GdB
Posts: 154
Joined: Tue Aug 12, 2014 4:06 pm

Re: Let the Hacking Begin

Wed Oct 24, 2018 5:07 pm

jray3 wrote:I just found this university project that is building a microbus on the chassis of an i-MiEV. Don't think it's been posted yet.
https://iot.ttu.ee/iseauto-saamislugu-1/


If they use some open source as a basis, maybe we will get access to the be able to use it also for a self driving autopilot.

https://github.com/commaai/openpilot
My EVs: 2012 Mitsubishi i-MiEV and 2019 Hyundai Kona Electric SEL

drcat
Posts: 9
Joined: Tue Dec 08, 2020 3:21 am

Re: Let the Hacking Begin

Mon Mar 08, 2021 1:23 pm

With i-miev, you have no hardware feature that could do steering itself. No dice. Adaptive cruise is maximum of what could be done with it and not very reliable without at least extra ultrasonic sensors (for emergency braking if AI screwed up something)
It's easy to spoof the accelerator pedal (I did a super small arduino-based project for that for a non-adaptive cruise control that I'm using for daily driving for years already, but not on i-miev)

fifurnio
Posts: 2
Joined: Sat Mar 20, 2021 11:43 am

Re: Let the Hacking Begin

Sat Mar 20, 2021 12:53 pm

Hello everyone

I just joined this forum
Hope I'm in the right thread
For a year I have had a Peugeot Partner EV that shares the entire i-Miev drive chain with Citroen Berlingo
My car has a 22.5kWh battery pack and I believe it has 80 Yuasa LEV 75 cells
I inserted a diesel heater to limit the autonomy cut, inserted in the central tunnel.
But the reason I signed up is another ....
I built a range extender consisting of an additional 7 kWh battery, inserted in the trunk with a thickness of 6 centimeters using 24 e-bike batteries recovered for BMS failure
My car battery is split into two units, I have already opened and inspected both, the components look the same as the i-Miev, obviously with a different arrangement.
I connected my extender in parallel to the original battery, but the traction management stops my vehicle with half-discharged batteries after traveling the same distance as before.
Now to overcome this problem and get the range of 200-210 km that I should have, I have to overcome the limitation due to traction control; in addition to this I would like to obtain a correct indication of the battery level and the remaining autonomy: now both mark values ​​as if my extender were not connected.
I am wondering if it is better to do a test by connecting my extender "after" the current Hall sensor in the original battery box (now it's directly connected to the pole of the battery pack), or to attenuate the signal read by the Hall sensor or rather to hack the protocol can on board and insert a can-can bridge that allows me to send the real data of the battery capacity to the on-board electronics (as long as the system accepts to calculate the autonomy with a battery of non-standard capacity).



https://www.forumelettrico.it/forum/ran ... 30-10.html
the Peugeot extender

https://www.forumelettrico.it/forum/pos ... tml#p64104
my EV conversion of a Fiat Doblò

JoeS
Site Moderator
Posts: 3996
Joined: Thu Dec 15, 2011 5:47 am
Location: Silicon Valley, California

Re: Let the Hacking Begin

Sat Mar 20, 2021 6:00 pm

fifurnio, welcome to the forum.

I wasn't aware of the Mitsu i-MiEV drivetrain being used in anything other than the Peugeot iOn and Citroen C-Zero. Interesting.

Congratulations on paralleling two different packs! We've often talked about it, but you've actually done it! Do you have any current monitors and how well do the two packs share their current? Does your auxiliary pack have any internal BMS? Where exactly did you attach this parallel pack?

In regards to your question and purely speculating, and as you had suggested, it would seem to me that a crude but perhaps simple spoof of the i-MiEV Hall sensor circuit would be to reduce its gain to be proportional to your capacity increase [i.e. 22.5kWh/(7kWh + 22.5kWh)] which might make the car think that you're using it's energy at that proportionately-reduced rate.
EVs: Two '12 Wht/Blu SE Prem., '13 TeslaMS85, three 156v CorbinSparrowsLi(NMC), 24v EcoScoot(LiFePO4)
EV Conv: 156v '86 Ram PU, 144v '65 Saab96
Hybrids: 48v1kW bike
Mothballed ICE: Orig.Owner '67 Saab96V4, '88 IsuzuTrooper; '76 MBZ L206D RHD RV

fifurnio
Posts: 2
Joined: Sat Mar 20, 2021 11:43 am

Re: Let the Hacking Begin

Sun Mar 21, 2021 12:34 pm

JoeS wrote:fifurnio, welcome to the forum.

I wasn't aware of the Mitsu i-MiEV drivetrain being used in anything other than the Peugeot iOn and Citroen C-Zero. Interesting.

Congratulations on paralleling two different packs! We've often talked about it, but you've actually done it! Do you have any current monitors and how well do the two packs share their current? Does your auxiliary pack have any internal BMS? Where exactly did you attach this parallel pack?

In regards to your question and purely speculating, and as you had suggested, it would seem to me that a crude but perhaps simple spoof of the i-MiEV Hall sensor circuit would be to reduce its gain to be proportional to your capacity increase [i.e. 22.5kWh/(7kWh + 22.5kWh)] which might make the i-MiEV think that you're using it's energy at that proportionately-reduced rate.


I put the two packs in parallel, I made a first attempt by connecting one of the two poles in common (now I don't remember if the positive or the negative) and the other "after" the Hall current sensor, but immediately I had an error that blocked the machine, luckily I got the original diagnostic tool and software and cleared the error.
Then I tried to connect the poles of the two batteries together (the power relay is delayed by 2 seconds compared to the original relay through a timer) and there were no errors, but at the moment the range of the machine has not increased...
I don't know if the SOC and autonomy calculation system are the same between I-Miev and my car (the ECU is Mitsubishi though ...) but I hope to find some help here.
I believe that the calculation of the autonomy and SOC are done by subtracting from the state of "full battery" (therefore the nominal capacity of 22.5 kWh with the cells at 4.1 V) the energy measured by the sensor on the battery, so when this has measured 22.5 kWh the machine stops even if the battery is not discharged.
The limit voltages of the original 80s battery are:
255 v SOC 0% (3.18 v per cell)
328 v SOC 100% (4.1 v per cell, verified at end of charge)

With the extender connected I tried to do the range test until the vehicle stopped, 156 km, at 146 the light with the tortoise light came on, about a km before the stop an important limitation of performance intervened, then it went off everything and I had to push the car all the way home. Cells was up to 3,6 v, not at zero SOC level
Returning to my project: I think that the first error occurred was caused by the detection of an anomalous current from the Hall sensor caused by a voltage difference between the two batteries, in addition to this I had inserted a relay that connected the two batteries together with the original relay single on the front battery box, and this was not good because I found that this closes together with the precharge relay (the precharge relay trips before the power relay and I should connect simultaneously with this) and so I definitely made a timing error : my pack connected to the drive chain without precharge.
Now the two batteries have equal voltage levels and I will try to redo the test by connecting the extender "after" the Hall sensor as in the first attempt, but with a 2 second delay; in this way I should have the correct indication of the SOC in the instrument panel (the current from my extender would not be measured), this unless a current check is performed between the current measured by the sensor on the battery and the current flowing in the module distribution above the engine.
Abut balancing: I used four of those active balancers working with bootstrap scheme, they are 4 x 21s, each board connects with the neighboring one sharing the extreme cell.
I'm still not sure if these devices keep the promised balancing current data, I still have to set up a monitoring and surveillance system of the voltage values of each single cell (a 80s BMS is expensive more than the cost of whole battery pack...).
I will have to do it: both for safety reasons and for checking the status of the system, since I have used recovered batteries it is likely that some of these will need to be replaced over time if their performance is constantly different from the average.
For now I have verified that the system charge stops with 4.1 v cells, that despite the imbalance of my extender-pack the unbalance values do not exceed 50 mV, and I have already inserted 10 thermal switches that disconnect the relay power when the temperature exceeds 46 ° C

Return to “Modifications & Accessories”