Cell Monitoring Unit "recall"

Called up Capitol Mitsubishi in San Jose today to set up an appointment. They bent over backwards to be accommodating. Will post if anything surprising happens.

JoeS said:

Called up Capitol Mitsubishi in San Jose today to set up an appointment. They bent over backwards to be accommodating. Will post if anything surprising happens.

Click to expand...

I'm sure they got some grief from Mitsubishi headquarters after I complained about their service department's lack of interest in answering the phone, E-mail, etc. Plus since they have the replacement SRS sensor parts sitting there they had ordered for my car the one time I actually got them on the phone. In any case I hope it all goes well for you.

Went in to Capitol Mitsubishi in San Jose today to have both the airbag recall and CMU software upgrades done. I was on time for my 11:30 appointment and they took me right away and were done by 1pm. They were very friendly and accommodating.

Anyway, two minor and one possibly major outcome of this process:

Minor
1. Radio presets wiped out
2. Both Trip A and Trip B had been reset and had 2.4 miles on them.

The Trip A reset bothered me as that's what I use for accumulating mileage in-between full charging, but all I lose is decimal-point accuracy for this interval. Like I said, minor inconveniences.

Significant
After arriving home I immediately plugged in the iMiEV into my L2 SPX EVSE as we were going to take a trip later on in the day. I checked on it a couple of hours later and it had STOPPED CHARGING. Unplugged it and plugged it back in and it started charging normally. I checked my T.E.D. (power monitor) on the computer and it showed that the charger had turned off about 1/2-hour after it had started charging and stayed off. I had not touched the Remote. Normally, the charger turns off and takes a ten-minute break sometime during the charging cycle.

Called up the service manager at Mitsubishi to let him know what had just happened, but that I was not going to worry about it unless the problem recurred. He tried to reassure me that it's all covered by warranty and tried to be helpful by suggesting that perhaps I had let the 12v battery run down by perhaps leaving a dome light on... which he said would cause the charging to stop as that was the problem they found on another iMiEV. My 12v is just fine.

Anyway, my speculation is that the reprogramming perhaps disturbed the Remote timer settings and that's what caused the charger to stop. I'm hopeful that this was simply a one-time occurrence. Although I can only recall one instance in the past of an unexpected charging stoppage (very soon after I purchased the car), I don't remember if it was using L1 or the SPX L2.

Anyway, after plugging her back in this afternoon she properly charged up to about 10 bars using the L2 SPX EVSE and we went for our visit and returned home tonight and right now she's happily charging away using the Mitsu L1 - I'll let her go to 100% overnight (it's cool and I want to reset my TripA odo) and if any further associated issues occur I'll post them here.

Like I posted earlier, I'm actually quite happy that Mitsubishi is staying current and implementing whatever they've learned into the battery management software. This little hiccup I consider trivial.

Update, a day later, following the Cell Monitoring Unit upgrade.

The overnight (11 hours) Level 1 charging to 100% worked normally, with the usual ten-minute timeout by the charger occurring this time about 3-1/2 hours into the charge cycle.

This evening, starting with 9 bars, I tried using the Remote with my SPX L2 EVSE. First I set the turn-on time to "0" (zero) - it accepted this, no problem. Then I set the on-to-off time to be two hours, and the car refused to accept this, giving me an ERROR on the Remote. I tried this five times, with differing time intervals before pushing the top button, and each time I got an error and I also noticed the warning on the dashboard. This is a change from before, as in the past it would either accept immediately or after one or two retries. Basically, my take is that my Remote is now unusable with the SPX L2 EVSE for charging.

Not to worry, as because of it's previous flakiness I had installed a mechanical timer on my input circuit (along with a backup kWhr meter) - so I set the mechanical timer for two hours and connected the SPX L2 EVSE and the car charged happily for two hours this evening. It did its 'timeout' routine about 35 minutes into the charge, but this timeout only lasted six minutes.

Inasmuch as I had already previously given up using the Remote with my SPX L2 EVSE and have been using my mechanical timer, this incompatibility doesn't bother me. If I had one of the Mitsubishi-recommended L2 EVSEs and it did this, methinks I'd be unhappy.

I tried using the Remote to turn on the heater in the car using the SPX, and it took two tries (first one gave an error) but it did activate the heater on the second attempt. This is nice to know as the heater (on paper) takes away a bit of energy from the battery even when plugged into L2.

I don't understand how the car knows if the EVSE connected to it is 'acceptable' or not - Someone smarter than me needs to explain this

When you study the 2 way 'communication' between the car and the EVSE I don't see anything which would tell the car what brand of EVSE is connected? You plug in the EVSE, it signals the car what charge current it available, the cars says OK and sends a signal that it's OK for the charging to commence. The EVSE closes it's relay(s) and the car begins charging. I understand that this 'communication' is via signal levels and resistances, but what else is there to tell the car that this particular EVSE is one of the accepted ones?

Could it be that when any L1 charger is connected, if it signals that any amount of current OTHER than 8 amps, THAT lets the car know that it's not the stock EVSE? (The stock EVSE apparently works with the remote, while the upgraded one does not) Likewise, maybe the approved L2 EVSE's all signal exactly 16 amps (or some other preagreed to number) so then any other level would let the car know that the connected EVSE isn't one of the accepted ones?

I just don't understand how any J1772 spec EVSE is accepted by the car for charging purposes, but only certain ones work with the remote? Did they really do this on purpose to steer the customer into buying an EVSE only from vendors with whom they have financial agreements? I understand that they do have to allow any J1772 EVSE to charge the car, otherwise it wouldn't be able to opportunity charge and THAT would have owners complaining by the hundreds, but are they trying to prevent us from using the remote functions if we don't buy 'their' charger?

An ICE must function normally (and under full warranty) with any brand of gasoline, so long as it meets the specified octane requirements - The car maker cannot tell you where to buy gas. Likewise, any EVSE which meets the standard (J1772) SHOULD be acceptable to the car and it should function normally, regardless of who made the EVSE

I think this is a legitimate complaint for Mitsubushi - If everyone griped about this, the next reprogramming of the car would eliminate all these problems I think

Don

My suspicion is that its the EVSE's fault and not Mitsubishi. Its all really a dumb system, right? Plug it in, the car sends a pilot signal and the EVSE reads it and throws the relay - simple. I can't imagine the car charger doing anything to the EVSE that would cause these problems. I'm fortunate and don't have any of these issues with my setup (Schneider Electric L2) and I use the remote a lot. It could be the EVSE is being too sensitive when the user is making a request its possible the car is sending a signal to the EVSE to hault, I usually hear a thunk when I just turn on the remote to check my SOC. Its possible the EVSE is interpretting this as a fault instead, so when the user presses the top button to execute the command the EVSE is in fault mode or in a cycling mode, the car then senses its disconnected from the EVSE and thus refuses to accept the command from the remote giving the dashboard warning. Its just my theory because some of us have problems and some of us don't but the only differing factor is the EVSE. In the cycling scenario, when the EVSE completes its reboot, the car senses its now connected and accepts the command. We all have basically the same car and the same remotes. From what I read about EVSE's is that yes there is a standard but not all manufacturers follow the standard, its more of a guideline.

I can not agree more with Don. My EVSEUPGRADE works well without remote at both level 1 and 2. Level 1 also works well with remote, but level 2 would casue a problem. I thought my bringing this issue to Mitsubishi's attention was doing a favor for them. However, their response was that I either returned the charging cable or I was on my own. I really hope there would be a recall to fix this problem.

MLucas said:

My suspicion is that its the EVSE's fault and not Mitsubishi. Its all really a dumb system, right? Plug it in, the car sends a pilot signal and the EVSE reads it and throws the relay - simple.

Click to expand...

Not exactly. I would describe the EVSE as a 'smart cord' and not a 'dumb system.'

First, the car does not generate the pilot signal as you stated . . . . the EVSE sends the pilot signal to the car. The J1772 Pilot is a 1khz +12V to -12V square wave, the voltage defines the state and the duty cycle defines the current available to the EV. The EVSE sets the duty cycle and the EV adds resistance between the pilot and ground to vary the voltage. The EVSE reads the voltage and changes state accordingly. It is a two-way communication between the car and the EVSE which is supplying the charge current. Doesn't sound dumb to me - I built one myself and I think I'd recognise 'dumb'

What I don't understand is how the car knows if it's connected to a 'recommended' EVSE as opposed to the one I built myself, which is actually quite a bit 'smarter' than the one supplied with the car

derminghsieh said:

I thought my bringing this issue to Mitsubishi's attention was doing a favor for them. However, their response was that I either returned the charging cable or I was on my own

Click to expand...

I wonder how they think you're going to charge the car anywhere else but in your own garage? Surely they don't think all opportunity charging stations are using Mitsubishi 'approved' equipment? Public stations are intended for ANY J1772 spec car to use, including ours so the car shouldn't have a problem with any of them, so long as they all conform to J1772

Don

J1772 is a Standard. I seriously doubt that the Mistsubihsi recommended (not approved. Its only recommended) is deviating from the J1772 standard. I never have any issues with EVSEupgrade, Leviton EVSE, or the OpenEVSE I built. Why? Because I tossed that remote control into the useless gadget drawer as soon as I saw all it can do is say error, error, error.

Got a new car. Cant close doors any longer (without opening windows first

Yesterday it's been our 2010 i-MiEV (3 month old), who got the update.

We could not burn more than two bars yesterday so we had to wait for another ride today to empty half the battery and to recharge. Well, we have been on a pirates convention spending some 350 kilometers and seeing the turtle thrice, not so long ago. Today we had a hard time getting that battery half empty. Whenever I had the electric grin on my face the turtle was smiling back at me showing deep blue regen at the next traffic light. No sound from our i-MiEV but a lot of noise from the other cars who dropped far behind. Next we tried the hills. No way, after every hill there comes a valley and the faster we go up the deeper blue the regen when we go down. The car feels like a new car and I feel 10 years younger.

Some 62 kilometers took us an afternoon. We are living on a country side, traffic jam, hills and narrow curves and we had fog today so we had to use AC and heater. 10ºC = 50ºF and we have got our winter shoes yesterday. Charging took us longer than I expected. I circled her with my hearing aids switched to T-coil to listen for anything unexpected. The fan was working but everything sounded like always. Except the girl was really thirsty today and she deserved every kilowatt.

She got us 133 kilometers. That is more than I expected after what I have done to her today and it is more than she used to give us these days.

J1772 is a standard and we are still on the Panasonic 230V/16A brick that Leaf and i-MiEV used to deliver but today we have ordered a 230V/6A EVSE that Opel is selling for the Ampera here in europe. Some camping grounds only offer 6A and with their thinner cords we might be able to charge with closed doors in winter nights.

One of the features of a good Battery Management System is that during the charging process it temporarily suspends charging and assesses the health of individual cells (or, perhaps, cell groups) by looking at their 'sag' with no voltage being applied. I was happy to see this 'timeout' feature in our iMiEV charger, which is readily visible on my T.E.D. (The Energy Detective) graph.

Prior to our cell 'update', this timeout was almost exactly ten minutes. Now, after the 'update', I see that it has been reduced to six minutes. I guess the battery engineers must have figured out that if there's a cell issue then they don't need to wait the full ten minutes to identify it. The benefit to us is that we now get another four minutes' worth of charging, which can translate into 0.8 miles. Hey, every little bit helps.

Ignoring the SPX L2 EVSE Remote incompatibility, this is the only difference I've noted since the Cell Monitoring Unit software update.

The EVSE Pilot signal should be + / -12v, but some Evse's only
output +12v to zero volts, "stretching" the standards.
However, most cars will accept that variation.

The car changes its resistive load on the Pilot, which should have
a 1000 ohm output impedance (resistance) from the EVSE.

Slightly simplified (excluding the diode in the car):

Loading the +12v of the Pilot down with a 1000 resistor would
result in a 6v Pilot instead of a 12v Pilot. The 6v level tells the EVSE
to close its Relay to provide AC power to the car. Only then, can the
car determine what the input voltage will be.

Changing the resistance to 3000 ohms should load the Pilot to
9v, signaling the EVSE that there is a car attached, but that
AC power is not yet wanted. These resistances, provided by
the car, must be lower when the effects of the diode are considered.

The incompatibility usually comes in several forms:
1. The timing of the response by the EVSE to a power
On or Off request from the iMiEV might be
too slow for the car's expectations.

2. The EVSE might expect only one Power On request per
attachment to a car, and think anything else is some error.

3. The timing of some interaction between the EV and the
EVSE is slightly too slow (or too fast) for either the
EV or the EVSE to consider "normal".

4. Less frequently, the diode in the car might be "blown",
either open or shorted.

Update after 300 km.

The final bar takes a lot longer charging than the other 15 do.

Kilometers remaining is less forgiving than it used to be - or is it showing the results I should expect from the early Winter? After a full charge it remembers A/C and heater used.

It is charging faster, the first 15 bars at least.

I am still using the original Panasonic EVSE 230V/16A but I have ordered an EVSE for the Opel Ampera (Chevy Volt) that should do 230V/6A as a backup and for power sockets that wont allow more than 6 A. I'll update as soon as get it.

I had my CMU software upgrade today. I saw no difference driving home. I'll let y'all know if there seems to be any difference in charging or RR estimates over the next week or so.

I can't say enough good things about my dealer, Bell Mitsubishi of NJ. They got me in and out in less than an hour...and they washed my car inside and out. When I bought my i, I had the choice of buying from a very closeby dealer (less than a mile) that sells at least 12 makes of cars. Instead I chose Bell (15 miles away) because they only sell Mitsubishi and they actually have staff knowledgeable about the car.

I'd like to listen what all those 88 battery cells are talking - eavesdropping into the canbus. Could give each of them a name and say hello to them. :mrgreen:

Ordered an Arduino today for learning, Machina will be the way to go for canbus hacking. Elm is far to slow.

Well, so far, owning a Miev is still better than a 787 !