Buy FOCBOX Unity

FlexiBMS Lite - New approach to get past Vaporware stage



Anything else is blind faith.

I know we’re getting off-topic so I’ll try to shut up :slight_smile:


This is great. For my money I want small, charge only, canbus for metr connection and user configurable charge termination/ balance. Balancing is a must…

No faith needed with can and metr.


Did you mean charge only??


I did…



I don’t know anybody that wants to take their pack apart and charge everything manually. It pretty much defeats the purpose of having a BMS???


@amazingdave @deucesdown @akhlut
Talking about Metr

Is the Canbus meant to connect the BMS to the VESC/derivative and then from there with USART to Metr and then over bluetooth for status monitoring on your phone?


Metr Pro issues command to DieBieMS through CANbus (Bluetooth to Metr Pro -> UART to VESC -> CAN to DBMS).
If you design commands the way as this is coded into DBMS, it would be compativle straight forward with Metr Pro (I assume).


The DieBieMS for one. It’s great, but spendy, large and not readily available.

there are a crop of smart BMS from China. I’ve seen a few that have cell monitoring but can’t remember the names.


Nice to hear back from you @SimosMCmuffin

  • One more vote for keeping CAN if integration with Metr is to happen, also since it’s charge only, you can use software to cut power to the motors instead of cutting power to all electronics

  • Current sensing module I think it’s not needed, VESC already do it, and with the new firmware it’s even more integrated when using multiple VESC’s, the downside is that charging current is not monitored, if you want the BMS to now discharge current and energy used is just a matter of requesting the VESC data via CAN

  • about the full blocking, how does it work? is there any way to make the charge port live when the board is turned on? or at least make it configurable by a jumper or something? I see a lot of uses for using the board as a portable battery, such as powering a computer and etc

  • size, most compact charge low current BMS’s are 60x30mm, if possible this would be the perfect size

And put me up from one, I have space reserved for it since your last iteration


999 out of 1000 times you just bulk charge as normal, but to 90% not 100%. The telemetry would let you know this is safe. “know” is the key concept. Just because you have BMS installed, do you think your pack is perfectly healthy? Does the BMS let you know if something is amiss? Do you just assume it is?


If I can see the p groups in metr I’m going to be happy for a month between inspections…


The INA180 is used to monitor charging current. Protection from overcurrent and allows setting cut-off current at the end of charge.

This is a fully blocking switch setup. in this case it’s 2 N-channel mosfets either in common-drain or common-source configuration.

Here’s an equivalent circuit with switches with a “body diode” instead of mosfets, but take one away and your switch can only stop the flow of current in one direction.

It can be opened up through firmware, but how to communicate that to the BMS when to do it, is the real question.


Ooh, just found an interesting IC for power saving use:

Gonna dig through this datasheet…


Understood, if you were integrating a switch it should be easy, something like hold the button for 10 seconds

And about the current drain, give the drama from the past week about switches drawing to much current when idle, in my opinion in no way a switch or BMS should be able to to kill a board if left connected in a reasonable amount of time

Isn’t any way to make the BMS only active when charging and when the board is on? Such as connecting it after the main switch and using a transistor to turn it on?

Lets say for example a 300 Wh battery left with 20% charge should no be dead after 6 months


IMO the solution is to maximize sleep states when not charging to minimize quiescent draw. On the other hand you need to have something active to check every now and then if any comms or if a charger is detected or enable signal from ESC is detected.

The first BMS’ problem to the quiescent current were all the resistive voltage dividers, which pulled multiple hundred microAmps from the battery regardless if the MCU was even soldered onto the board. That’s why I’m a bit hesitant to change the charging mosfets away from the fully blocking setup. IMO better idea to allow for the roll-to-start function to signal that system is active and then open the charging switches so, if somebody wants voltage from the battery on the charger plug then it can be activated with the roll-to-start. (EDIT: Or just ESC normally powered up)

I personally would prefer to not put any buttons on the BMS, because some people essentially integrate the BMS into their pack by wrapping the whole thing up.

ST’s L-series MCUs can go into reaaaaaally low-power states (attached some datasheet tables from current consumption from normal run states to the deep low-power states)


The micro controller yes, but what about the voltage regulator that powers it? genuinely asking


Thats not a battery management system. Its a battery monitor.


When a battery is connected to the BMS the LTC6803 is powered but stays on standby mode until SPI communication happens. The IC has a linear 5V voltage regulator output. So if it’s loaded 2 mA, it should pull 2 mA from the battery. This is then passed through 3V3 linear regulator to power the MCU.

EDIT: So if the MCU is in low-power state with few ten microAmps of current use, that’s how much should be pulled from the battery + LTC6803’s standby current.




(LTC6803-3 datasheet)


Awesome bms! Would definitely buy one once available. However, I would really want to see them with a better connector. Something graded for automotive industry or similar.


Do you actually even want roll to start on a charge only BMS though?