TeenyBMS, a small form factor 16S smart BMS

Esk8 Electronics DIY Solutions
, ,
#1

Update 13.2.2023

I’m moving from the 0.1 HW version to 0.2 with big updates. I have decided after tinkering with the 0.1 prototype board, testing it and more importantly looking at other possible IC solutions to make the design have a better feature set, same price and same footprint. Well I found an affordable battery stack monitor IC from Texas Instruments, which supports up to 16S configurations and I can cut my total component count to less than half, fit all components on one side of the pcb (cheaper/faster pcb assembly). Now this does mean that the BMS now needs a microcontroller, but as it doesn’t have any crazy connectivity, it can be a pretty basic one, but this then enables a lot of smart features for it.

So the originally simple and dumb BMS is now becoming smart!

Update 10.12.2022:
Thread got started with some initial general project ideas, but it has now converged more towards a small form factor dumb BMS, as the market already has decent and affordable dumb BMS’ from Daly for example, so folks threw around the idea to make it a really small form factor to make it distinct from the competition.

early PCB mock-up at 18x65mm size:

image
image737×438 66.3 KB

image
image745×401 67.3 KB

So currently the project parameters are:

  • Same PCB size as a 18650 cell (18x65mm)
  • 3S-12S cell support (11.1V-44.4V)
  • 8A-12A charging current
  • 5A discharge path control for accessories (undervoltage protection, so they don’t drain the battery empty)
  • SPST switch control for discharge control, to turn accesories on/off
  • Overcurrent protection
  • NTC-temperature probe support for over temperature protection

Are you interested in the project?

  • Yes, I’m interested
  • No, I’m not interested

0 voters

OLD OP
I have done some analysis of the cheap chinese BMS’ and have figured out how they are hardware wise implemented, so I’ve been wondering about making my own “dumb” (no programmable logic) BMS design based on the chinese “dumb” ones, but made with actually known components and I’m wondering would there happen to be any interest if I were to sell some as DIY kits with PCB and components included? I’m gonna designing a “dumb” BMS platform for my own use anyway for use in integrating battery packs into projects, so that’s why I’m asking.

The design can be scaled pack S-count wise between 3S-18S (fully or partially populate components), but same PCB. Both charge and discharge path control can be implemented (over-charge and -discharge protection). Adding a thermal probe is possible. Overcurrent protection is also possible. Should be pretty affordable especially if I only need to supply PCBs and can volume buy the needed components for the kits and there is no firmware development needed.

Basically the minimum BMS implementation with no extras, but enough to protect from overcharging, overdischarging, overcurrent, balancing and over temperature (under temperature also implementable).

Chinese “dumb” BMS for reference. We can improve on the connection placement though.

image
image1000×695 84.8 KB

  • Yes, I would be interested in a DIY kit
  • No, I’m interested in a “smart” BMS
  • No, I’ll buy a chinese one instead

0 voters

#2

I dunno man, DALY has a pretty good hold on the market with the 15a water resistant ones, for a mere $15 a pop and $10 shipping for a bunch it’s hardly worth fucking around

9 Likes
#3

If this bms could be 12S and skinnier than an 18650 so it took up a tiny area that could still be worth making.

4 Likes
#4

So width wise <=18mm? Connector is gonna be tricky but otherwise I think the balance circuits could fit in the space.

Would you have any opinion for the length?

2 Likes
#5

Or at least some multiple of 18mm, if you can get a maybe 5a charge and 5mA balance (no discharge) to fit in the space of 2x 18650 that would be really nice for smaller builds

Ideally all wires come out the same end so it would go in the last row of the battery, a 1.5mm pitch 13 pin connector for balance and some pads to solder B- and C- for 18awg leads

7 Likes
#6

So this is both charge and discharge? I think if you made a small charge only you’d have a better chance selling it IMO.

3 Likes
#7

Well the reason why I started this thread was to have this discussion about what people would want and not want from a “dumb” BMS, but you’re correct, most likely if it’s going to be size constrained, then you wouldn’t be able to fit enough FET for a reasonable discharge current, so it would be easier to keep it charge-only for that reason.

3 Likes
#8

I mean 90% of us set up our BMS as charge only so, not sure how useful providing discharge is, unless it’s for accessories

For example, my lights are connected to the discharge on my bms

3 Likes
#9

Would be possible to make a couple Amp discharge current path control with a small FET for this exact purpose, so accessories don’t kill/drain the pack accidentally?

1 Like
#10

Could be an excellent addition to the features if we’re going from the ground up

If you could have a connector for a SPST switch that attaches to it, would be even cooler

1 Like
#11

That’s why I started this discussion. It’s gonna be a custom jobby, but I’m interested in hearing the use cases from the user base.

3 Likes
#12

I think being able to control an accessory with a switch would be pretty nifty. I use the LLT bms app to do mine, but I’m developing a little switch board for my lights so they can take bare power input

If this was already in the bms, that would be pretty convenient.

#13

I think this should be easy to implement for a SPST off-on switch.

1 Like
#14

I’m very interested, but only if its very small. Dumb BMS often come with a discharge integrated, which makes them big (sort of) and hard to fit. I used just rudimentary balance boards for this exact reason, which are really less than ideal.

A small discharge path (lets say 5A) would be a nice bonus, but the small form factor would be the main point here. The smaller the better.

I also think for this reason you should maybe redo the main poll, as the current one does not indicate any features that makes you stand out.

5 Likes
#15

So I think that the size is gonna be the biggest thing that I can focus on for this BMS.

I will preliminarily christen this the “TeenyBMS”, at least I couldn’t find any hits on google with this, so I’d believe it’s not used yet.

3 Likes
#16

some method on reading out cell voltages would be very nice. im always wondering whether my packs are still good so i recently switched to those bluetooth bmses, however they are chonkers. The feature is super nice however and i wouldnt go without it anymore.

2 Likes
#17

At that point, it’s a “smart” BMS. Because you need actual logic to get that cell voltage data out and also quantifiably measure them.

And I’d say it’s better to approach the BMS design as either “dumb” or “smart” and not try to make a compromise between them.

1 Like
#18

+1 for making it as small as possible. If you need discharge then that means you also have a buck converter which means you have the space for a larger BMS anyway.

1 Like
#19

I can implement discharge path control without a buck converter though?

What is the buck converter needed for?

1 Like
#20

For lights! (And other accessories) 12V 3A or 5V 6A would be amazing :smiley:
Especially with a switch so that you could turn the whole circuit on/off separately.

2 Likes