Coulomb counting

[the1andonly]

I intend to make my own battery monitor (Pb wet).
My reasons are to stimulate my little grey cells not to save the odd 100 quid ,that has been done with a simple Chinese one.
1st question the comercial ones do they monitor current in and out with separate shunts or hall effect sensors.
My setup : solar 120w votronics 30a controler feeding both hab and starter batteries 100w solar dealer fitted( cheap) mppt controller, 2 x 110Ah (PB wet) CBE mains charger.
Does the CBE charger connect directly to the battery or to the cbe200 300 distribution centre (rapido)?
My intention is use an arduino mega to measure current in and out of the battery bank
The solar is easy to isolate and measure with hall effect acs712 or shunts, how to identify the alternator and mains charger feeds.
I've already used a uno and acs712 to check solar but up in the wardrobe is not the best place for all the wiring (swmbo). Down in the battery box also allows me to measure battery temperatures while I'm at it.
Previously measured temperatures in 6 places in the bongo (ds1820 with uno)
So back to title.
Has anyone allready used arduino or pi based processors to measure soc. I'm thinking of simple coulomb counting to start with, or any known algorithms I could use. Refinements using battery/ ambiant temperature could be Included if i knew what they were.
Sorry for the long post but I hope all relevant information is here.

 

[Augusta08]

I'd be tempted to simplify things by using a single shunt and calc ampere hours in and out from a known condition. You could always calculate battery health from temperature and voltage. Overcomplexity doesn't really buy you anything but poor MTBF as I see it. Gerry

 

[the1andonly]

Calibration is vital and fully charged a good starting point, actual capacity is unknown as they are now 2 years old. One problem was in my head as the arduino only measures poitive voltages so could not see direction of current (numpty) ive now sussed it. I am hoping to get a slightly more accurate method than voltage curve. Is there a source of curves at different temperatures? Thanks for your i put

 

[hilldweller]

Great fun, it will keep you off the streets. Expect zero understanding or encouragement from a certain female.

I once fitted a NASA one. Did what it said on the tin, using a single shunt.
Got the Chinese one in a drawer, looked like it would work but basically I had that nagging doubt "too cheap".

I can picture your moment of glory, maybe this time next year. "Look dearest beloved, today we used 41Wh, our solar delivered 47Wh, yippee we are ahead of the game". You will remember her answer for the rest of your life.

 

[the1andonly]

Allready had the reply several times been married approx 39.22yrs.
Last time was when i fitted an arduino to measure battery voltage in the toad whist towing ( normal setup is not to charge battery, but i wanted to know what happens, will know next year). 'Why?' Would have done.

^{_{Subscribers  do not see these advertisements}}

 

[hilldweller]

normal setup is not to charge battery,

It beats me how you could charge when towing on the basis every car I've known you would have to spin up the engine to get the alternator turning.

 

[the1andonly]

Charge From the towing vehicle, pin 10 on a 13 pin socket. Pin 12 is also reserved/ not used (depending on issue of spec) for trailer awareness . Pin12 to earth used by a number of ECU s in early days.

 

[hilldweller]

Charge From the towing vehicle

Ahhhhhhhhhhhhhhhhhhhrg, Of course. I'm losing it.

 

[Mike800966]

Great Stuff, it will keep you occupied for months.
Installed a NASA battery monitor many years ago and its been really helpful.
Also I do have a spare nasa 100amp shunt if its of interest I think produce's 100mv at full current if its of interest
Mike

 

[the1andonly]

Currently (forgive the pun) got a few 30A acs712 and a 50A shunt (used for the Chinese meter) used for my initial trials. Collecting parts for the other project as well.

^{_{Subscribers  do not see these advertisements}}