Designing yet another BMS (yaBMS)

[CatNewBee]

Well I think, if you want to sell any gadged nowadays, it needs either an BTLE / WiFi App to configure it and show the values or an optional a touch screen display. I dont think, you will have to flash a new firmware for any change on the ports.

But first it has to work.

[s/v Jedi]

Yeah I’m not gonna sell anything... I will document everything and upload all pictures and code so that anybody able to hold a soldering iron and laptop can replicate or modify/improve. This stuff really is childishly simple, you can buy the little boards and wire them up on a piece of breadboard and that will work.

The prototype will be like that for sure... with just some minor modifications to the modules used, like set an address or remove a component etc.

These low power relays cost nothing, it all is very cheap.

[seandepagnier]

Originally Posted by s/v Jedi

I’m not balancing with the BMS. During the build of the pack, I would bottom-balance them, which should keep them balanced for years. If balancing afterwards, I would plug in one of the $10 active balancers that charge the lowest cell from the highest cell. I may even help a bit by top balancing using a bench power supply on the low cells.

I agree.

Over voltage: the thing is that we want to not charge to 100% all the time. We would stop charging at 80% to extend life. This can only be done by counting Ah so I plan to use a Victron battery monitor for that.

This is difficult because the needs are variable. On a sunny day in winter you do want to charge when you can.

I agree that counting Ah or coloumb counting is the way to go for a variety of reasons.


I question the use of contactors because of power draw. Why not use only mosfets?

A complete power management system as you describe is not what I’m making. But using multiple of these BMS’es must be possible as I will have at least two house banks.

Well.. if you have a simple bms like this it will limit the whole system's efficiency. If you have separate batteries and/or different voltages and variable charge sources all of this has to be linked like "smart grid" to get the best overall efficiency with the most flexibility.

[s/v Jedi]

Originally Posted by boat_alexandra

I agree.


This is difficult because the needs are variable. On a sunny day in winter you do want to charge when you can.

I agree that counting Ah or coloumb counting is the way to go for a variety of reasons.


I question the use of contactors because of power draw. Why not use only mosfets?



Well.. if you have a simple bms like this it will limit the whole system's efficiency. If you have separate batteries and/or different voltages and variable charge sources all of this has to be linked like "smart grid" to get the best overall efficiency with the most flexibility.

The contactors use no power at all because they are latching bi-stable. They only use power while switching over to the other setting. In case of 1,000Ah cells, MOSFET’s can’t realistically do the job, while one Blue Sea Systems 7700 switch can.

The coulomb counting is just an input signal that can be enabled or disabled. You can do that either on the Victron monitor or on the BMS.

I don’t see how you find this BMS setup would limit efficiency. Can you do a specific description of a situation where it would not work efficient and what setup would improve upon that? In my view, the setup is brilliant

[seandepagnier]

Originally Posted by s/v Jedi

The contactors use no power at all because they are latching bi-stable. They only use power while switching over to the other setting. In case of 1,000Ah cells, MOSFET’s can’t realistically do the job, while one Blue Sea Systems 7700 switch can.

Interesting.

The bistable takes care of power consumption mostly but there is still contact resistance which is probably at least as high as mosfets.

As for 1000ah cells, what is the issue? It is not the cell rating but the current rating of the bms. You can get individual mosfets that can handle 60 amps for $3, and to parallel 10 or more is possible.

I don’t see how you find this BMS setup would limit efficiency. Can you do a specific description of a situation where it would not work efficient and what setup would improve upon that? In my view, the setup is brilliant

If your bms only protects against over-voltage over current and under volts over temperature then it won't matter. The voltage limit you set in this bms can be higher than what the solar charge controller uses.

There is a balance of having the highest state of charge and having the longest battery life. It seems both are desirable but work against each other for lithium unlike lead acid which likes to remain charged. To get the best mix of this, the charge controllers should communicate because anticipated power future solar and future wind etc.. can factor into the controllers decision for what to do now.

[s/v Jedi]

For using a MOSFET as a DC switch, in the ON state, you can count on a 1W power dissipation for each 10A DS current. I have a 450A charging capability so that would result in a 45W power dissipation for which I have no room for the heatsink. Most (cheap) BMS’s that use MOSFETs cannit sustain their rated current without switching the battery off the bus, which is unacceptable for serious work.

Also, we need to be able to manually switch the bank, which is not possible with MOSFETs. Every boat I have seen has a manual switch and these Blue Sea Systems switches combine that with remote operation and losses so minute (silver alloy contacts) that there is no heatsink requirement at all.

I do not understand your specific example of why my suggested setup is not as efficient as possible. It seems it does everything you suggest, incl.the possibility of not fully charging to increase cycle life... for which it communicates with the coulomb counter as well as the charging sources... without switching the battery offline.

Originally Posted by boat_alexandra

Interesting.

The bistable takes care of power consumption mostly but there is still contact resistance which is probably at least as high as mosfets.

As for 1000ah cells, what is the issue? It is not the cell rating but the current rating of the bms. You can get individual mosfets that can handle 60 amps for $3, and to parallel 10 or more is possible.

If your bms only protects against over-voltage over current and under volts over temperature then it won't matter. The voltage limit you set in this bms can be higher than what the solar charge controller uses.

There is a balance of having the highest state of charge and having the longest battery life. It seems both are desirable but work against each other for lithium unlike lead acid which likes to remain charged. To get the best mix of this, the charge controllers should communicate because anticipated power future solar and future wind etc.. can factor into the controllers decision for what to do now.

[witzgall]

Maybe you are too far along to turn back, but ZEVA sells their cell monitoring boards that communicate via Canbus, and they publish the protocol if you wish to roll your own controller and display. I have the complete system and it is working well for me thusfar.

If you don't want the board to balance, they it can be used that way.

https://zeva.com.au/index.php?product=133

Chris

[s/v Jedi]

Originally Posted by witzgall

Maybe you are too far along to turn back, but ZEVA sells their cell monitoring boards that communicate via Canbus, and they publish the protocol if you wish to roll your own controller and display. I have the complete system and it is working well for me thusfar.

If you don't want the board to balance, they it can be used that way.

https://zeva.com.au/index.php?product=133

Chris

Thanks... but products like that are the reason I’m building my own. Nice for EV’s, not for boat house banks (complete overkill)

[witzgall]

I guess that depends on your needs. I used it on a house bank, with their display and "Main" unit, as it was one of the few that would work for me:

qty 5 - 10 cell 24v modules (https://altairnano.com)
LTO chemistry

I know you are of the camp of not doing any balancing. These do very mild balancing, either only at the top of the charge or dynamically. I am using the dynamic balancing, and it is light handed, but I think it is a good approach.

I don't think the 12s boards are overkill, if you are going to develop the main board and or display yourself. One could use a simple canbus enabled arduino or Pi and use a web interface for all viewing and configuration. The development time to do that would be, IMHO, exponentially less than doing it all yourself. You would get the benefit of a very well vetted cell monitoring board, and then you could concentrate on the logic on what to do with that info tailored to your needs.

Chris

Originally Posted by s/v Jedi

Thanks... but products like that are the reason I’m building my own. Nice for EV’s, not for boat house banks (complete overkill)

[hzcruiser]

Originally Posted by boat_alexandra

Interesting.

The bistable takes care of power consumption mostly but there is still contact resistance which is probably at least as high as mosfets.

The on resistance of a MOSFET depends on a lot of factors, but as a ball park figure, it's around 150 mOhm for P-MOSFET, which is too high for currents above 10A.

The N-channel-MOSFETs have an Rds(on) of less than 10 mOhm (e.g. 4 mOhm for the IRF1404). The heat dissipation is R*I^2, or as an example, 400*R for 20A or 4W.
And just putting 5 or 10 in parallel is not as trivial as it may seem.


Relays have a contact resistance of about 2 mOhm, but on top of that you need to connect to their tabs, just like you would to the FET. Hence one challenge with FETs is to get the current to the legs with as little losses as possible.

[...]

There is a balance of having the highest state of charge and having the longest battery life. It seems both are desirable but work against each other for lithium unlike lead acid which likes to remain charged. To get the best mix of this, the charge controllers should communicate because anticipated power future solar and future wind etc.. can factor into the controllers decision for what to do now.

That's part of the problem: we don't want to replace (all) the charge controllers (CC) with a "compatible" one.

Instead I disconnect the CC from the bus when the batts should be kept at 70 or 80% SOC. The CC also monitors how many Ah (or Wh) were being produced, which is very useful to compare different panels.

[s/v Jedi]

Originally Posted by witzgall

I guess that depends on your needs. I used it on a house bank, with their display and "Main" unit, as it was one of the few that would work for me:

qty 5 - 10 cell 24v modules (https://altairnano.com)
LTO chemistry

I know you are of the camp of not doing any balancing. These do very mild balancing, either only at the top of the charge or dynamically. I am using the dynamic balancing, and it is light handed, but I think it is a good approach.

I don't think the 12s boards are overkill, if you are going to develop the main board and or display yourself. One could use a simple canbus enabled arduino or Pi and use a web interface for all viewing and configuration. The development time to do that would be, IMHO, exponentially less than doing it all yourself. You would get the benefit of a very well vetted cell monitoring board, and then you could concentrate on the logic on what to do with that info tailored to your needs.

Chris

I have a 4S setup, 12V. Cell monitoring is done. The total cost of all cell monitoring, central microcontroller, real time clock and display is $40. I need to add 6 relays so cost will be $50. Why would I do CANbus for a house battery?
I understand your choice for this unit because like you say, there isn’t much available. But I can make my own that is exactly like I want it for $50

[hzcruiser]

Originally Posted by s/v Jedi

Next projects I have lined up are an alternator regulator and power distribution modules that replace conventional breaker panels... with a microcontroller, display and pushbuttons on each branch. Float switches, timer solenoids, timer outlets etc. can all be replaced with software routines.

You got big plans there! Lots of time on your hand?

[s/v Jedi]

Originally Posted by hzcruiser

You got big plans there! Lots of time on your hand?

None at all... currently building new refrigeration cabinets, just tore out all the old

I always start early with projects like these, complete them in my mind a couple of times, come up with new versions before prototyping

Right now I’m thinking of features for the new style breaker panel... auto-off after x minutes idle, auto-off after x minutes of y current, immediate off when current higher than z, dimming functions, float switch routine, timer solenoid routine.
There was a guy who built a chain counter with magnet and reed relais on windlass shaft. I’m thinking of a windlass controller that you instruct to deploy 3:1 chain instead of so many feet and this list never stops

[hzcruiser]

Originally Posted by s/v Jedi

None at all... currently building new refrigeration cabinets, just tore out all the old

Oh, I did the same thing, after the old and very inefficient fridge gave up the ghost. It had a built-in inverter and the compressor was hence running on 240V only. Insulation wasn't too good either. Maybe we should open a new thread and discuss how you build your cabinets, as I could use some input there?

I always start early with projects like these, complete them in my mind a couple of times, come up with new versions before prototyping

Same here.

Right now I’m thinking of features for the new style breaker panel... auto-off after x minutes idle, auto-off after x minutes of y current, immediate off when current higher than z, dimming functions, float switch routine, timer solenoid routine.

Yes, the list of requirements can be endless and also varies from user to user. Therefore you'd want a system where you can easily configure all those parameters.

We came to the conclusion that a simple programming language was needed in order to implement a cause and effect system across several components on a boat.

A simple example would be a night light: a PIR module is in the galley and detects movement on the boat. During particular hours one or more night lights comes on for a specific time, allowing the person to move around the boat safely. The night light can be dimmed to a user-settable level and be set to turn on to, say, 25% at night and then increase to 50% when movement was detected.
For full brightness in the evening hours, just turn on the switch or click a button on your phone or tablet.

Another example is to disconnect some of the LFP batts once they reach a particular SOC. Right now with no one on board, only one or two batts are on the bus continuously, the others are in stand-by. The active ones are being charged as usual.

There was a guy who built a chain counter with magnet and reed relais on windlass shaft. I’m thinking of a windlass controller that you instruct to deploy 3:1 chain instead of so many feet and this list never stops

Yes, I've had the same idea. My windlass motor is below deck and adding a magnet and pickup wouldn't be that hard.
Again, the sensor is just picking up rotational information and feeds it wirelessly to the small central computer. From there one can make decicions like lowering or weighing anchor or how much rode to deploy, based on the depth.
On top of that, I'm measuring the current to the windlass motor which would allow to (briefly) turn off the motor if the current gets too high because the hook is stuck. That can avoid burning out the motor windings.

So yes, there are lots of ideas, not every idea should be deployed on every boat, but at least we have the option of doing it, right?

[s/v Jedi]

Originally Posted by hzcruiser

Oh, I did the same thing, after the old and very inefficient fridge gave up the ghost. It had a built-in inverter and the compressor was hence running on 240V only. Insulation wasn't too good either. Maybe we should open a new thread and discuss how you build your cabinets, as I could use some input there?

Are you on Facebook? On the “s/v Jedi” page I document daily; when the project is finished, I often write a blog article on it. This time around (it’s the 2nd time we completely redo the refrigeration) we went for a radical change from custom boxes to the stainless expedition boxes from Snomaster and build cabinets with sliding trays for those