openinverter forum

teslas new(ish) 12v li-ion battery, a switch up from the old lead acids they used.
they look like a promising choice for a “12v battery “ in a conversion -CATL prismatic Cells
-1p4s
-99 Wh
-4 lbs / 1.8 kg
-16v internally
-14.8v at connectors?
-LINBUS
-internal BMS: monitors temp, controls charge and discharge
-part number 1598486-99-D

its theorized that the cells are the following specs:
-NMC
-6.9 Ah
-3.7 V nominal (4.2 V max charge voltage, 2.75 V minimum)
-charging temperatures: from -10°C to +45°C
-discharging temperatures: from -20°C to +60°
-max charging: 10C
-max peak discharging: 50C
-cycle life: >2,000 cycles (80% of initial capacity at 3C rate)


BMS uses stp410n4f7ag fets it looks like

any one have any LIN captures or data on talking to these batteries?

Nice find!
Last month i tried to connect 1S4P LiFePo4 CALB 200Ah cells to our Ampera car just to see what would happen.
When car was at a stop cells were kept at 13.2Vish. At the moment of start DCDC pushed 135A into those cells and voltage increased to 14.8V!!!
This would mean 3.7V per cell, but in reality three cells were at 3.6V and one was hanging at 4V!!!
Not a good prospect for LiFePo4 cells as replacement for AGM.

While Liion chemistry would work from 3.3V (sort of empty) to 4.2V (full) this means 13.2V to 16.8V range and one does not put more than 15V on legacy car components... just because. This should use aux battery only to about 60% SOC. A bit lame, but in reality aux battery is there only to supply 15min emergency power and power for starting the car electronics... of course supplying the car while stationary counts too.

Sounds like a bad cell!

looks like the internal bms will limit charge and discharge accordingly


Also looks like the battery talks to VCfront(vehicle control front, pdm/digital fuse box) via linbus, which then talks over can(?)

The lid can easily be removed if heat is applied to the edges to soften the epoxy

There is some type of Li-ion battery, and max voltage near 16v.

After deep discharge BMS locked battery, and need charging cell and reprogram BMS chip.

Looks like they “lock out” in a crash event and require a firmware reset.

The bms chip is a mm9z1_638
https://www.nxp.com/docs/en/data-sheet/MM9Z1_638D1.pdf

Could be reset by glitching reset pin or rewriting eeprom part

how to reflash with vvdi prog http://blog.obdii365.com/2024/07/21/rep ... vvdi-prog/ okay so we have vcc, gnd, sdc and scl connections on the board it looks like

so would be as simple as pushing "clean" firmware on the battery. dump the eeprom from a "good" battery, upload to a "bad" one?

Yes, or clear "crash" data at your dump. I can help you with that.

Well as soon as my posts will be shown I think, I’ll add addresses in eeprom which should be zeroed to removeverror

Ah sorry, your first post looked so generic that I thought it was AI generated. We've got a spambot problem here, they post something remotely meaningful and later add links to casinos or something

Are people finding these for cheap somewhere? I looked on some wreckers sites and the few I found were pretty expensive - perhaps more than a new off the shelf LiFe battery?

Zieg wrote: ↑Mon Jan 13, 2025 7:14 pm Are people finding these for cheap somewhere? I looked on some wreckers sites and the few I found were pretty expensive - perhaps more than a new off the shelf LiFe battery?

yes. online wrecks have no idea what they are posting. ones that are in good working order i have found for 150 $ online, but ones that need eeprom cleand from wreckers for cheap

dimonlipko wrote: ↑Sun Jan 12, 2025 8:19 pm Yes, or clear "crash" data at your dump. I can help you with that.

if we can do this with generic tools that would be great

MrZe wrote: ↑Mon Jan 13, 2025 2:09 pm Well as soon as my posts will be shown I think, I’ll add addresses in eeprom which should be zeroed to removeverror

awesome, please do.

Picked up the 16v battery cable harness from a model y, the connector is a jonhon EVL1-P2TJ-25WA
tesla service docs pleasantly have the parts numbers for the pins:
https://service.tesla.com/docs/ModelX/E ... ctor/x001/

housing:
EVL1-P2TJ-25WA

current pins:
85562302-01

signal/ linbus pin:
857111838-01


currently no lucky on sourcing them, EVL1 is not even a product line publicly shown for jonhon

You need external programmer, I use VVDI2. And after you read dump, I can help you clear it.

I’ve used multilink fix. Will find my writeups soon

dimonlipko wrote: ↑Tue Jan 14, 2025 9:08 pm You need external programmer, I use VVDI2. And after you read dump, I can help you clear it.

yes that is clear, seeing if there is more generic options available. i should have my hands on a battery soon.

MrZe wrote: ↑Tue Jan 14, 2025 9:56 pm I’ve used multilink fix. Will find my writeups soon

cheers, right being a S12Z processor, did you use the following?:

https://www.nxp.com/products/processors ... :PROGS12ZZ

Yep. Used that... I don't have my writing still, but as i remember errors could be reset by FF'ing eeprom addresses from zero to 320, but of course you should precharge it's elements...

Very special eBay deal later….i have 4 sitting on my desk. 3 read 12v or more, and 1 below 2v.

they had already opened the dead one. Looks like they burnt a trace and possibly a diod? Unpluged the bms, and trickle charged it at 0.1amp in till 9v then slowly increase current to no more than 2 amps. it been sitting for days and holding at 16v. each cell in is perfectly in balance.

I DO NOT RECOMMEND THIS, anything below 9v should be considered worthless. The cells are potentially damaged, it is not worth the risk.

But for testing and reverse engineering this should be "okay"

When the bms is plugged back in, it dosnt take long befor it starts to self drain through the bleed resistors, so this should be in a lockout, safety state.

Prefect we have some diversity in our eeprom memory notice all the residual flux on the cell tabs? thats how them come from factory. all of them are like that

the nxp u-multilink debug probe is around 450$cad. pricey.

UBSDM flash programmers:
https://usbdm.sourceforge.io/USBDM_V4.1 ... index.html

started a wiki page: https://openinverter.org/wiki/Tesla_16v_li-ion_battery

Would these be reasonable in "dumb" 12V systems such as you might find in a classic car without having to mess with communication or special alternator/regulators?

Tremelune wrote: ↑Wed Jan 22, 2025 6:03 am Would these be reasonable in "dumb" 12V systems such as you might find in a classic car without having to mess with communication or special alternator/regulators?

For a starter battery in an ice vehicle? No definitely not.

I am interested in them because:
-lightweight
-sealed connectors
-bms with coms on SOC and voltage
-safety shutdown

Bratitude wrote: ↑Wed Jan 22, 2025 6:11 am I am interested in them because:
-lightweight
-sealed connectors
-bms with coms on SOC and voltage
-safety shutdown

That single green comms wire is LIN? Does it only send soc and voltage data over it or does it needs some commands as well to balance etc? As far as I have seen tesla keeps those cells 80% charged so at ~3.88v. So with another dcdc converter for 12v lead batteries that charge to 13.6 -14v cell voltages should be lower so even less capacity? Thinking to use one but what would happen if the car is left for a while and parasitic drain would discharge it. Would it prevent itself from discharging too low and would it be possible to then simply recharge it or would it get locked?

m.art.y wrote: ↑Wed Jan 22, 2025 9:20 am So with another dcdc converter for 12v lead batteries that charge to 13.6 -14v cell voltages should be lower so even less capacity?

The wiki says that internal it has 16V. So it has a bidirectional DCDC converter. I assume it will not care whether it sees 13V, 14V or 15V externally, and just charge according its internal needs.

Edit: or, wait, do they have NO DCDC? Do they just connect the 4S directly to the cars 12V line?

Bratitude wrote: ↑Wed Jan 22, 2025 6:11 am For a starter battery in an ice vehicle? No definitely not.

I am interested in them because:
-lightweight
-sealed connectors
-bms with coms on SOC and voltage
-safety shutdown

But what about an EV converted classic car...?