No documentation on the MQTT stuff - however if you are interested in what fields are output, you can probably glean that from the code.
You may want to try MQTT-Explorer and connect it to your broker, you should see all fields that DIYBMS sends to itâŠ
1 Like
Hello, i have mqtt fully working in diybms with home assistant. I can provide you with help if you need.
1 Like
Thank you both @SergioF @StefanK for your assistance.
I ended up taking Stuartâs advice to read through the code and eventually figured it all out.
Thanks though for replying. Much appreciated!
just getting around to playing with 64 cells 4 banks of 16,
soon to have my v3 powerwall operational
using v4.4 modules and v4.2 controller
3 Likes
Hi @stuart , Saw your interesting video on the Power Basket build. Are you ultimately targeting to get this purchasable from somewhere ? Donât know of any JLCPCB equivalent who make cases 
âŠ
Just FYI , I purchased the SEPLOS MASON case WITHOUT the BMS and use the DIYBMS instead . Front panel that came with SEPLOS case is not fully usable in this situation .
I also got in touch with an Alibaba store that does cases for the LF280K with customisable front panel - Currently they offer versions for JKBMS etc⊠If you like I can share the details of the seller and that might make it possible for people to order cases for DIYBMS.
CheersâŠ
Hi, thatâs an interesting idea. I wasnât planning to make the case public, as itâs more for my specific needs.
Hi there
I have completed a V4.2 board (9.5.21) and deployed the ESP32 test code on it.
The green LED is flasing, but there is no monitor.
This is the same result I got yesterday, before I switched the TCA6408 to a new one. The logs show diynms: TCA6408 error. I got the chips (TCA6408APWR) from LCSC, so should be originals.
> Leaving...
> Hard Resetting...
> Hard resetting via RTS pin...
> Done! Flashing is complete!
> Showing logs:
> [08:40:05][[08:40:05]I (13) diybms: Configure I2C
> [08:40:05]I (14) diybms: Scanning i2c bus
> [08:40:05]I (20) diybms: Found i2c device at address 0x38
> [08:40:05]I (28) diybms: TCA6416A not fitted, assume v4.2 board
> [08:40:05]I (29) diybms: Found TCA9534A
> [08:40:05]E (29) diybms: TCA6408 Error
Anybody has an idea what could possibly be faulty?
what is this on your cells? Where did you get it from?
I was just checking jlcpcb to see what parts are in stock. I would need a few more cell boardsâŠ
Unfortunately I already canât get past the first step, I get the following message:
âStandard PCBA service requires each side of the PCB should be larger than 70mm.
Your PCB size is 32mm152.84mm, do you need JLCPCB to help you add edge rails and enlarge
the size to 70mm152.84mm? Or you can also enlarge the PCB yourself then upload again.â
I havenât seen this before, is this new or did I set something wrong?
Choose âEconomic PCBAâ instead if standard.
1 Like
Yes that seems to be the problem, I wanted to make a collective order with 75 pieces.
I can then order a maximum of 50 pieces?
Yes. Itâs limited to 50. But you can do 2 orders simultaneously.
50 and 25. Will be delivered in one boxâŠ
1 Like
Good idea with the 50 + 25 
The minimal cell voltage is 2.9V, but it is better to stop discharging at 3.3V for life of cells
@Nico_F IMHO that depends on the cell. If you have batteries that are rated for 6000 charge cycles, i.e. ten to 20 years when used with solar, it seems rather wasteful to not use almost half of its capacity. (Assuming that the battery will die of old age after 15âŠ20 years anyway, no matter how often you discharge it.)
Stopping at 3.2V seems more reasonable to me.
There is no evidence for dying if cells after 15-20 years.
In my eyes this is all bullshit.
Cells are told to loose capacity.
They tell you they are dead after xxx years because they reach 80% of capacity.
For me at 80% the cell is not unusable.
In aviation Lange uses Li-Ion cells since 20 years.
Now the first cells have to be replaced
Why? Because they canât explain the authoritys why these cells still work.
In industriell applications these cells are replaced after 15 years because if the capacity.
Lange says: the cells in the aircraft are still OK.
Because they are treated well.
But now they have to be changed because the approval in aviation is only for 20 years.
Furthermore the capacity by weight is rising since the 70s.
Ist a straight line up.
Lange could climb his aircraft with the cell from Anno 2000 to 3000m.
With the same weight he can climb the aircraft today to 4800m.
In 10 years itâs expected to reach 6140m.
So in 30 years more than double the energy density.
And he uses standard industry cells. No special development.
Fast loading stresses the cell. Deep deloading stresses the cell.
Threat your cells well and you enjoy them for a long time.
I charge them in summer only by 12A. Now in winter with 32A.
Thatâs max of my Inverter. The days are shorter, the nights are longer.
So I have less time to charge the cell and need more capacity for the night.
I deload only to 20% SOC. To avoid deep deloading.
The same with panels. They Garantie 87% performance after 20 years.
And they tell you this is the limit and the panels have to be changed.
Are they mad?
1 Like
i cut the power if one cell go to 2,7V this is the bms setting on my lifepo4
i dont care for more this does the bms for me
charge voltage is 3,5v this is more than enough, no need go for 3,65v because 3,5 is same full charge voltage.
i dont use my battery regularly to 2,7v its just the cut voltage if something goes wrong.
my battery is discharging almost regularly to 60-70% soc and only a few days in the winter to 5%
the diy bms works in a mixed system with victron and sma sunny island over canbus it working very well.
but one minor problem with the displayed warnings in my inverter if one cell starts balancing.
and the slow response time of 3sec is a minor problem for me
hi @stuart
is possible use this current sensor with your board, a LEM CAB500/300, working in the can line communication âŠ
1 Like