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

[kiev]

Did you just use the circuit as shown in the picture with a fuse on the HV input?

i was wondering if it needed a resistor to limit current into the viper, like the other application version not using a fuse-resistor?

Did you measure any R, L values for the inductors?

 

[coulomb]

Did you just use the circuit as shown in the picture with a fuse on the HV input?

Yes.

i was wondering if it needed a resistor to limit current into the viper, like the other application version not using a fuse-resistor?

It's a worry, but it seems to me that these Vipers are designed to work with high voltages, so they should be OK. I'll find out in a few days.

Did you measure any R, L values for the inductors?

No, I don't have inductance measuring equipment, and I didn't measure the resistance, sorry.

 

[kiev]

And some more of my rambling thoughts about this circuit,

i'm wondering if the hidden diode might be a Zener?

And if Rf is meant to be a filter resistor on the input, such as Rf * C601, followed by another filter, L601 * C602 ?

And if L603 *C104 is a filter on the Input leg to the 5V regulator and C105 is on the Output side?

 

[coulomb]

i'm wondering if the hidden diode might be a Zener?

It's possible. But I think that the diode has to be pretty fast, and moderately high powered, so I think that the regulation is via extra resistors on the FB pin that I haven't shown or traced.

And if Rf is meant to be a filter resistor on the input, such as Rf * C601, followed by another filter, L601 * C602 ?

Yeah, that makes some sense. Also, it's probably a good idea to have some damping for the LC circuit, so it doesn't ring. I might put in the lowest value I find in the application notes, rated at 1 W for safety and voltage withstanding.

It's not too late to introduce a resistor. But I have no idea how to choose one. Also, a quality resistor would mean that the fuse would never blow. 1 A is the lowest I could get. It's probably a reasonable idea in case the resistor fails low resistance.

Edit: I've decided I'll add a resistor in series with the 1 A fuse, if only to limit the in-rush current when 360 V is suddenly applied across the capacitors. 360 V across 0.1 μF is only 6.5 mJ, but that's not insignificant. I'll use a 150Ω 1 W ceramic composition resistor, because I happen to have those in stock from Elcon charger repairs. That will limit the inrush current to some 2.4 A, with τ = RC = 150 · 10⁻⁷ = 15μs. In operation at say 1 mA, the voltage drop will be just 150 mV. A better value might be around 1kΩ.

Thanks for the thoughts Kiev, without the resistor I might have been in some trouble, perhaps not immediately, but after a number of starts. Recall that this circuit gets stressed every time the car goes to ready.

 

[vik013]

It's not too late to introduce a resistor. But I have no idea how to choose one.

Soon i will know the resistance of this resistor. I'm going to unsolder a working part.

 

[kiev]

@Vik, many thanks that would be a great help to understand this circuit.
Was the resistor in your picture blown--it appeared intact but can't tell by looking? Were your cracked capacitors reading open or shorted?

 

[kiev]

@ coulomb, take a look at Phil's board at the rh end--somehow that circuit looks familiar even though this is a different mini-waffle, IC305 looks like a Viper, also take a look at D315 or 311 or 314--one of those might be the missing diode on your board? i would guess a big one.

 

[coulomb]

@ coulomb, take a look at Phil's board at the rh end--somehow that circuit looks familiar

It sure does! Thank you so much! I never thought about the other modules and how they get their power for isolated measurements.

IC305 looks like a Viper,

I'm sure it is, with pins 5-8 and 1-2 commoned. And L301 looks identical to L601.

also take a look at D315 or 311 or 314--one of those might be the missing diode on your board?

Yes, D314 looks like it's the "missing" diode, and it appears to be bigger than the ones on the output module. I can't read the marking codes; it might be K8 or KB on one line and D6 on the one below. Possibly D311 provides the Vcc feed to the Viper, along with the nearby capacitors. If that's all there is to the Vref circuit (pin 3), it's very simple and clean.

But most important of all, the value for R389 is clearly readable! It will hopefully be the equivalent of the blown R601, code 101 = 100Ω. It seems smaller in size to R601, and the clearance for DC arcs is shockingly small, so let's hope that C334 and C335 hold up better than C601 and C602. My guess is that this Viper doesn't run on main battery volage, which is not available in this rectifier module, but rather from rectified mains, which peaks at some 340 V (or half that for 120 V charging), but has a lower RMS value (240 or 120 V). But instead of running all the time when the car is in ready, this will be present only when plugged in for charging, so I'm hopeful that these capacitors will last much longer than their equivalents in the output module. It's something to keep in mind as a potential fault pattern, though, if mains measurement stops working in future repairs.

It looks like this module has two analog measurements, and two digital outputs (via opto couplers PC303 and PC306). IC302 will of course be the PFC chip. I read UC285480W, nothing like the L4981AD PFC chip on the Elcon chargers. I see that the 100Ω resistors between the isolation amplifiers and the outside world are on top of this board (R329 etc). Maybe they were able to fit all the parts on one side of this board.

Thanks so much for your thoughts, Kiev! It validates adding the 150Ω resistor in series with the fuse, though of course @vik013's measurement on a known-good output module would be even better to know.

Edit: here is the direct link to Phil's photo on Imgur: https://i.imgur.com/guRRm5e.jpeg . That makes it easier to try and read marking codes, etc.

 

[vik013]

https://imageup.ru/img238/4476791/out-plate.jpg.html
330 Om

 

[kiev]

Good job there vik, thanks for sharing the picture and value.

 

[coulomb]

330 Ohm

Thanks a heap! I wish I'd used two 150Ω resistors now. Sigh.

 

[coulomb]

When re-assembling the charrger, I noticed a familiar-looking inductor (L402 below), and low and behold, there is a third Viper chip on the third module. I took these photos without taking out the module, using a mirror and bare phone out of its case (to squeeze it in to the tight spots):

The part in the bottom left corner above that looks like it might be designated F401 is actually L401; I suspect that there is a reflection from the front of the mirror glass.

So that's 3 Viper chips from 3 modules. I'm guessing that this one operates on the "DC bus" of the charrger, i.e. the boosted output of the PFC stage. So this would operate on a higher RMS voltage than the one in the PFC stage (perhaps 400 VDC versus rectified and possibly mildly filtered 240 VAC for the PFC stage). But like the PFC stage, this Viper chip and its associated capacitors would only be operating when the car is plugged in for AC charging. Hopefully this will be a lot less stressful than the output stage which operates every time the car is in ready state. But that's three modules with the potential cracked capacitor failure mode.

I don't see an in-rush resistor here; it's possibly below L401 buried in the soft potting compound.

Strange that they cantilever this thin PCB out of this module rather than burying it like the others. It seems like it would be subject to vibrational stresses.

 

[kiev]

Good catch to make all these findings.
Like in a blues song lyrics "can't have just one woman gotta have three" the Vipers are all over those boards.

Phil's pictures are useful again, we just didn't know what we had at the time; L401 shows up on the left board here. The other boards may have been too crowded to do the vertical trick, but i like it better than the epoxy-buried version for service work.

 

[coulomb]

L401 shows up on the left board here.

Actually, this particular photo shows the Viper chip; the inductor must be on the other side. I'll go chase up the other photos soon.

The other boards may have been too crowded to do the vertical trick,

I noticed that the charrger got hot mainly in the top right hand corner, where that PWM module is located. Maybe they found that the control board too close to the heat source was failing too often.

The charrger worked at this afternoon's in-car test:

It was very satisfying to see it working after all this time. Do I get my own entry in the index now? I think that this saga started roughly here.

Edit: BTW, I was surprised to see that this 2010 charger worked with a proper EVSE. I thought it would only work with a cord that connects the mains directly to the J1772 AC input pins. Do American 2010 iMiEVs work from an EVSE? If so, do they obey the control pilot signal?

 

[kiev]

Good job on the repair, glad to hear it is working.

The US didn't get these cars until 2012. It's interesting that your car will charge by those types of EVSE cables. Not sure anyone has tried just the direct AC cable over here.

Piev (Paul) had a very nice 2012 in silver such as yours until yesterday when he was hit FR corner and the car is a total loss by insurance standards. This was his development and test car for an updated battery with a new ~90Ah pack.

 

[Matias]

Hi.

I drive one Mitsubishi I Miev 2011 (in Norway) and face these days limited charging time. Charging a car at home using a type 1 conector is not a problem, as well as using a Chademo conector on a public charging stations.

But issue is I can not charge the car near my jobb, using a public charging station with type2 - type1 cable. As I investigated a potential source of the problem, it seems that car is programed to be charged only in afternoon/evening hours with this type of connection.

Anybody can help with info how this can be fixed?

BR Matias

 

[Tokarsky]

https://ibb.co/3BGRzSL
https://ibb.co/WgzPjbG

 

[kiev]

Great work, that is a really clear x-ray image of the H-bridge switching section from an early generation OBC,

 

[kiev]

...
But issue is I can not charge the car near my jobb, using a public charging station with type2 - type1 cable. As I investigated a potential source of the problem, it seems that car is programed to be charged only in afternoon/evening hours with this type of connection.

Howdy and welcome Matias,
Unfortunately this is a new issue that i have never heard before. Do you have a remote controller that can send commands to the car? Those can be used to program a timer for when to start charging (after being plugged in). Maybe Joe or someone else has experience with this to share.

 

[coulomb]

As soon as the top board is unscrewed from these charrgers (2012 onwards), you have to deal with a flexible ribbon that connects the top and bottom boards. Conductors on this ribbon are spaced 0.5 mm (less than 20 thousandths of an inch) apart, so the connector is a delicate, precision part. The ribbon carries vital signals, such as coil drive for the relay; if this drive is lost, the pre-charge resistors try to carry the full mains power and fail open circuit. I've seen too many of these high density connectors damaged. It was not obvious to me how to disconnect them without damage. Here is the place on the connector to apply a small screwdriver, outlined in yellow:

Pressing on the area outlined in yellow seems to lift the left part of the clip, allowing it to clear the barb outlined in red. To unclip the ribbon, apply a small screwdriver as shown below, pushing down and towards the ribbon:

The ribbon should move away from the white connector with very little force. If you need more than a little force, you're not doing it right, and will likely damage the connector / clip.