Turtle with no warning and 25km range remaining

[Trevor Jack]

My 2010 iMiev has turtled twice in the last week. Each time it showed 25km range remaining and battery at about 1/3, the latter being consistent with a typical full charge reported range of between 70 and 85km. Each time the reported 25km range was consistent with the charging and driving I’d done to that point. After disconnecting the 12v battery and reconnecting the first time, the dash showed zero battery, zero range and turtle. The second time (yesterday), it showed the same as it had before disconnecting – ie 25km range, 1/3 battery and no turtle (perhaps I didn’t leave it disconnected as long/long enough…).

The vehicle had shown an HV error (car!) about a week earlier. That was cleared and has not been seen since. I didn’t clear the codes but made these notes. P1A2F, P1A4B, U1111. This was before the first turtle event. My buddy read the following only about 5km before the turtle yesterday (ie before the 2nd turtle event):
- P1B01, stored DTC, Quick Charger vehicle error;
- P1AF6, stored DTC, On board charging time over;
- U1111, Active DTC, Display CAN timeout, Not equipped.
The U1111 could not be cleared. The two stored DTCs were cleared.

I had used a Chademo charger, for the first time in the 10 years we have had the vehicle, on the trip that gave rise to the first turtle event. That charge was after having done about 25km (having started from a full charge) and only added 0.6kWh before stopping automatically. (Although I don’t see how these two things could be related, I observe that the reported expected range at turtling, twice, (25km) is the same as the capacity that had been used at the time of the Chademo charging (also 25km).)

I put a new 12v battery in yesterday before the second turtling event. The old battery showed good voltage and passed cranking tests.

After each turtling event I have charged, with the granny charger at home, and the vehicle has shown full charge and about 80km range.

Any suggestions?

 

[MickeyS70]

Hello and welcome to the forum

My 2010 iMiev has turtled twice in the last week. Each time it showed 25km range remaining and battery at about 1/3, the latter being consistent with a typical full charge reported range of between 70 and 85km. Each time the reported 25km range was consistent with the charging and driving I’d done to that point. After disconnecting the 12v battery and reconnecting the first time, the dash showed zero battery, zero range and turtle. The second time (yesterday), it showed the same as it had before disconnecting – ie 25km range, 1/3 battery and no turtle (perhaps I didn’t leave it disconnected as long/long enough…).

The tortoise is generally set as a warning when SoC drops below 10%, another reason is battery overtemperature or like in your case a cell voltage dropping below 3VDC (either due to a weak cell or faulty CMU data). This usually happens under a heavy load (acceleration or driving uphill) and generally stops once you ease off the accelerator.

P1A2F (Cell Error, BMU) P1A4B (Voltage of each Battery Cell Abnormal)

The vehicle had shown an HV error (car!) about a week earlier. That was cleared and has not been seen since. I didn’t clear the codes but made these notes. P1A2F, P1A4B, U1111. This was before the first turtle event. My buddy read the following only about 5km before the turtle yesterday (ie before the 2nd turtle event):
- P1B01, stored DTC, Quick Charger vehicle error;
- P1AF6, stored DTC, On board charging time over;
- U1111, Active DTC, Display CAN timeout, Not equipped.
The U1111 could not be cleared. The two stored DTCs were cleared.

Ignore U1111, the other two are CHAdeMON related and are unlikely to have caused your current issue.

Any suggestions?

You need to have a close look at your cell voltages under load @ 1/3 battery

 

[Trevor Jack]

Thanks Mickey.

Both turtle incidents occurred going uphill. In neither case did stopping or getting away from the hill bring things back to normal.

A buddy has an analyser with which we can monitor cell voltages. Could I be reasonably confident of getting the 50km range I've got before turtling these last two times out? I charged the vehicle when I got home last time so I'd need to do 50km or so to get to 1/3 SoC. If I could confidently do, say, 40km, I'd do that to get the battery SoC down then go to a nearby steep hill to load it up. Or perhaps just put it in drive with hand brake on (near my regular home charging port)?

 

[MickeyS70]

Both turtle incidents occurred going uphill. In neither case did stopping or getting away from the hill bring things back to normal.

Strange, a weak cell usually recovers more or less instantly once the load is removed?

A buddy has an analyser with which we can monitor cell voltages. Could I be reasonably confident of getting the 50km range I've got before turtling these last two times out? I charged the vehicle when I got home last time so I'd need to do 50km or so to get to 1/3 SoC. If I could confidently do, say, 40km, I'd do that to get the battery SoC down then go to a nearby steep hill to load it up. Or perhaps just put it in drive with hand brake on (near my regular home charging port)?

It seems the problem starts once the battery voltage drops below a certain level (around 1/3 SoC), you should be fine above that.

Rather than risking IGBT failure with your handbrake idea, bring a second car to tow the imiev back (in case it won’t recover after being switched off for a while).

 

[Trevor Jack]

Thanks again Mickey. I've done some load tests.

I ran at full power, albeit briefly, with a buddy reading module and cell level voltages. Range when we started was about 32km – so slightly more than 1/3 (with 75km taken as full battery range). We finished at about 1/3 (25km). Each module showed about 3.9v/cell (31v/module ...and comparable for the 4 cell modules) when stationary and close to 3.1v/cell (25v/module) at full load. One module was closer to 3.0v/cell. Cell voltages were within about 0.2v range within modules. I drove home, cautiously, keeping the power gauge in the green, about 8km. We did further tests at home, starting with a range of 17km (~23%). Module level voltages were close to 26v when stationary. The car did not develop full power even on short acceleration tests. The vehicle power gauge got to perhaps ¾ then dropped back. My partner observed 5v “for one frame” in module 2. All others dropped to around 23-25v under load as did module 2 apart from the single 5v frame (which seems likely to have been an aberration). The vehicle seemed close to turtling when we got home. Seems like I should assume that the range is from 75km to 25km - ie 50km range - but not much more. Any suggestions, of course, will be eagerly considered. At 75km range the vehicle works for me. At 50km range it’s a lot harder for it to be worthwhile.

The ABS and handbrake warning lights and TCL came on at various times when we were testing - there was no heavy braking. All off again now and no codes showing.

 

[MickeyS70]

. Any suggestions, of course, will be eagerly considered. At 75km range the vehicle works for me. At 50km range it’s a lot harder for it to be worthwhile.

Very comprehensive testing!
There seem to be no obvious outliers and therefore only a complete pack replacement/upgrade will increase your useable range, unfortunately.

 

[Trevor Jack]

Very comprehensive testing; there seem to be no obvious outliers and therefore a complete pack replacement/upgrade is required to increase your useable range..

Thanks Mickey. It's not exactly what I was hoping to hear....but is what I was expecting... Bummer.

 

[MickeyS70]

Thanks Mickey. It's not exactly what I was hoping to hear....but is what I was expecting... Bummer.

Most likely the car is also ‘overestimating’ the remaining battery capacity as 75km for a 2010 Imiev with the original LEV50 cells sees a tad optimistic?

I would no be surprised if your RR will reduce after a battery calibration, this won’t improve your range but should prevent you from being left stranded with a permanent tortoise..

 

[Trevor Jack]

Mickey, Could you direct me to any documentation on doing a battery calibration. If that makes the range estimate realistic, albeit lower than I'd like, that would be worthwhile. (I'm just glad this has happened with me driving rather than my wife....her "finding" the problem would not have been good.)

 

[MickeyS70]

Could you direct me to any documentation on doing a battery calibration.

There are two ways to do this:

1) gradually, according to the manual:
“The progress of the battery capacity loss depends on the vehicle usage and environment. We recommend to do regular charging from 2 segments or less to charge completely at least once every 3 months. The procedure lets the battery remaining indicator adjusted automatically.”

Above method works, but can be incredibly slow (if the value is way out) as from experience, it only changes the capacity about +/- 2Ah/month

2) In 'one go' using Hobdrive app, or MUT3/Diagbox diagonstic tools, same procedure as above, except the new calculated value is set immediately.

Note: ODBZero app has a CAP2 function that will accurately calculate the actual capacity but can't change the value in the BMU, this gives you at least an indication where you stand..

 

[Trevor Jack]

Thanks Mickey.

I could get stuck several times if I have to go to fewer than 2 segments with procedure #1.

The only local I’ve found who could use MUT3/DiagBox does not recommend the procedure as he doesn’t think it produces any better information. (As a general rule I’m inclined to believe people who advise against doing something when such recommendation is clearly not in their commercial interest…but….?)

I’m quite happy to dive into this and spend a bit of money on it…..though time precludes diving too deeply for at least a month. This post, https://myimiev.com/threads/what-actually-is-mut-iii-in-detail.4144/, gets a little complicated. Is that the best source of information on the software/hardware that’s required?

The guessometer (you referred earlier to RR – is that Range Remaining?) seems to be moving faster, in both directions, with energy flows than it used to. For example, it might increase marginally (a few km) from a relatively short downhill run that would previously have not made any difference. It will also suddenly drop a few km going up a small hill. If I know to basically ignore the guessometer but can have (reasonable) confidence in the battery remaining gauge, that would be fine. From recent experience, I’m starting to think this is reasonable – however, I’m cautious given that on the first turtle event, battery remaining showed about 1/3 but was then zero after disconnecting and reconnecting the 12v battery.

Is there any documentation covering the guessometer algorithm? That's more for interest than for practical use. Perhaps on a more practical level, how is the battery gauge calculated? Does it count coulombs in and out or is it voltage based - surely it would need to be the former?

 

[MickeyS70]

I could get stuck several times if I have to go to fewer than 2 segments with procedure #1.

You have an underlying issue that stops the charging process from succeeding

The only local I’ve found who could use MUT3/DiagBox does not recommend the procedure as he doesn’t think it produces any better information. (As a general rule I’m inclined to believe people who advise against doing something when such recommendation is clearly not in their commercial interest…but….?)

He’s correct, this won’t work unless you can charge the car in one go to 100%

I’m quite happy to dive into this and spend a bit of money on it…..though time precludes diving too deeply for at least a month. This post, https://myimiev.com/threads/what-actually-is-mut-iii-in-detail.4144/, gets a little complicated. Is that the best source of information on the software/hardware that’s required?

The sky is the limit when it comes to diagnostic HW/SW solutions for an I-Miev. For the occasional user, imho, the combination of a low cost Bluetooth dongle and HobDrive (paid version on Android) is hard to beat.

The guessometer (you referred earlier to RR – is that Range Remaining?) seems to be moving faster, in both directions, with energy flows than it used to. For example, it might increase marginally (a few km) from a relatively short downhill run that would previously have not made any difference. It will also suddenly drop a few km going up a small hill.

The displayed RR is calculated based on consumption of the past 15miles driven and current SoC, it’s quite accurate if your SoC is correct and the drive profile remains similar. What you describe isn’t unusual.

If I know to basically ignore the guessometer but can have (reasonable) confidence in the battery remaining gauge, that would be fine. From recent experience, I’m starting to think this is reasonable – however, I’m cautious given that on the first turtle event, battery remaining showed about 1/3 but was then zero after disconnecting and reconnecting the 12v battery.

Is there any documentation covering the guessometer algorithm? That's more for interest than for practical use. Perhaps on a more practical level, how is the battery gauge calculated? Does it count coulombs in and out or is it voltage based - surely it would need to be the former?

The battery gauge displays SoC or current battery capacity. It’s updated using coulomb counting method.

Disconnecting the 12V aux will result in a SoC re-calculation based on battery voltage, if it dropped from 1/3 to 0 afterwards, the assumed battery capacity value was too high and needs to be corrected doing a battery calibration (after you fix your charging problem)