Increase charging rate.

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.
A

Anonymous

Guest
Hi.
CaniOn has shown our ev charges barely 2,4kW when our ev is charging from a charging point capable of 16A 230V AC phase-neutral (Europe) which would be 3,7kW minus inverter consumption. This rate goes for all times when charging below 95% SoC or so.

2YNQTln.png


I wonder if this limit is decided by the ECU or onboard charger.
From sharings on this forum I have read the ECU charge limit to be 4kW. On a table in the section of technical articles it shows the ECU accepts up to 10A which would equal 3,6kW on the HV wires. The total draw from the AC supply would be 3,6kW plus inverter consumption.

So I wonder how to adjust so the charger provides full 10A output. Anybody who knows how this is done?

Reading an article found in the folder of technical articles about how certified technicians carries out update of the ECU using MUT3 equipment and Medic 3 computer software I'm wondering whether such equipment is accessible and whether its realistic using this to adjust the max current rate of charge. Any knowledge and experience on this anyone?

I've seeing the fuse protecting the HV wiring from the onboard charger is rate at 20A. This one is located in the MCU on the HV wires from the onboard charger.

6PEumeQ.jpg


This would indicate the physical wiring being capable of handling a maximum of 20A equal to 7,2kW output at 360V.

I've got a second charger from a c-zero which I'm considering to parallel. However this would only be effective if finding a way to either
- limit either charger to max 5A each. A total of 10A output as is the current limit programmed on the ECU.
- reprogram the ecu to a higher rate.
- adjust/reprogram the onboard charger to provide more power.
- remove the current signal to the ECU (if even possible to still initiate charge)/ Adjust the signal to a value acceptable for the ECU while actually charging more within the physical limits of fuses.

So anybody with experience in dealing with reprogramming the ecu, adjusting the current signal and paralleling the onboard charger, please share.

Also does anybody know in which physical location the current measurement is done?
I know of the current sensor onboard the battery pack and wonder if the onboard charger has one as well.

Thanks.

Photos of the onboard charger:
0abHH70.jpg


ESlImea.jpg


CSQzJTf.jpg


rBr0OSi.jpg

Anybody knows what this does?
And explain what the orange wire passing outside does?

suaWXtW.jpg


I'm noticing the AC onboard charger is placed on top of the DC/DC charger. I wonder if wanting to remove the AC charger to use in another car is this able to separate at all from the DC/DC charger?
 
The set of coils in the little metal doghouse above the charger is an ElectroMagnetic Interference (EMI) filter.

i think the orange wire on the outside is the ground wire--no need to filter the ground.
 
If you are able to get two chargers working together, the easiest way to limit input current is to lower the EVSE pilot signal. That should theoretically reduce the output of both chargers. Setting the EVSE for 9 amps at 230 volts should give you a total of 3.9 kW. With 12 volt loads, though, you may be able to set it for 10 amp input (pay attention to EVSE wiring limitations, two chargers at 10 amps will draw 20 amps through the EVSE).

The charge rate in CaniOn appears lower because of the 12 volt loads drawing power before it reaches the battery (where the current sensors are located). Plus, such low amperage isn't perfectly accurate. CaniOn will show actual AC input voltage and amperage through the My Trip Timer screen.

What others have done is tie an external charger directly to the HV bus to allow fast charging at around 12 kW. I'm looking to mimic the setup of a local FFE (Ford Focus Electric) driver in creating a CHAdeMO charger from a pair of 6.6 kW FFE on-board chargers to allow for up to 13.2 kW charging from a pair of charging stations. It would simply connect to the car through the CHAdeMO port. As far as I know, though, successfully increasing default level 2 charging rate has yet to be accomplished.
 
Here are some photos of the onboard charger taken apart.

ow9mrgJ.jpg


UfhaIA2.jpg


Top lid off and top board. AC input at bottom right. Two orange wires.
r25cSdJ.jpg


Xoejzal.jpg


Bottom board. DC output top right. Orange wires.
Q9AFpcI.jpg


In curiosity to remove the boards all screw and plugs was detatched. Remember to mark the plugs before removal. It helps to figure out where they need to be connected.
As the bottom board stayed fixed I decided to take a look at the lower part of the box. So the unit was turned 180degrees and lid removed.
This will be the DC DC charger.
H9yOVhl.jpg


After removing and disconnecting lots of plugs, screws and bolts I noticed a alumium plate and took this off and found the cooling area.

zO1gesK.jpg


Interesting to look at. Yet no bolts or screws fixing the AC charger bottom board. So its probably fixed applied with paste of some kind.

I got the unit reassembled and left the DC DC charger disconnected in trying to use only the AC DC charger.

Today I had it set up and connected to check if it actually functions as a replacement to the fully functioning onboard charger.

h2g1eYB.jpg


Results shown on CaniOn when only the extra charger was connected instead of the onboard unit.
9e8bQb5.png


Seeing the AC charger actually charging was delightful. I had reached todays goal of testing whether the charger would work or not. Both due to being a part coming from a different brand (Citroen c-zero) and as its know how part are coded to fit a specif car. Not to mention to see if the unit would actually funtion especially after the disassembly and reassembly. So I stopped the charging, disconnected the wires and reconnected the onbaord charger.
With the extra charger disconnected and only the onboard charger reconnected I engaged charging once again and took a look at the Canion app to check if charging was working once again when reconnected with the onboard unit.

VB2njbw.png
 
Very cool. Can't wait to hear if they both run together. If they do, that should speed charging by better than half if the limit is 4 kW.
 
From my experience with paralleling a 12 KW charger to the battery pack, you might be able to add the additional input into the pack without a problem. In my case, if the car was connected to the EVSE and charging normally I could apply additional input to the pack up to a point. I don't remember exactly at what point it was but I believe it was about 6 KW of total DC input (internal charger plus external charger) and the car disconnected the internal charger and shut down. There must be some software limit that senses an overcharge in the BMS or the EV-ECM. Two paralleled internal chargers at full power are less than 6 KW DC to the pack so you have a little latitude there. The other issue is the final state of charge when the internal charger backs down to a very low output and the battery balancers turn on. Balancers are just small resistors that are introduced into the circuit to shunt individual cells to balance the pack and can be destroyed by just the partial output of the internal charger. If you could get CAN buss control over both chargers its possible you could make this work automatically--if not the extra charger must be shut down prior to balancing. Good luck on this and I think a 6 to 10 KW charger would be perfect in a car with a pack the size of the I-MiEV. BTW--the charger is also the DC-DC converter for the car so you would need to parallel all of the connections of both chargers and fuse them separately at the inverter to pass the power bi-directionally from the battery pack without blowing the 20 amp high voltage fuse.
 
I think he's using only the charger board out of the second unit and not even hooking up the DC-DC converter.

I was thinking more about how this would function. At the least, it would need an Arduino to control the second charger so that it won't trip breakers by pulling pilot x2. It would need to intercept the CAN bus and have a tap on the pilot signal, and do some math to do pilot - AC amps = set charge rate (say a public station has a pilot signal telling the car 30 amps are available. The Arduino would see this, subtract the AC amps value on the CANbus, and set the 2nd charger to however many amps are left over, up to the max power input). This all depends on how automatic the OP would like this setup to be, and for full automation, would be much more robust than this.

I've heard the cutoff being anywhere from 4-6 kW.
 
So the second charger is hooked: AC input, DC output, Ground, communications via split on printboard.
When all 13 wires in the communication plug are connected to both chargers the evse contactor will not engage.
When the two CAN bus wires (n. 6 and 13 - Plug E-03) are disconnected from charger n.2 the evse contactor engages however canion shows no charging. When also n.4 (going to DCSW on ECU) are disconnected from the second charger then the evse contactor does engage and canion proves charging is ongoing.

Can someone clarify the use of n.4 connected to DCSW on ECU?
I suspect this to be part of the DC-DC charger setup, and in this set up without use as we use the AC/DC charger only for this.

Second and most important. How to ensure the 2nd charger gets the CAN bus signal. I wonder if this takes to block the CAN signals from the second charger so it doesnt reporting back on to the CAN bus?
Or could it be its only the DC-DC part which does so. In such case disconnecting the communication plug to the DC-DC charger would solve this.
Someone suggest to use two of these MCP2551-I/P and put them on the CAN bus wires before the 2nd charger. The person suggest to use one microchip setup for reading and another to pass on. I'm unfamiliar with this and input will be appreciate as to how this could work.

Another suggestion received is to use a CAN bus repeater which can be set up to cancel input on to the CAN bus. In this case from the 2nd charger. This solution requires a device costing around 50 times that of the microchips from a brief show online.

Again input is welcome also as I'm trying to get the stuff ready for an ev event in 2 days.
The event is an official record attempt of number of ev's driving in a parade. I think we have some 300+ ev's signed up so far. More info https://fdel.nemtilmeld.dk/4
 
Why not set up the chargers in a master-slave configuration such that only the master allows the J1772 signal, so when the master shuts down the EVSE, the slave gets starved. A blocking diode on the slave's J1772 wires may do it.
That shouldn't result in full-current shutdowns, but a big magnet alongside the EVSE relay would help prevent damage from arcing. (DIY EV converters have been using 12V or AC-only rated relays on high voltage for years with this blowout magnet technique for bending the arc to that it is effectively longer breaks more quickly to prevent erosion of the contacts). I'd expect that EVSE already have blowout magnets, but haven't confirmed. I think we can cause full current shutdowns by pressing the handle release button, which immediately opens the EVSE contactor before the charger ramps down.
 
Thanks for input jray3. I wonder how your suggestion solves the CAN bus situation?

I assume both chargers require CAN signal to charge. Yet only one can be allowed to send a return message on to the CAN bus. Any input on this assumption any one?
 
(Taking a stab here) The charger continually broadcasts info on the CANbus (AC Voltage, AC Amps, etc.) using designated PIDs. The EV-ECU uses different PIDs to control the charger. But, when there are two chargers both trying to use the same PIDs, it causes errors on the CANbus. Now, the car is supposed to fallback to a different bus (K-line?) should there be too many errors on the CANbus, but again, if both chargers are feeding info to the car, there will still be errors and the car will shut down.

You may need to have a microcontroller in between the CANbus and both chargers. That way it can be programmed to make two chargers appear as one and properly make use of the EVSE pilot signal.

Jay, I've noticed a slight delay between pressing the button on the J1772 handle and the EVSE contactor dropping out. Maybe the charger does cut out before dropping the contactor?
 
Thanks for sharing the pictures of the boards.

i see at least 2 fuses inside the dcdc. Fuses are good to protect wiring from overcurrent, but if they fuse and open then the box is dead unless someone can troubleshoot it and determine the problem. Then who even knows about the fuses that could change them? Seems a shame to throw out otherwise good electronics just for want of a fuse.

You didn't happen to trace any schematics while you had it open did ya?
 
Thx PV1 and Dani. I can almost think analog, but can't comprehend CAN. Didn't recall the delayed EVSE disconnect either. Without a cookbook, my approach to faster charging would be either a second non-CAN charger along the lines of what siai47 did through the CHAdeMO contactors, or a primitive forced regen approach (pusher trailer or powered rollers!).

My hat's off to you digital dEVelopers, hopefully you'll blaze a trail to maximize this minicar without burning anything down....
 
Hi Dani,

Thanks for posting the pictures of the charger internals. It's really interesting what you are trying to do.

I'm not a specialist in electronics but here are my thoughts on getting the second charger working.

I think if you want to make progress on this you will need to understand how the charger is controlled by the can messages.
Typically you get some hardware that can datalog the can bus and then play back and edit the log file.

You take a CAN datalog of the car charging normally and then put the charger on a bench and playback the CAN datalog into the
charger standalone and see if it turns on. When you can get this working then you start removing data from the playback until
it wont activate to isolate which can messages control the charger. I think this is kind of tedious and is mostly a computer job.


Hopefully you can figure out how to activate and control the charge rate of the charger on a bench and then think about ways of integrating
the second charger back into the car.

Alternatively I also think that the charger could be controlled bypassing the CAN bus. I suspect the charger is made in two sections a high voltage
section and a control section. I would guess that there is an analog control signal somewhere that controls
charge rate if you can bypass the can controller of the charger and use the analog control signal then you can do this without any programming or
CAN hardware in the same way siai47 did it.

This would require no computer skills but you you need to figure out how analog side of the charger works. Again another bench test exercise.

Good luck with your efforts I will definitely follow your progress.

Don...
 
Thanks for all input so far.

Status update:
Communication wires needed for the second charger are identified. An additional plug is attached taken from the almost identical wrecked car I've got at disposal.
AC power has been wired to the second charger. Attempt of drawing in another live-phase to balance load on the power grid was done, however this resulted in the charger wouldnt engage in charging during solo test. Seamingly there is some sort of measuring going on inside the charging unit from the evse or AC inlet. Any clue what this is and takes to introduce a second phase. Same neutral is used.
DC wire is connected in to the MCU and a new lid for the trail has been produced.

When charger 1 (stock) and charger 2 (additional) are parallelled on communications, AC input (same phase) and DC output connected each in the MCU. Then Canion displayes charging engages on both chargers for some 5-10 seconds and then charging is stoppes by ecu and evse told to disconnect, which I can tell by the evse contactor dropping out.
In the 5-10 seconds Canion shows how charging rate is indeed doubled.
If the two CAN bus wires are disconnected from the second charger, only charger 1 will charge and will charge continueously as regular charging. From this I get clear how the CAN bus is the "only" piece needing adjustment to have both charger charge continously.

This was tested at different pwm charging rates and showing same charging-stopped result when CAN bus signal wires are connected identically to both chargers.

During the ramp-up charging on both chargers Canion reached 3.2kW before charging was stopped by the ecu. Regular single max charging rate is around 2.3-2.5kW.

I've been suggested to try filtering the CAN signal for the second charger by using two mcp2551 i/p can.
Word is this setup could need a resistance connected to it which matches the size of resistance used in this particular CAN bus system at the terminating points. Anybody knows which size resistor will match this ev CAN bus?
I've understood the point of using 2 such microchips to make sure only the CAN bus signal is let through from the ecu to the second charget and blocked in reverse.

So far it was great to see the double charge rate which showed on first attempt. The pwm signal was set for 1.0kW (DC level) of charge which I registered on Canion when at first using only the stock/onboard charger. Then as also connecting the additional charger the displayed power was 2kW at the same pwm signal. This was great to see.
 
According to the service manual there is 120 Ohms between the Can H and L at each end. So measuring the resistance between Can H and CanL would read 60 Ohms (2 in parallel). During diagnostics of the CAN, if the measurement reads 120, then one end is open and the lines will be noisy.

Good luck with your investigation.
 
A CAN bridge is on its way and the manufactor explains this device is capable of removing the return CAN bus message from the second charger.
What is needed is to find the address of the charger whichs needs to be filtered. Then to put this address on the CAN bridge. Anybody who has found out the address from which the charger communicates back to the ECU?
 
Back
Top