D316)

Did someone solve a similar problem?

I wonder why the relay turned off during charging (which caused the r106 resistors to burn)? this relay is controlled by a microcontroller located on the top board.
It is possible that power factor improvement creates too low voltage, and for this nothing further works? The capacitors are at 420 V, and it creates 350V. Maybe it should be 370V or 400V for example?
Diagram of the entire layout ( https://drive.google.com/file/d/1tHilCxdM7tmisq-a1U_R6NnYbFyOVV85/view )
 
DBMandrake said:
Good news. :) The fuse finally arrived today and the rest was a bit of an anti-climax!

I fitted the fuse, put the cover back on the MCU, refitted the traction battery safety link, reconnected the 12v battery, plugged the car into charge listening for a bang but all I got was charging. :D I checked the 12v battery and it was charging at 14.5 volts so the DC/DC is working.

So back in with the seat over the link plug, reconnect the airbag and then reset all the fault codes with Diagbox. I then let it charge to 100%, ran some errands to the tune of about 40 miles and on the way home used a rapid charger just to double check the Chademo port was working which it was.

So it's all fixed. :mrgreen: Thanks to everyone in this thread - without those who have trailblazed this issue I don't think I would have been able to solve it by myself, at least not nearly as quickly! Most of the delay in getting the car fixed has been waiting for parts and materials and good weather! The actual repair itself could be done in a day, allowing a bit of time for the silicone to cure properly.

Very well done SBMandrake, bravo! I am happy for you that your car is on the road again.
So what do you think about the cause of such fault? An AC swell, surge due to poor quality of your AC 240V grid?

If the blue caps got that swell, they did their job (becoming a short circuit) protecting the rest of more expensive component and blowing the fuse.

But have you replaced them with more robust ones with higher voltage treshold? If so don't you think next time they will not protect the other expensive components?

Or is the fuse that blowing for any mechanical? reason like vibration, opening the circuit so suddenly to create a swell that triggered the blu caps?

Please kindly let me / us know your view on what could be happend in your OBC, thanks in advance.
Sandrosan
 
Here is a simplified functional schematic of the OBC, MCU and Pack, showing details of the HV DC Output section with the snubber caps and MCU fuse that seems to blow.

During charging there is about 10 Amps of current from the OBC to the Pack thru the MCU 20A fuse.

The 20A fuse in the OBC has never been found to blow even though it is in the same current path as the MCU fuse?

zFwVgUm.png
 
"The 20A fuse in the OBC has never been found to blow even though it is in the same current path as the MCU fuse? "

How/where is the OBC 20A fuse mounted ? you had hypothesized previously that the MCU 20A may suffered from mechanical vibration. I don't know where the OBC 20A fuse is.
 
Back
Top