openinverter forum

I'm beginning to doubt if the Nissan Leaf is a good source of batteries for EV conversions.

I'm wondering if they're unlikely to be in the best condition due to the lack of active battery cooling in the design of the car itself.

And I expect it could be challenging to add cooling when no provision for it was part of the original design.

Crashed Leafs sell for about half the retail value. And 90% of that is the value of the batteries alone.

I think these things are probably more popular as off-grid power walls than EV conversion donors. And the prices don't seem to be coming down despite availability going up.

I noticed some uneven voltage across cells. So they all started at 3.9V and after some discharging some would sit at 3.5 and some at 3.45V. Definitely a sign of depletion. Will know more when the BMS is in place.

Have a look at Wolf's "Pack Sniffer 2" Leaf battery tool on DIY electric... it plugs into the I/O connector on the battery pack and reports cell voltages, SOH, SOC, and other data. We use it on Leaf packs that have been removed from cars and on cars that don't power up. I believe that the latest software can also be used as a complete battery controller for the Leaf BMS, contactors, shunt, etc.

For the record we have purchased many Leaf packs and have never had a problem with battery health. Lack of cooling has also not been a problem unless you're pushing the pack very hard (worth remembering almost 500,000 Leaf's have been sold).

Just queried my BMS via CAN and SOH reads as 52%. Seems like I bought a lemon Byte 4 & 0x7F = 0xB4 & 0x7F = 0x34 = 52%

johu wrote: ↑Sun Oct 13, 2019 8:53 pm Just queried my BMS via CAN and SOH reads as 52%. Seems like I bought a lemon

Code: Select all

0x5BC DLC: 0x8 Data: 0xFF 0xC0 0xFF 0xFF 0xB4 0x42 0x5F 0xFF CRC: 0x720C

Byte 4 & 0x7F = 0xB4 & 0x7F = 0x34 = 52%

Hm... i wouldnt sag just yet J. It could be coded to record from 70% to 100% in smaller increments. You could have like 85% health.
Like i said use 6x or 9x H4 bulbs connected in parallel and measure time from 4V to 3V and calculate Ah from that. I did it for my used Kokam cells and i consistently got valid results. From 9 bulbs you get like 40A from single cell. That is a good load approximation.
You do have to prepare to do this though.

So...
1. Fill up all your cells up to 4V per cell.
2. Connect H4 bulb load directly to +/- of selected cell.
3. Measure voltage by good voltmer to disconnect load when cell hits 3V.
4. Once you start you have to do this for every cell in your car! Or fill up the measured cell by single cell charger.

At first i had to stay guard, but after like two cells were done i set myself a timer and when discharge was like 15min before end i went and checked on the cells. I got consistent 2hrs and 5min per cell +/- 2min. That means they were in some 85Ah usable range. It is a flintstones method, granted, but works without BMS. Yet you can use a time/voltage logger with BMS and also use shutoff trigger to disconnect load automatically.

A

Well, since I don't want to pull out the pack again I'll just get a BMS going (I have to, anyway) and charge to average 4V per cell. Then do it the fun way and drive until the first cell hits 3V. That gives me the total usable capacity and the remaining voltage in each cell even gives away per-cell SoH when looked up in the SoC curve.

I once spent several months bottom balancing a Leaf pack. I built an automated load very similar to this;

johu wrote: ↑Sun Oct 13, 2019 8:53 pm Just queried my BMS via CAN and SOH reads as 52%. Seems like I bought a lemon

Code: Select all

0x5BC DLC: 0x8 Data: 0xFF 0xC0 0xFF 0xFF 0xB4 0x42 0x5F 0xFF CRC: 0x720C

Byte 4 & 0x7F = 0xB4 & 0x7F = 0x34 = 52%

Well it turns out I misinterpreted the SoH byte
It actually stretches from Bit 7 to Bit 1. That makes it 0xB4 >> 1 = 0x5A = 90%
Sorry for the stir.