Tough BMS
-
- Posts: 3265
- Joined: Mon Dec 24, 2018 1:37 pm
- Location: Slovenia
- Has thanked: 80 times
- Been thanked: 234 times
- Contact:
Tough BMS
I have used another distributed BMS for a long time now. Surprisingly i had absolutely NO issues with it.
It was designed to dissipate energy from cell at 4V per cell. Not 4.2V but 4V since this way i am assured of longer cell life and from 4V to 4.2V it is like 30minutes of charging. During that time even the dummest charger starts to go to CV.
I used TL431 reference diodes to trigger PNP/NPN pair. NPN transistor is actually NFET capable of 5A continuous drain. However i noticed even 1A in closed space can be a fire hazard. So i use 0.5A drain. It works in my car for 4 years now. No events
Also i designed a LV alert into each module. I used Microchip TC54 voltage trigger to signal EMPTY at 3V per cell. While there is some power still in the cell at 3V it can drop to 2.5V in under a minute! I use optocooupler PC817 to signal events for a single cell. Currently i have this alarm set so it lights red LED on my dash. But i could just as easily connect BMS pin of Rev 2 inverter to put the car in "turtle mode".
There is also provision to use HV alert when a single cell reaches 4V BMS loop can slow the charger to 1A so that all cells align.
I tried to eliminate PNP transistor, but i found out its base hysteresis keeps TL431 from turning on too fast. This was designed explicidely to work with LiPo cells. It is a rather crude system with low parts count, but it is also EMI proof and reliable. I made 8S boards because i use cells in packs of 8.
You are free to use this system and further develop it if you like.
It was designed to dissipate energy from cell at 4V per cell. Not 4.2V but 4V since this way i am assured of longer cell life and from 4V to 4.2V it is like 30minutes of charging. During that time even the dummest charger starts to go to CV.
I used TL431 reference diodes to trigger PNP/NPN pair. NPN transistor is actually NFET capable of 5A continuous drain. However i noticed even 1A in closed space can be a fire hazard. So i use 0.5A drain. It works in my car for 4 years now. No events
Also i designed a LV alert into each module. I used Microchip TC54 voltage trigger to signal EMPTY at 3V per cell. While there is some power still in the cell at 3V it can drop to 2.5V in under a minute! I use optocooupler PC817 to signal events for a single cell. Currently i have this alarm set so it lights red LED on my dash. But i could just as easily connect BMS pin of Rev 2 inverter to put the car in "turtle mode".
There is also provision to use HV alert when a single cell reaches 4V BMS loop can slow the charger to 1A so that all cells align.
I tried to eliminate PNP transistor, but i found out its base hysteresis keeps TL431 from turning on too fast. This was designed explicidely to work with LiPo cells. It is a rather crude system with low parts count, but it is also EMI proof and reliable. I made 8S boards because i use cells in packs of 8.
You are free to use this system and further develop it if you like.
- Attachments
-
- LiPo_8Cell analog.zip
- (38.52 KiB) Downloaded 142 times
-
- LiPo_1Cell analog.zip
- (685.61 KiB) Downloaded 112 times
-
- Posts: 260
- Joined: Sat Jan 12, 2019 12:39 am
- Location: UK
-
- Posts: 260
- Joined: Sat Jan 12, 2019 12:39 am
- Location: UK
Re: Tough BMS
Ah OK
Code: Select all
$ file LiPo_1Cell\ analog.pcb
LiPo_1Cell analog.pcb: Composite Document File V2 Document, Little Endian, Os: Windows, Version 5.1, Code page: 1252, Revision Number: 128, Total Editing Time: 1d+08:43:51, Last Saved Time/Date: Thu Aug 1 12:24:52 2019, Create Time/Date: Thu Feb 12 11:07:50 2015, Name of Creating Application: DesignSpark
-
- Posts: 3265
- Joined: Mon Dec 24, 2018 1:37 pm
- Location: Slovenia
- Has thanked: 80 times
- Been thanked: 234 times
- Contact:
- Jack Bauer
- Posts: 3563
- Joined: Wed Dec 12, 2018 5:24 pm
- Location: Ireland
- Has thanked: 1 time
- Been thanked: 87 times
- Contact:
-
- Posts: 3265
- Joined: Mon Dec 24, 2018 1:37 pm
- Location: Slovenia
- Has thanked: 80 times
- Been thanked: 234 times
- Contact:
Re: Tough BMS
OK untill this time i was looking for a complete schematic on my PC, but the only thing i found was my hand drawing when i was experimenting with LiPo cells.
I will post this for now so you can start experimenting. Later however i will try to make a propper schematic to put into DS.
This schematic i s only for the sensing part of BMS. I used R1 and R2 as symbols. It is really 3 resistors in series that allow for fine tuning of turn on point of 4.03V per cell.
R1 = 2K7 + 1K + 10K
R2 = 22K
NPN is MJD122
PNP is BC516
I didnt use LED between transistors since i figured PNP base has enough hysteresis. I only used R3 470R to reduce PNP and TL line current.
I will draw a complete schematic for you when i am able to focus on on the design .
A
- Attachments
-
- Shema za BMS.pdf
- (280.75 KiB) Downloaded 218 times
-
- Posts: 3265
- Joined: Mon Dec 24, 2018 1:37 pm
- Location: Slovenia
- Has thanked: 80 times
- Been thanked: 234 times
- Contact:
Re: Tough BMS
Sorry, missed that post...
Here are 8S modules of this BMS. Each module is connected through opto to others and is active high. So if a cell passes 3V LED on my dash will lit and i will have to slow down and find an AC source...
Later on i experimentally connected top balancing connection to reduce the charging. Also i will have to redesign opto signalling to active low. It will be more failsafe that way.
-
- Posts: 3265
- Joined: Mon Dec 24, 2018 1:37 pm
- Location: Slovenia
- Has thanked: 80 times
- Been thanked: 234 times
- Contact:
Re: Tough BMS
Hi
I revisited my design and i am wondering how to simplify manufacture and signals.
Does anyone know a good signal opto chip for 8, 12 or 16 signals? Of course this would be smd chip.
Or i will just use 4x SFH6916 chip...
I revisited my design and i am wondering how to simplify manufacture and signals.
Does anyone know a good signal opto chip for 8, 12 or 16 signals? Of course this would be smd chip.
Or i will just use 4x SFH6916 chip...