ActiBMS - Discussion about the DIY active balancer BMS

juan3211 · 1 June 2020 04:54

Hi I have done an ATMEGA328P version of DIYBMS version 4. You need to sleep with IDLE not power down. That way you can hear RX/TX serial port

hitech95 · 1 June 2020 09:42

Yup, but it will use more power. There was a pool some time ago.
People prefere a complete shutdown instead of a sleep.

juan3211 · 1 June 2020 09:54

I get 1 mA in IDLE sleep mode. 1.67mA connected to RX/TX bus. I am using arduino’s pro minis.

I think 2mA is quite low, think about how much power is wasiting WEMOS meanwhile

How much power get attiny841?

hitech95 · 1 June 2020 10:59

On sleep the at tiny can use up to 4uA with the WDT active. So the difference is quite big.

stuart · 1 June 2020 12:42

Power management on the ATTINY841 is very good, and I often get single digit uAmp power readings when in standby. This really makes a difference to small battery packs.

juan3211 · 1 June 2020 13:07

May be I can get less mA, but 1mA x 8 cells is 8mA. How much is the wemos itself with wifi connection?
What I mean is I think that it will be better to, for example, sleep wemos 2 seconds, than sleep all cells’ modules in idle vs power down.
In my case, I use arduino pro mini because of spare parts in home and I don’t feel confidence with smd components.

I also sleep more time beacuse I use 9600 baud rate instead 2400 (arduino pro minis have crystal/resonator), round trip is less. I didn’t do calculations but pro mini’s are wake up much less time than attiny.

And because round trip is less, we could also reduce wemos consuption. Is it wasting about 18mA-50mA ?? If we sleep it 2 seconds and wake up again, we are reducing its power 1/3 (1 second of up time). Also in that situation attiny will reduce power also, as cell’s modules willl be more time sleeping.

I think that 1 mA is a lot of in compare with 1uA. but we also have other places to reduce power consumption.

Attiny841 is a good IC. Stuart has done a great job searching and selecting it.

hitech95 · 1 June 2020 19:59

Yes, totally agree.
The ESP8266 can go up to 250mA on wifi TX & RX. But that is a special case only when configuring the controller.

Yes, on my plan there was an idea about using the crystal to increase the spped and reduce the communication ring time. but with the low available space on the board I decided to remove it.

We could crete a solder pad to configure the baud rate for an advanced controller VS the DiyBMS controller. (Lower baud rate due to lack of osc)

I was doing some research for a discrete Inductive balacing circuit. Dammit it require a lot of passives. I was wondering if we could use a asmall H bridge IC (Like the DRV8839) and feed it with a PWM signal. but we would still need the current measurement for costant current charging.

IF someone know a all in one package with mosfet and its drivers that is more versatile than an H bridge please share the chip PN!

stuart · 1 June 2020 20:27

I’ve already done this on my 4.3 version, but that’s not public (had a few errors)

hitech95 · 1 June 2020 20:32

Good to know, are the Attiny still cheaper than the AtMegas? when I’ve designed my board it was only 0.3 cents cheaper on LCSC.

Have you ever considered some active balancing circuit for your BMS?
After some analisys I still think that the ETA3000 chip is a quite good compromise for the price (0.50USD) , but I can only find thoose in the SOT-23 format.

juan3211 · 7 June 2020 15:29

Hi, today I have measure (my power supply internal ammeter) 80mA of continuos power in Wemos.
I dont know if it is more or less the value that have to, but imagine that we have only a 50% dutty cicle, we can save 40mA, and it is much much much more than 4/8/16 cells wasting 1mA more.
I think we have to improve Wemos code, don’t we?
Bye,

fhorst1 · 7 June 2020 15:57

Hi,
New here to this forum, experience with pcb, arduino, solar/wind/water energy and Lifepo4 storage.

We live off-grid in Thailand and have 13.5kw peal solar, lifepo4 736Ah@S16, about 37.5kwh (3152Ah + 1280Ah)

Pcb’s and parts are ordered at jlcpcb (60 + 5 controler V4.21)

After this short introduction, the question

Lifepo4 lives the longest between 15-85% SOC.
That’s where the challenge begins.
The Charger (Solar MPPT) doesn’t yet go into CV mode.
CC for us is not really constant, it varies due weather conditions, time of day (solar) and energy consumption.

The battery array is never in rest! Always active charging or discharging.

For me, I need to do active balancing.
The batteries between 10 and 60% charge for weeks or months, depending on the weather.
40 panels don’t produce 13.5kw during clouds and rain.

The DIYBMS can control 4 relays per control board.
can those be used for higher amperage balancing??

I have 1A active balancer, and it helps a little, during higher charge and discharge, it’s simply not enough.(aliexpress)
the larger (working) ones, 5 to 10A go above $ 500.
that’s kind of outside budget,

I have one 152Ah lifepo4 cell extra that can be used as dump load for the highest cell and transfer to the lowest
it would be great if that can be re-used.

i don’t know how to get it “walking” over the 16 serie sets, or connect to the highest, and then to the lowest cell.

The information about highest and lowest voltage is available via de DYIBMS, but there it stops for me.

Can you help me to re-use the information form DYIBMS to make “flying capacitor(or better lifepo4 cell)”
Or have better solution to keep the pack in balance during the low charged season?

I’m seeing 10 to 15% difference in SOC, after a few weeks.
now manually use this cell to charge at the highest and discharge at the lowest when the array gets below 25%
but I like to automate this.

As we live off grid, i can’t charge the cells extra.
Not to that extend. (we have generator)
Bottom balancing is just as difficult, empty battery is no power…

Help is highly appreciated!!

side step:
Active balancing IMO should not met V5 of DYIBMS, but better addition to add to BMS, for sure not all people would need.

G33RT · 7 June 2020 17:38

Where can we download the files for the ESP and firmware for the modules ?

At GitHub - actiBMS/actiBMS_module_firmware: actiBMS Cell Module firmware there are no firmware files available.

Am I missing something ?

Update:

I found FW the Arduino files here:

stuart · 8 June 2020 07:28

No I don’t think it can be used for active balancing.

Normally the “need” for active balancing indicates a faulty cell or a cell that is not the same capacity/state as the others. Active balancing “hides” this issue by automatically topping up its charge.

hitech95 · 8 June 2020 08:37

No the relays outputs have the same functionallyty of the Original DIYBms. In fact my module uses the same protocol, but instead of passive balancing it uses an active balance IC.

That said, the balance curent is around 1A. This is due to the limitation of the balance IC. (It is the same used on the cheap aliexpress boards)

hitech95 · 8 June 2020 08:53

Yea, it is not in the default branch. Due the fact that it has not been completly tested.
So you have to check that it’s working fine. Especially the wakeup from the deep sleep.

It have an HAL with the common methods defined (Interface), from that you have to implement the methods for your HW. Tts OOP. Withthe HAL you can port it to different uC you just have to implement the HAL methods:

github.com

actiBMS/actiBMS_module_firmware/blob/dev-arduino/diybms_arduino.cpp

/*
  ____  ____  _  _  ____  __  __  ___
  (  _ \(_  _)( \/ )(  _ \(  \/  )/ __)
  )(_) )_)(_  \  /  ) _ < )    ( \__ \
  (____/(____) (__) (____/(_/\/\_)(___/

  DIYBMS V4.0
  CELL MODULE FOR ATTINY841

  (c)2019 Stuart Pittaway

  COMPILE THIS CODE USING PLATFORM.IO

  LICENSE
  Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales (CC BY-NC-SA 2.0 UK)
  https://creativecommons.org/licenses/by-nc-sa/2.0/uk/

  Non-Commercial — You may not use the material for commercial purposes.
  Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made.
  You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

This file has been truncated. show original

The control logic is independent of the HAL implementation.

Take also note that some debug strings are still printed over the serial bus.

fhorst1 · 8 June 2020 09:49

Hi Stuart,

Thanks for the reply.
Yes, I’m aware that normally the pack should not need Balance.
That is, if you regularly charge it to full, +3.6v

Or bottom balance when you have grid to top up the lowest ones.

No grid, and months no full pack.
That’s just the weather.

We just ordered 14*340w extra.
That should fix most of the problems.
Once every 3 or 4 weeks top charge Balance is sufficient.
The balancing from the BMS (0.035a) a d active balancer.

2 or 3 months without top charge, gives imbalance.
You can call this a cell problem (all new cells) or just natural LIFEPO4 behaviour and the reason why we need BMS to start with
I like to call it the last.

Sometimes, I’m away from home for months, and my wife is no hero for electrical equipment.
She stays far away from it as possible.
And that is a good thing.

Result, they need to stay in balance automatically.

Sadly I can’t program the Inverter / charger to do top charge every 30 days.

So alternative is needed.

You have options?

stuart · 8 June 2020 10:31

Just a reminder, can you please rename the GITHUB repo and not use the diyBMS or v5 labels as this will confuse people thinking its still the original project. Thanks.

hitech95 · 8 June 2020 19:28

Yup,
I’ve asked you about a name suggestion or a preference on a previous post.
I’m going to rename the repo. About the forum topic I don’t know if I can edit the title.

borpin · 8 June 2020 19:34

Yes you can.

hitech95 · 8 June 2020 19:38

Noob here, how?
I don’t have the EDIT button available on the first post.