Page 2 of 2

Re: Battery current sensor

Posted: Mon Sep 14, 2020 8:02 am
by coulomb
I've been asked to look into this for a company that has done this already, and sees demand for many more if it can be solved. They are using NMC cells, which have a lower voltage characteristic, so the BMS learns a lower capacity that is actually present.

Also, these are 93 Ah cells, and I read somewhere that the limit on this value (or perhaps reporting of this value, it wasn't clear) is limited to 60 Ah.

With 30 km on the GOM, the lowest cell voltage was 3.825 V, meaning that there is still roughly 50% capacity left.

I'd certainly be happy if the BMS will eventually learn a higher capacity, perhaps even something approaching actual capacity, but it doesn't seem likely.

The proper way to solve this would be to modify the BMS to take into consideration the characteristics of the new cells and their initial capacity, but that option doesn't seem open to us. I patch solar inverter firmware, so I'd be prepared to try this, but getting the image to analyse and patch is the problem.

It was pointed out to me that we may have to change the current reported on the CAN bus to other ECUs to prevent other problems. I intend to try a gadget soon that should allow a reduction of measured current by 10-50% (adjustable). I hope that the 10% figure will be noticeable as a range improvement, and if it affects other systems, a 10% change should be noticeable, but hopefully not damaging.

If the digital adjustment is necessary, it would be nice to also adjust the GOM figure, a la Leafs. But I've not seen a CAN Id for the GOM display. Anyone seen one?

Re: Battery current sensor

Posted: Mon Sep 14, 2020 8:37 am
by kiev
i had some notes about CAN pids, i think it is pid x346. i haven't verified this personally. Canion reports the RR (GOM) value so it knows the pids to gather the data.

29A: VIN

346: range remaining
373: battery voltage and current
374: SOC
384: heating/cooling current
389: charger current and line voltage
3A4: settings of HVAC console

Re: Battery current sensor

Posted: Mon Sep 14, 2020 3:17 pm
by coulomb
Thanks, Kenny. The can triplet database confirms, but doesn't seem to have the data format. It should be fairly easy to guess given some data. I might even be able to figure it out from some of the published massive CAN data dumps.

[ Edit: this post seems to say that it's the first data byte, so the range of the range meter would be 0-255 km (0 - 158 miles). Hopefully that will be enough :D ]

BTW, does the Range Remaining meter display in miles in North America? I assume that the display processor does the conversion, and the CAN data remains in kilometers.

Re: Battery current sensor

Posted: Tue Sep 15, 2020 5:15 am
by PV1
Yes. US I-MiEVs show miles for RR. We’re stuck with 24-hour only clocks, but it does show miles for everything.

I haven’t heard about the 60 Ah limit. Good to know.

Re: Battery current sensor

Posted: Tue Sep 15, 2020 6:03 am
by kiev
coulomb wrote:...so the range of the range meter would be 0-255 km (0 - 158 miles). Hopefully that will be enough .


Hopefully it won't be enough if you can fit higher capacity cells--don't limit yourself, too much is not enough.