Peugeot E-Boxer NMC cells reuse modules.

[andyoffgrid]

Hi, just joined you all. I live off grid in the UK (25 yrs with a 24KWh FLA set up using Outback inverter charger. 6.5Kw of PV and 3 Kw wind turbine. 24V battery) Also 10KVA gen set on Auto start.

I wish to double my battery capacity and upgrade to a 48 V system.

I have just purchased the complete battery and case from a 2022 plate Boxer-E with just 1000 miles on the clock.

I have dismantled the case and disconnected the modules. So I have 18 modules @ 44V nominal, Each module is 2 packs of 6s2p (in series) so effectivley 12s2p. Module energy 2.74KWh.

I wish for help in using these modules in the simplest and most reliable way. I am a long time user of Outback equipment, for it’s rugged build and reliability and simplicity and ideally would use their kit as I am familiar with their programming and menus etc.

The spec of the Outback VFXR3048E allows for DC charging on a scale from 42 to 68VDC. To support a larger load in the house I would have 2 inverters in parallel which also enables battery charging at a higher rate from a gen set (2X 45ADC).

I could have 9 packs at 44V in parallel. If I have a low volt point for recharge set at 43.2 (3.6V per cell) (30%SOC) and top V for charging at 49.2 (4.1V per cell) then that fits in with the Outback VDC capability.

So my question is can I manage the BMS side of the project using the onboard BMS circuit board, I note cell terminals are welded and the BMS board is an interference fit on a set op pins on the module cells.

Any thoughts please.

I have piccies but have not worked out how to upload them yet.

[Jacobsmess]

Something isn't quite right. If you have 18 modules at 44v each then you'd have a 792V pack if they were in series which is likely.
Please can you upload the pics and also dimensions of the modules?
You can do so by going to the "Full editor" and then scroll to the bottom to upload photos and other media.
Regarding the other stuff I'm not able to help but hopefully others can.

[andyoffgrid]

Hello Jacob. Thanks for pointing out my error, caught by terminology. And thanks, I've found the upload as well. Each module is 6s2p 22V nominal. Each module is built paired up in its water cooling pack, so 44VDC, so 9 x pairs of modules is 400V as built.

So what I'm looking for is the simplest way to build a nominal 48V battery and can I re use the integral BMS board, and if so what else would be required, please?

Below is spec for the complete battery
51 KWh.

Pack voltage 400V

Capacity 127.5Ah.

Total cells 216

Battery pack 108s2p

Modules 18

Module energy 2.74kWh

Module nominal 22V

Cells in module 12

Module configuration 6s2p

Format Prismatic.

Cell chemistry Lithium ion NMC

Cell nominal Voltage 3.7V

Standard charge and discharge cut off V 4.2V/2.75.

Cell capacity 62Ah

Battery voltage before dismantling was 388VDC and individual cells measure 3.59

[yasko]

Probably, those are the same modules used in Peugeot e-208 and other cars from the Stellantis group.

The on-board BMS is described here: viewtopic.php?t=2181.
It's possible to reuse it, but it requires some specific hardware for communication (PCB with MAX17841B).

I'm unfamiliar with the Outback inverters and their communication protocol for Li-Ion batteries. I hope it is publicly available.
I and other people on the forum use these modules in complete stack configuration (400VDC), but this requires an HV inverter.

[andyoffgrid]

So I notice that the 9 modules are all connected daisy chain in series all the way back to the main control module. The daisy chain loom consists of only 4 wires . From this I deduce that the individual modules board must be doing the actual "work" ie balancing cells, etc and the 4 wires are (perhaps) supply power to the individual module boards (sub 9V) red and green/ and the 2 white wires are carrying signals either to or from the module and main control "box".

So simple (stupid) question. If reconnect the daisy chain back together on all 9 modules and reconnect to the main battery control module, will the cells be checked and balanced if required? Power to this main BMS is 12 Vdc from the main car battery?

I only need the simplest control as i would plan to use voltage based signal to trigger ac charging (gen set) and high bat voltage to signal off to the gen set. I don't see that the inverter needs to comm with the bms. I don't need a SOC reading as in an off grid set up the battery is never at a static point. There is always a load, the house; and there can be DC charging 24/7 from solar and wind. And if battery is low the genset just starts automatically and stays on until it gets an off signal. Charging rate/voltage is set in the Inverter/charger.

[yasko]

I guess that the board over the cell looks like this:

There are four wires between each module, but they are only communication (RX and TX pairs). The cells power the balance PCB.
It's possible to reconnect all modules (18) to the central BMS and to get cell data. Some people try to do this here:
viewtopic.php?t=2675

I tried some time ago and even connected an OBD adapter to the CAN bus, and it's possible to get a lot of data. But I don't know whether the central BMS will balance the cell in this setup.
Indeed, the cells inside are very well-balanced, and you can run a battery without balancing for a long time.
Anyway, you will need some cell monitoring at a minimum, and I think the easiest way is to use an OBD scanner. I can give some directions on how to connect it.
Other solutions that exist require hardware and programming.

[tom91]

Please get details of the chip, better picture and markings on it.

This looks like some ISO-SPI comms.

[andyoffgrid]

Hi Yasko, hi Tom. Thank you for your replies. Yes the board over the cells appears to be same/similar to the illistration you provide. See mine below. I have the complete battery pack with all modules and the wiring loom. The daisy chain terminates at what I assume to be the main battery control module. I have removed the cover. The main chip conveniently is marked with black text on black so I can't get a picture, no contrast. I have copied the marks as follows.

INFINEON. TRICORE
SAKTC 275TP/64F200 (WCA I think)
hard to read last line Infineon 14 84 62018.

Yasko, you say that Indeed, the cells inside are very well-balanced, and you can run a battery without balancing for a long time. I have read other peoples posts with similar view. Do you mean that cell balancing is not a required function?

To recap, what I wish to achieve is the most simple of basic functions to maintain the cells in good working order for the rest of their lives. I would envisage monitering high and low battery voltage through the inverter charger with it's auto command to start and stop the genset as dictated.
Not concerned about SOC as in off grid there never is a static SOC. There is always draw on the battery (house load) and most of the time charging from RE sources.

Over temp on cells. I read that this battery could accept 80% charge in 1 hr. I would imagine that might demand a cooling system. The max charge rate I am likely to achieve would be 7Kwh, which might not cause over heating. I could very easily connect up the water cooling jackets on the modules with a temp sensor to run a circulation pump on demand. Or run it as a heater to keep cells warm in mid winter?

Any help or guidance would be much appreciated. I am old and this is new to me. Being old does mean it takes longer to learn new stuff.

[andyoffgrid]

[tom91]

We want to IC markings on the module boards not the main controller.

[uhi22]

These ones:

image.png (231.04 KiB) Viewed 11064 times

[andyoffgrid]

I'll try again but by eye, these 3 items appear to have no markings visible. They are coated with a varnish like treatment which renders them opaque. I will try to find some magnification device and come back with any info obtained. Thank you for your interest. Andy

[andyoffgrid]

I realise that "looking the same" is not very scientific, but the thread here appears to be covering the same battery modules set up. viewtopic.php?p=35581#p35581

For the sake of discussion. IF the modules I have are "the same" what will be my 1st step to repurpose the battery, albeit re connected as a LV battery not 400Vdc.

[EV_Builder]

This BMS is available. You can configure them the way you like.
(But practical is obviously per 6S).

[andyoffgrid]

I plan to put them in 2s8P to give 44v and use an Outback inverter charger. The Outbacks nominal 48V settings have a wide enough range to accommodate the lower nominal battery voltage. I will have 2 inverters in parallel to give larger supply and charging capacity.

Sorry, which BMS system was that.

[andyoffgrid]

Sorry missed your link at first glance. Found it now; that looks really useful, great.

[andyoffgrid]

I wondered if any one who is reuseing these modules has a view on the integral cooling system and whether in a static off grid set up the active cooling is needed. I would not expect to be charging or discharging at a greater rate than 7Kwh .

It would be fairly simple to plumb the modules and connect to a water pump and radiator to dissipate heat, triggered by a signal from the onboard PCB. Is this overkill, are the cells likely to overheat in the aboce scenario?

The battery will be located in a purpose made cabinet in a large dry ventilated building. Heating can be provided in the cabinet if the ambient temp drops to freezing.

[yasko]

Based on my observations, liquid cooling is not needed for this application. This battery can handle 10kW without problems, and the temperature increases by a few degrees. Here, I talk about 400V configuration, but with proper reconfiguration, this will be the same for an LV application.

Regarding your previous question about cell balancing, a friend has been running a similar battery (Pegeout e208) without balancing for over a year, and the cell difference stays around 20mV. The battery was balanced manually before connection to the inverter.

Anyway, I recommend a suitable controller for battery monitoring. Let's not forget that these cells are Li-Ion, not Lead-acid, and require more attention.

[andyoffgrid]

Thanks for your input Yasko. I was of the opinion that the water cooling would not be required with my relativley low level charge/discharge, nice to have a second opinion.

I plan to be in touch with http://www.wdrautomatisering.nl/ as suggested by "EV Builder" for the BMS controller. My son is a senior software engineer and I've just discovered, conversant with both UART and SDI protocols so I've got him lined up to help with this project. Saves any more wear and tear on my old (70+ ) brain.

Thank you.

[andyoffgrid]

So I thought an update on this project may be called for.

I'm pleased to report that I have commissioned my new off grid power system reusing the battery from a 2022 build Peugot 208e. The battery breaks down to 18 modules, each consisting of 12 NCM cells 2p6s. Nominal voltage is 22V and capacity of each module is 2.77Kwh. In the car, modules are built together in pairs. Each module has an on board managment chip which operates on a cell level, adjusting the charge rate balancing etc, each module also has a communication data 4 wire daisy chain, linking all 18 modules back to the main BMS. Alarm and temp data can also ultimatley turn off the main battery contactors. Because the on car bms is configured for 400V battery I had to source an alternative. I found a small manufacturer in Holland and after a couple of hiccups have the BMS in place and the parameters set for my requirments. Thank you Walter for your patience.


So I have built my battery by keeping modules in pairs giving a nominal voltage off 44V, theoretical max V 50.4 and min 39V. I have made sets of 3 in parallel to a bus bar which then links to the main 1000amp bus bar. I discovered that these battery voltage set points can be used with the Victron multiplus.II so I have a new 10000/140 unit on my power house wall alongside the battery bank.

[andyoffgrid]

]I removed the pairs of modules from the water cooling system. I then rebuilt the pairs. I used a piece of 3mm steel plate, placed between each pair of modules then used the plastic material supplied on each side of the plate. then I bolted the pairs of modules back together using 6mm stainless steel threaded rod with captive nuts and tightned the pair of modules together. Make sense?

The dimension of my pairs of modules as reassembled is 390mm long X 225mm wide X 153 mm high. I have 3 pairs of modules on each shelf. Although the shelving is chipboard each set of 3 pairs is in fact sitting on a section of corrugated steel so not in direct contact with the wood. The steel plate can dissipate heat therefore.

I have a 250amp main fuse in the system and never discharge more than 5Kw, normally less than 1Kw. I charge from the genny at a max of 5Kw. I have now got around 8Kw of PV panels, all on controllers and the wind turbine is capped at 3 Kw so I am not stressing the system. I discharge to 41V and charge up to 48V.

My battery as built, is inside a heated insulated wooden cabinet. There is a 60w electric tube heater at the bottom (on a stat set for 5@C. The generator is just outside the power house and I get some heat come through when the genny is running.

I managed to re use the original data daisy chain but due to the modules being spread out I had to make up some extension pieces. Watch out if you do this as it is easy to get the wires reversed. I have the wiring diagram on file.

I re used the contactors from the car battery (there are 4) and have one on the main battery positive and one on the negative so if the contactor trips the battery is totally isolated. Note I have not used any crocodile clips.