sorry What i meant was that if you don´t have any fets on your main board, you definitely need to have a second board with fets for charging. Otherwise you´d have a small board with a 40A charge/discharge capacity, which is nice and small and enought for people to charge and bypass discharge. So that only people with higher loads need to get a second FET board.
If you just want to charge you donât need any FETs on the main board for that. So the main board alone could be used just for charging, if youâre using the boost charger (controllable end-voltage).
Bulk charger would present a problem, because that you canât disconnect if it for some reason was charging too fast. Then if you want more control add the Switch board and route the battery connections through it for more control.
I must say Iâm a bit intrigued by this concept nowâŚ
Iâm interested in hearing peoples thoughts on these two design/concept approaches.
All-in-one BMS module: Single board that has all the features on-board (as pictured in the first post). Battery connection is controlled. power button support. Supports both discharge and charge control and current measurement. Otherwise good approach, but non-ideal for people who want to just charge through the module under control and donât want to pay any more than necessary. Also non-ideal for high performance/current users. which the board canât support. Increasing current capability also increases price, which the charging only people are not interested in.
Split board design BMS is split into mandatory MAIN unit and optional SWITCH unit. Main unit supports battery stack monitoring, charging with itâs dedicated charging FET and decreases entry point price for controlled charging with flexible battery configs. SWITCH unit allows for current measurement, load disconnection, power button control. Allows for flexibility in discharge current capability with bigger SWITCH units and optional heatsinking.
Both boards support:
- Supports 4S-12S battery stacks
- Easier and simpler charging for different battery configs with only single charger ** 12 V charger can charge 4S-12S configurations ** 15 V, 5S-12S ** 18 V, 6S-12S ** 21 V, 7S-12S ** 24 V, 8S-12S and so on
- Still supports Bulk chargers with built-in CC-CV feature, if you want faster charging for specific series cell battery pack configuration. (for example, a 42V 2A charger for 10S pack)
- Adjustable limiting of charging power when using integrated boost charger
- Adjustable end-point voltages
- Balance current 100mA @ 4,20 Vcell
- Overcharge protection during charging (Load is able to push current to battery for example during regen braking even if cell voltages exceed limits)
- Bluetooth comms (HM-10) support for wireless diagnosis and monitoring of cell voltages and pack current during rides.
- Firmware flashing through USB via STMâs built-in bootloader
- BMS configuration via USB on a graphical UI
This is not a definitive poll, but just to gauge peoples interest in what kind of BMS they would rather have?
- All-In-One BMS
- Split-Design BMS
0 voters
I think there are really only a few people who do 40A+ cont. Current peaks yes, but only a few seconds.
Not even on hill climbsâŚ
I heard a lot from different people and also read that from here (in French, sorry) : http://www.ni-cd.net/accusphp/theorie/charge/liion.php
Of course, C rating for charging is meant for CC stage.
That makes sense. Canât wait for the result ! Feel free to have a look at my code (not sure the latest version is commited yet) :
Iâm finishing assembling a 12S1P bench battery with unhealthy and healthy cells (to observe a bit better balancing process). Iâll let you know if youâre interested in.
http://lithiumbatteryresearch.com/Plating.php
When the charging current is very high, as might occur during regenerative braking, the transport rate of Li+ ions to the graphite negative electrode exceeds the rate that Li+ can be inserted (intercalated) into the graphite. Under these conditions, Li+ may deposit as metallic Li, which can lead to a short circuit, degrading the batteryâs life and durability.
https://batteryworkshop.msfc.nasa.gov/presentations/1-Lithium_Plating_AZimmerman.pdf
Quick look into the matter would suggest that plating occurs when charging current is too high, not low. Can you find any other sources confirming the low current plating?
Very interesting links ! I read them carefully, and no mention of low curremt causing plating. Iâll try to find more documention (or maybe find some specialists at work).
I just googled with âlithium charge current platingâ Those were one of the first hits.
What I gather is that plating seems to incur at high charge rates and/or low temperature while charging.
http://www.upsbatterycenter.com/blog/lithium-plating/
During this process, which is called intercalation, is when lithium plating takes place. There are two main reasons for this. One is the high charge current forcing the lithium ions to move at a faster reaction rate and accumulate in the surface of the anode (usually made of carbon compounds). Another reason is charging at low temperature. In this environmental condition, the reaction rate slows down thus affecting the intercalation of lithium ions.
Saw that article as well. Iâm focusing on plating under low charge current, but nothing really clear and studied like the over-charging. But I still read during my search :
The advised charge rate of an Energy Cell is between 0.5C and 1C
Is it for another reason than lithium plating ? Which one ?
Still searching.
So I saw that it only supports up to 6A for charging - wonât that be a pretty big hindrance for braking situations? People commonly brake at between 12 and 20 amps on 12s packs, from what Iâve seen on the boards.
Also, since it doesnât look like weâll be getting this feature in our ESCs anytime soon, do you think itâd be possible to include a power shunt inside of a BMS for resistance heating circuit? So that we can brake at higher amps, regen charge up to a certain amperage ceiling, and then shunt the excess amperage off to a different circuit, that we can connect whatever we want to?
great design so far, and very impressive work!
Where did you see or where did you get this number? I donât think I have mentioned that anywhere.
Itâs in the âFeaturesâ section on his blog post:
âBoost converter circuit with adjustable voltage for CC/CV charging at up to 6Aâ
There are a lot of words in that post that I donât understand though, so maybe I misinterpreted this sentence and it can in fact support higher charge currents for braking events, and this statement only applies to wall chargingâŚ?
Ah, yes. Thanks for the context.
So the 6 Amps is the maximum input current for the boost converter, which you connect your charger into. As I mentioned in the original post, I have been testing the boost converter by charging my battery with a 16V replacement laptop power supply. The boost converter takes the 16V and then âboostsâ it up to a controllable voltage, which allows me to charge my 10S battery pack from 29V to 41.9V.
It doesnât affect the max current the motor controller can pump back into the battery during regen braking through the discharge FET.
Ahhh, OK, noted. Having a transformer on board is a really great feature, that would be fantastic. Thanks for clearing that up!
Ok, so Iâm gonna do a concept PCB design for the split-design and post updates as I get more progression to get more thoughts. I spent a little time last evening fleshing out the idea, looked for components to actually execute it and made a better block graph for it how it would actually work wiring wise.
Iâd appreciate more votes on the POLL though! (http://www.electric-skateboard.builders/t/poll-announcement-flexible-configuration-and-charging-bms/46117/23)
I also never saw that mentioned before, but saw a few times in forums that low charge current donât do bad to the battery
This is the same when people say that float charging damages the battery (Lithium), true float damages because it switches again to CC after the CV phase, but keeping the CC until it goes exponentially to zero doesnât have an influence on life
I think the best argument against the 0.5C recommendation is Tesla, take the S 100D, it comes with a 17.3kW charger, thatâs approximately 0.2C, or even less since you can set lower in car, and I bet they are the company that have the most of real world data about 18650 batteries
@SimosMCmuffin do you think your firmware would be compatible with @raphaelchang hardware? Quite a few people including myself have a 100 USD dead weight
If you managed something in the same size as the original, I had the exactly space reserved for it in my board thatâs is now occupied by an anti spark and cheap Chinese charge only BMS
Going to be tricky, Added to that is the fact that I have bigger components on the board due to the beefâd up boost converter (bigger inductor, double diodes and input/output el.caps). Okay, I donât have the discharge FETs or high current pathways, but I need the connector for the Add-on Switch-module.
At this point, I canât really promise anything on the form factor as I donât have layout even mocked up to see whatâs the smallest area or shape I can go. Plus I donât actually have practical test data on the new board too see if the layout works properly IRL.
Different microcontrollers. Battman has STM32F303CBT6, Iâm planning on using STM32F070CBT6 + different peripheral connections, at least pin out wise. Whatâs the problem with the original firmware? Does it now work at all, or does it work badly? The thing is open source, so you could code your own basic firmware for it to make it work. Iâm going to prioritize my own firmware first and once thatâs working, then maybe port it with a basic feature set to the Battman, so people can at least get use out of their hardware.
EDIT: Could you take measurements of the Battman board?
The actual firmware allows us to use Battman only as a basic Vedder powerswitch. I already code a lot of features now and Iâm glad that @Pedrodemio will join the coding process The size of Battman is pretty compact but if I could rework it, Iâd make it a bit bigger to allow more space between neg and pos battery and output leads (they are verryyyy close from each other )
Making a small BMS is a real challenge. DieBieMS isnât small either, for good reasons I guess.
Yah these wires tend to shorts! A cool BMS doesn´t have to be extremly small, i think the DieBieMs seems to have a great size. Others like my first China BMS are taller than my battery and are about 25cm long Smaller is of course nice to have, but it´s totally fine to have all the cool features and the size of a batterysupport bms.