diy solar

diy solar

12v BMS design. What would you like to see

G

Goboatingnow

Solar Enthusiast
Joined
Jul 3, 2022
Messages
1,325
Im looking at a bms project for my boat , so will design one around the esp-32C series which is RISC V based

standard features
* 5A switched capacitor based top balancing
* CAN, wifi and BLE interfacing but no idea what protocols to do ( possible Victron VE.CAN ) possible RV-C protocol
* ISO ABYC compatible
* lve hve , overcurrent and individual cell temp monitoring
* re start charge signal , stop charge signal , disconnect non priority loads signal
* local status panel ( lcd )

Any killer features you think might be good.

Remember this is not for big domestic solar but boat and RVs
 
Master/slave BMS approach (for better comms regarding to expansion)...

I kind of like the idea of the BMS123-Smart approach where they just use single-cell BMSs that daisy-chain the comms (expandable up to 255 BMSs, to support high cell-count systems), where you have one master which comms with the inverter via one plug, and other connections for the backend on the daisy-chain network, using relays to control charging and another for loads.

Or have a separate master control unit (we'll call it an MCU) which isn't a BMS data sensor itself but talks to all the cell BMS modules (which would be the slaves on the network), and that MCU would be the gateway unit the inverter interfaces with, that compiles all the comms and data together and communicates with the inverter...

Ref:

European battery storage manufacturer and distributor | GWL Group

A wholesale distributor of LiFePO4 and LTO battery cells, photovoltaic solutions, and electric car charging systems. Worldwide delivery of Industry grade batteries directly from Czech (EU) warehouse.
shop.gwl.eu shop.gwl.eu

(Oh, I guess you didn't say what size boat or class of RV, perhaps this type of architecture is too elaborate for smaller pleasure boats or normal class RVs. I just liked the idea of a BMS setup where it is not tied to a particular cell-count)
 
Last edited:
Its always a struggle between enhanced functionality while keeping the PCB down to a reasonable size. I wanted to replace my 200A golf cart contactor (140mA parasitic load) with a magnetic latching contactor but other than Batrium which is limited to 30V, none of the other BMS brands seem to have the capability to generate the momentary On/Off pulse. Not to mention that the 12V version of such a magnetic contactor requires 3 to 6A to function. Would be a cool feature though. Eliminates the need to use Power FETS or always energized contactors.
 
I think the cell level bms is a very interesting I could seperate out the active balancer as this is a seperate module per battery

There is a ATTINY cell based open bms project already
 
Goboatingnow said:
Im looking at a bms project for my boat , so will design one around the esp-32C series which is RISC V based

standard features
* 5A switched capacitor based top balancing
* CAN, wifi and BLE interfacing but no idea what protocols to do ( possible Victron VE.CAN ) possible RV-C protocol
* ISO ABYC compatible
* lve hve , overcurrent and individual cell temp monitoring
* re start charge signal , stop charge signal , disconnect non priority loads signal
* local status panel ( lcd )

Any killer features you think might be good.

Remember this is not for big domestic solar but boat and RVs
Click to expand...
You’ve described the REC active 4s BMS pretty well there.
 
toms said:
You’ve described the REC active 4s BMS pretty well there.
Click to expand...
Yes I like that product but I can’t change aspects of its operation , I’m hoping they will be at METS in November

I want a distributed drop in battery style BMS. , ie i want a 300Ah 12V . System , I’m going to build it from 3x4S 100aH Cells. The BMS will monitor each cell AND and every cell for voltage and temperature , ( and it’s 4S for current ) each battery BMS will network to its neighbour so that the combined system will act in unison. This means I can degrade my battery system back to 200aH to,100aH as needed and allows maintenance and cell replacement ( I’m using cheap Chinese cells ) every cell will be individually monitored. There is no “ hidden “ parallel cells that are not monitored.

The REC isn’t designed to do this

For me the future is networked drop in batteries. So I want to get ahead of the game

Having each and every cell individually monitored for temp and voltage , means this is the safest LFP configuration possible on the planet !!
 
Goboatingnow said:
Having each and every cell individually monitored for temp and voltage , means this is the safest LFP configuration possible on the planet !!
Click to expand...
Keen to see the finished product. My experience is multiple BMS creates a more unreliable system than parallel cells.
 
toms said:
Keen to see the finished product. My experience is multiple BMS creates a more unreliable system than parallel cells.
Click to expand...
Why can you outline the engineering reasons for that. ?

In my experience multiple redundancy creates reliability, like flight control computers.
 
I’ll post some prelim schematics on Nov. once my sailing winds down , plan is pcbs by dec
 
Goboatingnow said:
Why can you outline the engineering reasons for that. ?

In my experience multiple redundancy creates reliability, like flight control computers.
Click to expand...
More components = more chances for component failure.

If redundancy is important, build more than one complete system.
 
toms said:
More components = more chances for component failure.

If redundancy is important, build more than one complete system.
Click to expand...
There’s no engineering basis for your comments. Complexity where such complexity often involves internal redundancy and self monitoring brings significant increase in reliability. Good component choices and conservative design add That reliability.

God knows I’ve enough military space and industrial product design under my belt.

The philosophy of this design is in my case with 3x 4S packs , I have in effect three completely redundant battery systems each with it own BMS and local safety disconnect contactor. Even though the three can act in cohort , each one can also function without the others.

I don’t think it’s possible to design a better system in this regards.
 
12V systems are often used in harsh conditions so a system for ensuring Lifepo4 isn’t charged at low temperatures that could damage it

Ideally a system that redirects charge power to warm the battery
 
toms said:
More components = more chances for component failure.
Click to expand...

This is only true if the additional components are added for another reason than redundancy, which is not the case here.

If you add redundancy (and doing so, you're adding components) then your system as a whole will be more resilient to failures. That is the reason why RAIDs work in the first place for example. Also, why everything in the aero and spatial world is at least single and often double redundant.
 
ianganderton said:
12V systems are often used in harsh conditions so a system for ensuring Lifepo4 isn’t charged at low temperatures that could damage it

Ideally a system that redirects charge power to warm the battery
Click to expand...
In case each and every cell in my case there is temperature monitoring so it’s easy to have a charge inhibit line

In low temp environment it’s may ( under certain circumstances ) make sense to do battery heating in some cases like mine it makes no sense
 
Last edited:
Goboatingnow said:
There’s no engineering basis for your comments. Complexity where such complexity often involves internal redundancy and self monitoring brings significant increase in reliability. Good component choices and conservative design add That reliability.

God knows I’ve enough military space and industrial product design under my belt.

The philosophy of this design is in my case with 3x 4S packs , I have in effect three completely redundant battery systems each with it own BMS and local safety disconnect contactor. Even though the three can act in cohort , each one can also function without the others.

I don’t think it’s possible to design a better system in this regards.
Click to expand...
I agree with your sentiment.

In practice i haven’t seen anybody achieve this yet. I agree that this is predominantly due to the use of unreliable BMS systems.

Also while battery redundancy is good, that doesn’t address the issue of inverter or charge controller failures.

In your situation, i would build 3 completely independent systems using a proven reliable BMS (eg REC). That would be a fraction of the cost of designing and manufacturing my own BMS. I don’t follow the reasoning for redundancy in the battery system only.

I will follow your progress keenly though, i think your logic is sound and it would be good for someone to match the theory with a long term practical result.
 
toms said:
I agree with your sentiment.

In practice i haven’t seen anybody achieve this yet. I agree that this is predominantly due to the use of unreliable BMS systems.

Also while battery redundancy is good, that doesn’t address the issue of inverter or charge controller failures.

In your situation, i would build 3 completely independent systems using a proven reliable BMS (eg REC). That would be a fraction of the cost of designing and manufacturing my own BMS. I don’t follow the reasoning for redundancy in the battery system only.

I will follow your progress keenly though, i think your logic is sound and it would be good for someone to match the theory with a long term practical result.
Click to expand...
Well assuming I’m not charging myself professional EE labor rates my current rough estimate is each 4 cell bms will cost $40 in components.

This would be with mosfet overcurrent control as well as one single bank based contactor disconnect to ensure disconnect even if the mosfets fail.
 
Goboatingnow said:
Well assuming I’m not charging myself professional EE labor rates my current rough estimate is each 4 cell bms will cost $40 in components.

This would be with
Click to expand...

Goboatingnow said:
overcurrent control as well as one single bank based contactor disconnect to ensure disconnect even if the mosfets fail.
Click to expand...
I charge my time at $100AUD/hr when working for myself (i have plenty of projects i can charge at that rate)

If you can spend less than 10 hours on your project i tip my hat to you.

Since 2012 when i first started working with LiFePO4 cells i’ve seen dozens of people try to do the same thing as you. I hope you are the one to succeed.
 
You must log in or register to reply here.
Back
Top