skylogger
Posts: 74
Joined: Sun Dec 14, 2014 3:54 am

Re: Troubleshooting and repair for On-board Charger, DC-DC Converter

Tue Aug 07, 2018 4:57 am

Here is a drawing showing the signals from the rest of the car through the E-03 Connector outside of the charger box, to the CN101 Connector that plugs onto the top PCB. I have two chargers, one a Late 2010, and one an Early 2010.
The CANBUS Comes in on E-03 pins 6 and 13 and runs to CN101 PINS 10 and 9.
One differences between the two versions of 2010 Chargers are Pin 2 of CN101 Is a orange wire in the Late 2010, while Pin 8 is the orange wire on the Early 2010 (moves from top to bottom row of connector)
One strange thing is that the 12v supply from the battery on pin 2 of E-03 connects to CN101 PIN 12 which is a violet wire on the Late 2010 model but the scary bit is that this 12v supply connects to a black wire on the Early 2010 charger.
Pins 1,3, and 5 do not show up on the Maintenance manual schematics.

So connecting 12v to CN101 Pins 12 and 7, and connecting ground to Pin 6 of CN101 Should power up the PCB.

The remaining question is what to do with the CHGP signal CN101 PIN 3 (E-03 PIN 12) which connects to the EV-ECU. Not sure if this is a input or an output signal.
Also Pin 9 of E-03 ( CN101 Pin 5 ) is a Grey wire on the Late 2010 version that connects to PIN 3 of the AC Charging port.
I'm not sure if this is the Pilot Pin or the Proximity pin. In the Early 2010 Charger, this wire is not populated in the connector.

Image

kiev
Posts: 815
Joined: Sun May 03, 2015 7:15 am
Location: The Heart o' Dixie
Contact: Website

Control Board Power Supply

Tue Aug 07, 2018 5:00 am

The control board, aka top board, is a multi-layer construction with internal traces.

It appears that a 12V supply power enters on pin 12 of the control board, is filtered by caps and coil, then passes thru a 4A fuse and a barrier diode before exiting on pins 5-8 of CN1 to provide 12V supply to the bottom board.

top layer
Image

bottom layer
Image
kiev = kenny's innovative electric vehicle

kiev
Posts: 815
Joined: Sun May 03, 2015 7:15 am
Location: The Heart o' Dixie
Contact: Website

Re: Troubleshooting and repair for On-board Charger, DC-DC Converter

Tue Aug 07, 2018 5:20 am

@sky, great job on tracing the cable and noting the differences.

Looks like the 12V supply for the bottom board is thru the relay and is switched on by the EV-ECU, so part of troubleshooting would be to verify that the relay is coming on and 12V coming to the box.

http://mmc-manuals.ru/manuals/i-miev/on ... 101ENG.HTM

How about that fuse check for continuity?
kiev = kenny's innovative electric vehicle

skylogger
Posts: 74
Joined: Sun Dec 14, 2014 3:54 am

Re: Troubleshooting and repair for On-board Charger, DC-DC Converter

Tue Aug 07, 2018 5:37 am

When I first checked the fuse, I thought it was blown, but then I scratched the end/pads because they looked like conformal coating,
and once I cleaned up the end, it showing 0R so the fuse has not been blown on my unit. That conformal coating can throw you off if not cleared off.

kiev
Posts: 815
Joined: Sun May 03, 2015 7:15 am
Location: The Heart o' Dixie
Contact: Website

top board 12V power

Tue Aug 07, 2018 6:41 am

[edit] But first, does your top control board(s) look like the same general layout as the one i posted?

i think if you are careful you could verify that 12V is coming in to the control board on the 2 supply lines, the one that is powered all the time and the one from the relay contacts. The switched power could be measured at the little test point near the 4 vias circled in red near the bottom of the picture of the control board top layer posted above. If no voltage there then the bottom board is not getting 12V power.

Your measurement between pin 3 and the tab of TR310--was that a dc measurement or ac on the scope? If working properly then that signal would be pulsing between 12V and ground at the pin 1 gate frequency coming out of IC702 on the bottom side, the flyback switching regulator chip. If the regulator was not working and 12V was getting to the bottom board then i would expect to read 12Vdc.
kiev = kenny's innovative electric vehicle

skylogger
Posts: 74
Joined: Sun Dec 14, 2014 3:54 am

Re: Troubleshooting and repair for On-board Charger, DC-DC Converter

Tue Aug 07, 2018 9:35 pm

Hi KIEV:

Yes, my TOP Board appears to be same Version/revision as your picture.
Starting with the charger on Bench with nothing else connected, I've made up a cable and I'm supply 13.0v from a motorcycle battery to Pins 12 and 7 of CN101 with ground connected to pin 6.
I still have the wires soldered to the TAB and PIN 3 of transistor TR310 on the bottom PCB, and I am measuring 12.66v there between these points. I am also measuring 211ma from the 12v supply battery.
On the top PCB, I've also measured the following voltages across the electrolytics:
C702: 12.52V
C706 5.18V
C707 16.75V
C704 16.75V
C737 5.19V
C701 12.46V

while having the 12v battery connected to the pin on the CN101 Connector, I've also tried applying 240vac to the mains screw terminals on the top pcb, and checked to see if there is any AC at the primaries of the transformer that are on the ouptus of the IGBT's. This is still dead.
We probably need to work out what signal needs to be on the CHGP Line from the EV-ECU to pin 3 of CN101. This is probably the signal that actually enables/disables the running of the IGBT's.

I'm also having problems working out the The PE and CP Proximity and Pilot pins from the j1772 connector to the charger for bench testing.
I think that PIN 4 of the J1772 is an output to EVSE to tell it what charge status is on car based on change in resistor values. This was not fully implemented on 2010 I-MIEV, and since it's an output, it not required to get charger enabled for bench testing.
Pin 5 on the J1772 is the proxiity pin that confirms the charger cable is plugged in. I think this pin connects to ground when a charger cable is attached. If this is true, then this would connect to pin 9 of connector E-01 which in turn connects to pin 5 of the CN101 Connector on the TOP PCB. It looks like this pin is not connected on the early 2010 model, but is connected in the late 2010 model.
I think this would mean that Pin 5 of CN101 would need to be connected to ground before the charger would run.
Kiev / Coulomb: do you have any info on this to confirm this?

skylogger
Posts: 74
Joined: Sun Dec 14, 2014 3:54 am

Re: Troubleshooting and repair for On-board Charger, DC-DC Converter

Wed Aug 08, 2018 2:57 am

I finally got things setup to look at the pins of TR310 on the bottom board with the scope:


Image

Image

So this looks like TR310 is chopping ok, now I need to pull it apart and find out what test points to test for the other supplies.

coulomb
Posts: 114
Joined: Sun Jun 10, 2018 8:32 pm
Location: Brisbane, Australia

Re: Troubleshooting and repair for On-board Charger, DC-DC Converter

Wed Aug 08, 2018 5:55 am

skylogger wrote: C707 16.75V
C704 16.75V

Huh. So something is boosting the 13 V input, or else there are some spiky signals confusing the multimeter. But across large capacitors, this should not be the case. It seems such a non-standard value. And so unnecessary given that there is another power supply chopping it up on the next board. It's possible that it's meant to be some other value, say 20.0 V, and some fault is loading it down, but for now I think we assume it's meant to be "about 17 V" until proven otherwise.

... checked to see if there is any AC at the primaries of the transformer that are on the ouptus of the IGBT's. This is still dead.

This is to be expected, considering we still have about three other balls to juggle.

We probably need to work out what signal needs to be on the CHGP Line from the EV-ECU to pin 3 of CN101. This is probably the signal that actually enables/disables the running of the IGBT's.

Yes, this is the first one. I note that the proximity signal from the J1772 connector (that Mitsubishi are calling pin 4, but the rest of the world calls pin 5) goes straight to the EV-ECU. So when the user presses the button (meaning stop charging, I believe, but I'm no J1772 expert as yet), this must go via the EV-ECU and the only ways to stop the charge are via the CHGP signal, or by cutting the power to the charger relay. The latter seems way too crude to me; you want to gracefully shut down the power, after immediately cutting off gate drive to the main IGBTs to stop charging. Otherwise, you could have IGBTs with partial gate drive going into linear regions and overheating.

I'm also having problems working out the The PE and CP Proximity and Pilot pins from the j1772 connector to the charger for bench testing.
I think that PIN 4 of the J1772 is an output to EVSE to tell it what charge status is on car based on change in resistor values. This was not fully implemented on 2010 I-MIEV, and since it's an output, it not required to get charger enabled for bench testing.

My understanding is that proximity is a signal to the vehicle, so it can tell one of three states: no plug connected, connected but button not pressed, and connected with button pressed. You might measure the voltage to chassis at the proximity pin; not connected should be 4.5 V due to a 330 Ω pullup to 5.0 V and the 2.74 kΩ pulldown near the J1772 socket. If so, try putting a 150 Ω resistor to chassis so that the pin falls to 1.5 V. You might consider putting a switch in series with that resistor, perhaps a normally closed one, with a 330 Ω resistor across that switch. Then if something goes wrong, you press that switch in a hurry, and it's like pressing the J1772 plug button, which I believe should kill the charging. Maybe even make it a red mushroom type (an E-stop for charging).

Pin 5 on the J1772 is the proxiity pin that confirms the charger cable is plugged in. I think this pin connects to ground when a charger cable is attached.

Again, my understanding is that when the cable is attached, that pin 5 (according to the rest of the world, or pin 4 on the Mitsubishi diagram) needs 150 Ω to ground to indicate cable presence. A short circuit to ground would be 0 V, an error, and would likely prevent charging.

Pin 5 on the Mitsubishi diagram is the large earth pin at the bottom, which the rest of the world calls pin 3. Some diagrams show it as neutral in split phase situations, but fortunately you don't have that. All the diagrams I see show the vehicle connecting it to chassis / earth / ground. This seems to be the reference for the pilot and proximity signals.

I note that the pilot signal is generated by the EVSE, and "read" by the charger, although the charger modifies its amplitude with a diode and two different valued resistors. It might be necessary to generate a ±12 V 1 kHz square wave (24 V p-p) with a 1K resistor in series to keep the EV-ECU happy. It might be necessary to start with DC +12 V to initiate the handshaking (state A, EVSE ready, not connected).

If this is true, then this would connect to pin 9 of connector E-01 which in turn connects to pin 5 of the CN101 Connector on the TOP PCB. It looks like this pin is not connected on the early 2010 model, but is connected in the late 2010 model.

I'm confused. Pin 9 of E-01 connects to pin 3 of the J1772 socket per the Mitsubishi diagram. This seems to be the pilot signal, that the rest of the world calls pin 4. It makes sense if these are not connected in early 2010 models, since they don't do real J1772 charging, and merely connect the mains (large pins 1, 2) to a mains plug (forgive me if I got this wrong). I thought all Australian 2010 models are like this, unless or until they have the "Gelco" modification, which makes them more or less proper J1772 compliant, so you can charge at public charging stations. @Skylogger, has your 2010 had the "Gelco" modification done? Does yours connect to the pilot signal (rest of the world pin 4, Mitsubishi pin 3)?

I note that the confusion re pin numbering is merely the number that's the problem; the function (e.g. line / pilot / proximity / protective earth) is the same, as it needs to be to work with public charging stations.

I think this would mean that Pin 5 of CN101 would need to be connected to ground before the charger would run.
Kiev / Coulomb: do you have any info on this to confirm this?

As above, I disagree, but am happy to be proved wrong.

[ Edit: "no button" -> "button not pressed" ]

kiev
Posts: 815
Joined: Sun May 03, 2015 7:15 am
Location: The Heart o' Dixie
Contact: Website

Endeavor to Persevere

Wed Aug 08, 2018 4:33 pm

That is some good data on the cap voltages, and a great capture with the scope. That proves out quite a bit of stuff that is working, and every little bit helps solve the puzzle.

As Coulomb said there are several balls to juggle here before even being close to getting a charger to run on the bench. Here is my list:

1. CAN buss commands from the EV-ECU. The voltage and current commands are calculated and determined by the EV-ECU with data from the BMS and Cell Monitoring boards. There is a big ole processor chip on the underside of the top board that receives these commands and implements the necessary detailed control commands for the PFC and FETs, etc. i'm sure it won't work without this.

2. The J1772 protocol implementation (or not), probably a minor issue.

3. Troubleshooting the box to determine what has failed and why, and what else might also be damaged besides the obvious burned resistors or blown snubber caps. Why did they fail, what caused the chain of events to create the damage, etc. This is important to prevent repeating or causing futher damage.

4. The top board creates multiple power supply voltages, the bottom layer has the processor and a multitude of CMOS logic chips, plus any sensor and signal conditioning needed to provide data to the microprocessor. The bottom board is simple in comparison. We have to know what is normal operation to compare for troubleshooting.

5. Diagnostic Trouble Codes in the EV-ECU. Some codes will prevent the OBC from operating, so the codes need to be read and cleared if possible.

i did the easy part (bottom board) first because i've been bullriding in this rodeo before. It does take a lot of time to trace it all out, but i don't know of any other way to get the knowledge . You've got the best team working on this including the master boomerang thrower of all time in QLD--coulomb has special skills and magic with processor code.
Last edited by kiev on Wed Aug 08, 2018 6:13 pm, edited 1 time in total.
kiev = kenny's innovative electric vehicle

kiev
Posts: 815
Joined: Sun May 03, 2015 7:15 am
Location: The Heart o' Dixie
Contact: Website

Small bite of the elephant

Wed Aug 08, 2018 4:55 pm

i found a small diode rectifier bridge circuit on the bottom board coming out of the side of the AC input doghouse. It runs thru a series string of resistors to drop the voltage way down, then gets passed to an optocoupler and sent up to the top board. i expect that it goes to the microcontroller to provide a signal that AC is present in the doghouse, and that the relay, emi filter coils, and resistors are good. i traced the top board portion and rang it all out except for the final link to the processor pin--i'm sure it must be there but the multilayer board and tiny parts and traces are a real pain...

The 12V supply that is hot all the time Pin 7 of CN101, creates a 5V supply on the top board (and possibly more) that is used in this AC sense circuit on the bottom board. If possible try to repeat your test and measure the current draw with just that one 12V to the board with the CN1 cable attached to the bottom board. And measure the cap voltages. i suspect that only a portion will be energized and it would be good to know what the "OFF" state power draw of the OBC might be.

Thanks to your capacitor voltages i now know where to look on the top board to flush out the low voltage power supplies--that will save quite a bit of time.

i have hand written notes for this and will convert to schematics as time permits.

p.s. i checked the wiring colors in Jay's box and it matches your late 2010 colors. One of the brown wires with the CAN line is dark brown, the other is lighter such as tan.
Last edited by kiev on Wed Aug 08, 2018 6:16 pm, edited 1 time in total.
kiev = kenny's innovative electric vehicle

Return to “Batteries and Battery Management”