Aftermarket iMiEV battery upgrade project

Mitsubishi EV Interest Group
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: Aftermarket iMiEV battery upgrade project

Post by coulomb »

That current sensor (and much else) looks rather different to the one in the iMiEV: http://myimiev.com/forum/viewtopic.php? ... 509#p41509
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
fifurnio
Groupie
Posts: 36
Joined: Thu, 25 Mar 2021, 06:54
Real Name: franco vaccari

Re: Aftermarket iMiEV battery upgrade project

Post by fifurnio »

It is not at all rare that a spare part for the same car model has variations over time, perhaps maintaining compatibility with previous models (a certain item code is replaced by another and then by others) it may be an evolution of the project with another supplier or change of code of a piece that tends to break , or one of the adaptations made to install the drive chain on a different vehicle, the CAN protocol could also be one of these adaptations and be different from that of i-Miev.
In this regard has anyone done the hacking of the CAN protocol?
nuggetgalore
Groupie
Posts: 365
Joined: Sun, 11 Dec 2016, 18:02
Real Name: Andreas Tobler
Location: Rowville 3178 V
Contact:

Re: Aftermarket iMiEV battery upgrade project

Post by nuggetgalore »

coulomb wrote: Thu, 25 Mar 2021, 17:20 That current sensor (and much else) looks rather different to the one in the iMiEV: http://myimiev.com/forum/viewtopic.php? ... 509#p41509
And a lot of things are different between the 2010 and later(2012) models.
For instant the contactors on mine are nowhere near the same as the ones on fifurnio's 75 Ah pack that look more like the ones on the 2012 models iMiEV
Image
Except it appears there are no fast charge contactors only the negative and positive precharge and positive main (edit:on the 75 Ah vehicle, but I assume they are somewhere else?).
T1 Terry
Senior Member
Posts: 1952
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Aftermarket iMiEV battery upgrade project

Post by T1 Terry »

The range extender type battery in my Prius uses a higher voltage battery to top up the traction battery. Contactors close when the on traction battery drops to a preset voltage and open when the 3/4 full battery voltage is reached, in the case of the Prius, that is 228v.
If the range extender was connected to the output side of the hall effect sensor and the contactor only closing when the traction battery was getting close to turtle mode, would that do the job by putting the extended range battery capacity back into the traction battery and fool the on board system into seeing the traction battery magically recharge :lol: This could be a manual operation to verify if it works and trying different scenarios to find the best time to do the energy transfer, when the vehicle is stopped, on regen or what ever ...

T1 Terry
Green but want to learn
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

You can not just connect the two batteries together without matching their voltage.
If one battery is charged and the other discharged the voltages will be different. If joined there will be lots of current flowing (possibly damaging the contractors and the batteries)
If a battery has an internal resistance of say 0.1Ohms and you have one at full charge say 360v and the other discharged at say 270v then you would have about 450A when you close the contactor.
360 -270 = 90v. 90v into 0.2ohms = 450A. 0.2 ohms because you have two batteries.
So you need to wait until both batteries are about the same voltage before you close the contactor and them make sure both stay at the same voltage.
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

Coulomb tried to fool the current sensor but for some reason it didn't work. I never tried to find out why but in my opinion it should have worked.
You can not send messages to try to override the ones sent by the BMU because CAN will make sure both are received.
I have done a fair bit with the CAN messages on the Imiev which happen to be the same as the minicab. It may be worth trying the same messages on yours.
One way to get around the problem (if you are prepared to be nasty) is to turn the power to the BMU off. It will then loose it's memory on state of charge and it will make an estimate using the battery voltage. So if the SOC is at 0 and your battery is only half discharge and you cycle the power the SOC will go to about 50% when you put the power back on. I have tested this already. It was just to see how it initialises the SOC counter.
Estimating SOC is not as simple as just counting current. You have to have a way of dealing with leakage and offsets which integrate over time and also a way to make sure you do not damage the battery if your count gets out of whack.
T1 Terry
Senior Member
Posts: 1952
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Aftermarket iMiEV battery upgrade project

Post by T1 Terry »

francisco.shi wrote: Fri, 26 Mar 2021, 10:40 You can not just connect the two batteries together without matching their voltage.
If one battery is charged and the other discharged the voltages will be different. If joined there will be lots of current flowing (possibly damaging the contractors and the batteries)
If a battery has an internal resistance of say 0.1Ohms and you have one at full charge say 360v and the other discharged at say 270v then you would have about 450A when you close the contactor.
360 -270 = 90v. 90v into 0.2ohms = 450A. 0.2 ohms because you have two batteries.
So you need to wait until both batteries are about the same voltage before you close the contactor and them make sure both stay at the same voltage.
Theoretically speaking, you are correct, if you ignore voltage rise in the battery being charged and voltage drop in the battery supplying the charging current. No idea how fast you can charge an i-Miev battery but you certainly couldn't get 450 amps out of push bike batteries without a savage voltage drop ..... I think the theoretical 90v difference will drop significantly and act as a current limiter.

The Prius extended range battery is 270 fully charged, the system closes the contactor at around 198v mark in the traction battery and opens it again when the traction battery reaches 228vdc. It will continue to do this clunk on/off until the extended range battery drops to roughly 200v at which point the system will lock out the extended range battery until it gets recharged. The average current transfer starts around the 100 to 140 amp mark but drops significantly the longer the contactor remains closed as the two batteries get closer to voltage balance.

T1 Terry
Green but want to learn
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

Why not leave the extended range battery on all the time?
fifurnio
Groupie
Posts: 36
Joined: Thu, 25 Mar 2021, 06:54
Real Name: franco vaccari

Re: Aftermarket iMiEV battery upgrade project

Post by fifurnio »

Connect two batteries with a big different voltage level is a bad idea, and I think is not useful for my target.

Prius energy strategy probably is different than i-Miev, I think it's a waste of time to consider this strategy

Overwriting on the CAN line is possible, but it is not sure if it is effective and the result is obtained, in some cases the network goes into error (two different values arrive on an ID ...) and everything must be turned off to obtain a reset.
Nonetheless, it is a very quick and convenient way to test the effect of varying some values.
With this method I tested the response of the instrument cluster of the car which I converted to EV
If you look at the link you will find that the tachometer has become an ammeter, the fuel level indicator has become a SOC indicator (the SOC is calculated by an Orion BMS) and the engine thermometer has become a battery thermometer, the various engine warning lights such as that of the oil pressure they mark traction system errors (I had to change all symbols)
T1 Terry
Senior Member
Posts: 1952
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Aftermarket iMiEV battery upgrade project

Post by T1 Terry »

francisco.shi wrote: Fri, 26 Mar 2021, 12:15 Why not leave the extended range battery on all the time?
I think the problem with the iMiev was the system didn't register the additional capacity. My suggestion was to add the extended range battery capacity back into the original battery by sending it through the hall effect sensor so the system sees the capacity return to the original battery just like it was being recharged.

T1 Terry
Green but want to learn
nuggetgalore
Groupie
Posts: 365
Joined: Sun, 11 Dec 2016, 18:02
Real Name: Andreas Tobler
Location: Rowville 3178 V
Contact:

Re: Aftermarket iMiEV battery upgrade project

Post by nuggetgalore »

T1 Terry wrote: Fri, 26 Mar 2021, 11:35



The Prius extended range battery is 270 fully charged, the system closes the contactor at around 198v mark in the traction battery and opens it again when the traction battery reaches 228vdc. It will continue to do this clunk on/off until the extended range battery drops to roughly 200v at which point the system will lock out the extended range battery until it gets recharged. The average current transfer starts around the 100 to 140 amp mark but drops significantly the longer the contactor remains closed as the two batteries get closer to voltage balance.

T1 Terry
As far as I understand,please correct me if wrong,
in normal life the iMiEV (and other EV) contactors,while indeed rated at very high Amps, do not close or open under heavy load with the exception of the precharge contactor who's current is limited by the precharge resistor ( Question: is the precharge contactor always on or gets de-energised shortly after the main contactor has closed)?
I just wonder what happens to the precious metal contact faces if they are forced to operate at 100 + A or even much less? The relays are hydrogen filled and have magnets to draw the arc off, so oxidation is no problem but still would get very hot.
The EV-ECU (I think) does a weld check on all 5 HV contactors, not sure when and how,but I assume if it finds damaged points it flags a fault and refuses to go to READY.
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

The problem of not being able to see the added capacity is solved by using a can bridge and calculating a new SOC and replacing the message from the BMU with the new value. I already did this successfully and it works as expected. The only thing that does not seem to work correctly is the range estimate. It changes but not quite right. So it starts with 100km of range but after doing 100km it shows it still has 50km left and the SOC is still 50% so if you put a battery twice the size the remaining range goes down at half the rate.
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

You are right about the contactors. They switch under zero current except for the precharge.
So switching under load (especially two batteries with different voltages) will wear them out.
T1 Terry
Senior Member
Posts: 1952
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Aftermarket iMiEV battery upgrade project

Post by T1 Terry »

The after market extended range battery in my '06 Prius uses 250amp rated Gigavac contactors, two of them, one checks if the other one opened and sets an annoying screeching alarm when it senses one of the contactors has welded shut. Something that happens in frequently at first, then with annoying regularity until you fit a new contactor.
Most of the welded shut problems are the result of me driving it like I stole it, just love that acceleration with the combination of the electric drive and the petrol engine, it really turns heads to see a Prius whistle past when over taking the many mobile road blocks that seem to clog the roads around these parts (under powered car towing an over sized caravan driven by a hat driver).
The next version battery will be an LTO 55Ah celled unit so I can get rid of the original Toyota traction battery(6Ah) and drive straight off the LTO battery pack, that should remove the clunking contactors and the noisy fan that cools the NiMh traction battery ... and get rid of the leaking electrolyte smell from the GBS cells.

T1 Terry
Green but want to learn
fifurnio
Groupie
Posts: 36
Joined: Thu, 25 Mar 2021, 06:54
Real Name: franco vaccari

Re: Aftermarket iMiEV battery upgrade project

Post by fifurnio »

francisco.shi wrote: Fri, 26 Mar 2021, 17:10 The problem of not being able to see the added capacity is solved by using a can bridge and calculating a new SOC and replacing the message from the BMU with the new value. I already did this successfully and it works as expected. The only thing that does not seem to work correctly is the range estimate. It changes but not quite right. So it starts with 100km of range but after doing 100km it shows it still has 50km left and the SOC is still 50% so if you put a battery twice the size the remaining range goes down at half the rate.

I suppose that it is just the calculated SOC that limits the range, so if a modified SOC is sent to the system I should get the result of an actual increase in the range, which I really want to achieve.
If the CAN bridge in addition to this gets a correct indication of the SOC on the instrument panel I could also ignore the indication of the remaining range in chilometers
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

The car uses the reported SOC to decide when to go to turtle mode and when to stop driving. The SOC reading on the dash displays the reported SOC. So the can bridge will give you the correct SOC display and the correct behavior.
nuggetgalore
Groupie
Posts: 365
Joined: Sun, 11 Dec 2016, 18:02
Real Name: Andreas Tobler
Location: Rowville 3178 V
Contact:

Re: Aftermarket iMiEV battery upgrade project

Post by nuggetgalore »

francisco.shi wrote: Fri, 26 Mar 2021, 17:43 The car uses the reported SOC to decide when to go to turtle mode and when to stop driving. The SOC reading on the dash displays the reported SOC. So the can bridge will give you the correct SOC display and the correct behavior.
Yes all of that is most probably true until it isn't.
Because as I posted earlier,the age of the battery was reset(by accident) and that in turn reset the Ah capacity to what a *new* standard 88 cell iMiEV has. With the (degraded to ~65%) original cells back in,it happily charges to 360+ V ( showing 16 bars,tank full),but lying to me it can go for 120 - 130km yet emptying the tank in 60 km. I got caught out and drove it (once) until it nearly killed the cells without the tortoise sticking it's neck out!
So what I am really try to say is that there are more factors involved in how the EV-ECU and BMS calculate SoC and RR and when one or more parameters are out of spec one cannot rely on the result. I really would like to find out when and where the coulomb counting takes place, it does not add up with the reported Ah (BMS) or the energy input into the charger as measured with a dedicated meter from the switchboard (assuming about 10% charger etc loss).
fifurnio
Groupie
Posts: 36
Joined: Thu, 25 Mar 2021, 06:54
Real Name: franco vaccari

Re: Aftermarket iMiEV battery upgrade project

Post by fifurnio »

I think we are here just to share informations about this
fifurnio
Groupie
Posts: 36
Joined: Thu, 25 Mar 2021, 06:54
Real Name: franco vaccari

Re: Aftermarket iMiEV battery upgrade project

Post by fifurnio »

nuggetgalore wrote: Fri, 26 Mar 2021, 19:55
francisco.shi wrote: Fri, 26 Mar 2021, 17:43 The car uses the reported SOC to decide when to go to turtle mode and when to stop driving. The SOC reading on the dash displays the reported SOC. So the can bridge will give you the correct SOC display and the correct behavior.
Yes all of that is most probably true until it isn't.
Because as I posted earlier,the age of the battery was reset(by accident) and that in turn reset the Ah capacity to what a *new* standard 88 cell iMiEV has. With the (degraded to ~65%) original cells back in,it happily charges to 360+ V ( showing 16 bars,tank full),but lying to me it can go for 120 - 130km yet emptying the tank in 60 km. I got caught out and drove it (once) until it nearly killed the cells without the tortoise sticking it's neck out!
So what I am really try to say is that there are more factors involved in how the EV-ECU and BMS calculate SoC and RR and when one or more parameters are out of spec one cannot rely on the result. I really would like to find out when and where the coulomb counting takes place, it does not add up with the reported Ah (BMS) or the energy input into the charger as measured with a dedicated meter from the switchboard (assuming about 10% charger etc loss).
sorry if I have not read the whole thread, this information is important: therefore it seems that the system calculates the SOC based on the energy subtracted from the stored capacity, surely in addition to this there will be a calculation mechanism of the SOH, the SOH then goes to decrease as a percentage of the maximum withdrawable capacity.
I understand the choice: calculating the SOC only using voltage is difficult, it would take a system that compensates the internal resistance of the battery and the current (or the value would fluctuate with the variation of the current) and this in turn should be compensated with the temperature.
The Orion BMS that I used on the Fiat uses a complex system that uses both voltage and current values, and sets the SOC at 100% when the battery reaches the maximum charge voltage, but when the cell limit voltages are reached intervenes by limiting the output and input currents by sending orders to the inverter via CAN
However, I find it very strange that your BMS did not intervene at too low a cell voltage level.
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

The coulomb counting is done by the BMS. It then sends the calculated SOC to the car.
I did notice that the BMS does not appear to look at the cell voltages to try to protect the battery. I don't really know why. It may have been poor design.
What I do with my SOC calculation is I use the coulomb count but also check the cell voltages and the battery voltage. I keep two SOC estimates one based on coulomb count and one based on battery voltage. I only update the voltage based SOC when the current is below a small value. This gives me a rough OC voltage reading. If the voltage referenced SOC gets to below 20% I start interpolating between the two SOC values. The idea is that you do not suddently drop to zero without warning. I also reset the coulomb counter when the battery is fully charged. The charger does not care what the size of the battery is. All it does is puts out a constant current then constant voltage until the current drops to a low value for long enough then it turns off. So the charger will charge any capacity battery without need to modify. The BMS is active during charging and it updates the SOC.
nuggetgalore
Groupie
Posts: 365
Joined: Sun, 11 Dec 2016, 18:02
Real Name: Andreas Tobler
Location: Rowville 3178 V
Contact:

Re: Aftermarket iMiEV battery upgrade project

Post by nuggetgalore »

francisco.shi wrote: Sat, 27 Mar 2021, 06:30 The coulomb counting is done by the BMS. It then sends the calculated SOC to the car.
So it only counts the Ah in,and ignores the Ah used?
The CanIon app does show Wh in and out when driving and recharging, I should have checked if this also shows when charging,I do that tomorrow.
francisco.shi wrote: Sat, 27 Mar 2021, 06:30 I did notice that the BMS does not appear to look at the cell voltages to try to protect the battery. I don't really know why. It may have been poor design.
I am sure that with a properly maintained original battery all the ducks line up and it works fine. Us trying to outguess the design engineers with different and/or bigger batteries need to take some additional precautions.
francisco.shi wrote: Sat, 27 Mar 2021, 06:30 What I do with my SOC calculation is I use the coulomb count but also check the cell voltages and the battery voltage. I keep two SOC estimates one based on coulomb count and one based on battery voltage. I only update the voltage based SOC when the current is below a small value. This gives me a rough OC voltage reading. If the voltage referenced SOC gets to below 20% I start interpolating between the two SOC values. The idea is that you do not suddently drop to zero without warning. I also reset the coulomb counter when the battery is fully charged. The charger does not care what the size of the battery is. All it does is puts out a constant current then constant voltage until the current drops to a low value for long enough then it turns off. So the charger will charge any capacity battery without need to modify.
In a couple of days I should find out what happens when I parallel connect the original pack (approx 30 Ah SoH) with the *60 Ah* aftermarket battery (probably just a bit less than 30 Ah usable). With the (erroneously) reported SoH of about 45 Ah it may be interesting....
According to the manual, 60 Ah is the top range that may be set or read.
francisco.shi wrote: Sat, 27 Mar 2021, 06:30 The BMS is active during charging and it updates the SOC.
SoH is at least partly depreciated by an unknown , probably non linear factor by the number of months since the manufacture or installation (resp. resetting said number of months) of the battery. What role, if any, the coulomb counting up to the max cell voltage plays would be interesting to know. The BMS definitely gets confused (as far as SoC is concerned) with cells of different chemistry than the original LEV50s (voltage rise not matching the count during charging or is it that the balance drives could not handle the mismatch of cell capacity).
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

The coulomb count works both ways in and out.
If the count reaches 100% and the battery keeps taking current in the count stops at 100%. The same goes for discharge. Once the count gets to zero it stays at zero even if current continues to come out of the battery.
When the power is lost and the BMS looses the coulomb count it first uses the battery voltage to estimate the SOC.
I think not looking at the cell voltage is a bad decision because there are cases when the coulomb count can get out of whack and it could potentially damage the battery. There may be some other safety somewhere else that I haven't noticed.
nuggetgalore
Groupie
Posts: 365
Joined: Sun, 11 Dec 2016, 18:02
Real Name: Andreas Tobler
Location: Rowville 3178 V
Contact:

Re: Aftermarket iMiEV battery upgrade project

Post by nuggetgalore »

francisco.shi wrote: Sat, 27 Mar 2021, 19:27 The coulomb count works both ways in and out.
If the count reaches 100% and the battery keeps taking current in the count stops at 100%.

With the NMC 60Ah cells both the SoC (control) and (display) sometimes read up to 105% on the scanners or the EvBatMon app.
francisco.shi wrote: Sat, 27 Mar 2021, 19:27 The same goes for discharge. Once the count gets to zero it stays at zero even if current continues to come out of the battery.
I never saw SoC (control) and/or (display) at zero, even the time the car stopped or the other day when it slowed down to creep speed.
francisco.shi wrote: Sat, 27 Mar 2021, 19:27 When the power is lost and the BMS looses the coulomb count it first uses the battery voltage to estimate the SOC.
I think not looking at the cell voltage is a bad decision because there are cases when the coulomb count can get out of whack and it could potentially damage the battery. There may be some other safety somewhere else that I haven't noticed.
francisco.shi
Senior Member
Posts: 826
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: Aftermarket iMiEV battery upgrade project

Post by francisco.shi »

I was reading the SOC from the CAN messages coming from the car's BMS.
I was using the can bridge and was reporting my own SOC. Because the batteries were 94Ah instead of 50Ah and I was reporting a SOC based on 94Ah the original BMS would eventually get to zero but it would never go negative.
The turtle mode works by reducing the torque command to the motor as you get close to 0 SOC. I think it starts reducing from about 5% so you will never really get to zero unless you keep the accelerator to the floor until the car stops.
I tried that by reporting a lower SOC. (Just so that I could find out what made the car stop)
I also checked the SOC reported by the BMS while charging and once it gets to 100% it stops. (When charging the 94Ah cells I could get more than 50 Ah thru the counter)
fifurnio
Groupie
Posts: 36
Joined: Thu, 25 Mar 2021, 06:54
Real Name: franco vaccari

Re: Aftermarket iMiEV battery upgrade project

Post by fifurnio »

someone in this forum has posted something about the CAN protocol, in particular the BMS?
What hardware for the CAN bridge ?
Post Reply