me meither, I have maybe 4 different china chargers, never had any of these issues. They only “stop” if VBAT=end voltage
I’d imagine that they also have a CV, current taper at the end and stop after the charging current falls under some specific limit?
From what I know, E-fuses provide over current protection as basic and some more sexy complex functionality the more expensive they get. Think default for the FET is “open” unless theres a fault.
Chargers should see the voltage if so
well, yes and no. The current taper is a natural reaction to voltage rising on the pack, the potnetial between charger and BAT gets smaller → less current flows. So the charger really isnt doin jack shit.
Its just dimensioned so it works well
Same behaviour if you use a DC supply in CV mode and charge a pack.
1 Like
These mysterious china chargers, do they output voltage at all if no battery is seen? Cause if theres voltage, you can open a FET 
Pack charger’s don’t usually have sense lines for the voltage though, so the charger only sees the voltage at it’s output, not what is at the pack, so with:
So the charger would stop if it’s hits it’s floating output voltage at the end of the CC charging, even if there is still decent current flowing at this point, skipping the CV stage completely, so I’d say your description of the charging behavior is incomplete, but this is just splitting hairs at this point arguing over semantics.
I checked Digi-key for their electronic fuses with relevant filters and they don’t have anything to fit this specific application. only single available IC went to 56V in it’s voltage range and it was only for 800mA.
But I’d say at this point, that I’m not gonna rack my head over this longer, as the end conclusion kinda was:
It can still be added later, but feels more like wasting time currently.
1 Like
yeah of course, I worded it poorly. It monitors the current, when I = XmA the charger stop, which usually is charger V= Vbat
1 Like
Could this be adapted for 15s and onewheel pint form factor, I think this would be great for a whole bunch of people who run without a bms. For me personally, this would fit the bill exactly in a 15s variant.
1 Like
Yeah I would like to see a higher maybe up to 24S BMS with this skinny form factor.
This is how big the daly 24S is.
2 Likes
I have been getting the schematic side locked in for the first version and been working on the layout side. The 18650 sized pcb is possible.
I’ll have the first proof-of-concept board ordered next week.
7 Likes
I don’t know if mentioned already or too late but having an led that goes on whenever a cell hits 4.2 and is being balanced is nice for knowing what’s going on.
I use just discharge balancers with lifepo4 since seems safe enough; what other features are worth having with more dangerous chemistries? I had these 24s lifepo4 balancers custom made to fit a small space. The red one is the commonly available discharge balancer.
The thru-holes take less space
3 Likes
Got the first boards for the first prototype. Note, component and connector placement is not finalized yet, this is a proof-of-concept board, aka does everything work as intended.
I’ll get the parts ordered.
16 Likes
Looks awesome! What is the target price for these babies?
What does it do specifically?
Can’t promise any particular price point yet, as the hardware and design is not verified to work, but the major components for a batch of 100 units (digi-key sourcing), would be roughly in the range of 16-19 € for each unit.
Considering that there isn’t any software work needed, once the hardware design works as intended, I’d throw out a prelimiary price of 35-50€ VAT not included.
The design is relatively flexible, meaning that this 12S unit should work with 3S-12S packs and the design can be scaled up to higher S-count packs as well (more balance circuits and higher voltage chg and dchg path control and regulators), although the PCB size will increase.
4 Likes
Planned features:
- supported cell count 3S-12S (currently planned for Li-Ion 3.0V - 4.2V range)
- cell level balancing
- cell level under- and overvoltage protection
- charging current of 10A (at least)
- overcurrent protection
- under- and overtemperature protection (via external NTC probe)
- charging cut-off once any cell triggers over voltage
- discharge current 2A (at least)
- discharge control ON/OFF via external SPST switch for accessories (lights, charger etc.)
- discharge path shut-off once any cell reaches under voltage
I wrote the above all in sentences first, but decided a list would be cleaner, but here’s the paragraph:
At moment I’m planning on implementing cell level balancing, over- and undervoltage protection. Then on charging path control, overcurrent protection, over- and undertemperature protection (external NTC probe), charging cut-off comes from the cell circuits. Then there is discharge path control ON/OFF via external SPST switch and undervoltage cut-off to shut off the discharge path if any of the cells get too low.
8 Likes
This is probably going to sound all backward and messed up, but bear with me.
Could I turn on the BMS discharge with the 3V/5V output of my VESC via a MOSFET/relay? And if I had a buck converter downstream would I blow the MOSFETs on the BMS with inrush current? The use case being that I’d like to have lights turn on when I turn on my VESC. Would be nice not to have two switches on the board.
2 Likes
It would likely work by injecting the 5V directly from the VESC or via MOSFET and it will most certainly work with a relay in-between the VESC and BMS, because it would just replace the SPST ON/OFF switch contacts.
I can test this once I get the first prototype up and running.
3 Likes
Getting some testing done on the first board. Haven’t assembled it fully yet, as I have been testing it in it’s functional blocks and then adding new ones, as I have tested, modified and verified the earlier ones. So far nothing terribly wrong in the design although I had misread a comparator’s output behavior, which caused the logic output to behave inverted then what I had thought. At least with this specific linear regulator, the minimum battery voltage for it to work correctly is about 12V (in my testing with benchtop lab PSU), but I think 15V would be a bit safer, considering the datasheet recommends a minimum of 25V, but this affects the possibility of supported S-cell configurations, as to meet that minimum voltage requirement.
I have been thinking about the possibility of ditching some of the logic level ICs and plonking a cheap and small MCU on to the board to be honest. Would simplify the control circuits and allow for more control over the whole behavior, which would balloon out of control if it was to be implemented via logic ICs.
Something like this 20-pin low tier STM32, which in a QFN-package would be just 3x3mm in size. The 555-timer circuit on the bottom side (at bottom of the picture on the right) is 3.9x4.9mm. No connections outside, just programming pads for production. I’ll look into it a bit more. It might become just a “slightly dumb” BMS 
https://www.digikey.fi/en/products/detail/stmicroelectronics/STM32C011F6U6TR/17075954
12 Likes
Thanks for the update!
It’s always amazed me how many things have to be considered and dealt with even for what most people would think is a simple circuit.
2 Likes