SOC meters for small systems

cptdondocptdondo Solar Expert Posts: 96 ✭✭✭
in General Solar Power Topics #1
I would like to have a SOC meter on two different systems; one is tiny; 200Ah @ 12V and the other is not much bigger at 600Ah @ 12V.  What's a good system these days?  A lot RVers use Trimetric but I'm not overly impressed with them as it's strictly a coulomb counter.  Any others?  I don't really care about monitoring all the different sources; I just want to know how much is left in the batteries.
· Share on Twitter

Comments

  • EstragonEstragon Registered Users Posts: 4,484 ✭✭✭✭✭
    #2
    There really isn't such a thing as a perfect, or even excellent SOC meter.  There are just too many variable, and they'll all tend to drift over a period of partial charge/discharge cycles.  The trimetric does have some limited ability to adjust (eg battery efficiency setting), but will still drift.  Likewise with WbJr monitoring in MN classics.

    IMHO, the only way to accurately measure SOC and monitor battery health in lead acid batteries is with regular SG checks.

    "How much is left" is sort of a squishy number as well.  If running a big load relative to bank size, the effective watt-hours available will be less than a small load, for example. 
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
    · Share on Twitter
  • cptdondocptdondo Solar Expert Posts: 96 ✭✭✭
    #3
    Oh I know.  I spent a summer with a load bank, a 7Ah battery, a bunch of Hall effect sensors, and a crapload of Arduinos learning about SOC, measuring it, compensating for Peukert curves, estimating from current + voltage, I even had SG sensors and I came to the conclusion that I can pick one method that will show 10% SOC and another will show 90% SOC.  

    So that's why I use squishy terms, because with lead-acid there is no accurate way to really determine the charge of a battery.  I'd like to get close, relatively inexpensive, and convenient.  :)
    · Share on Twitter
  • AguarancherAguarancher Solar Expert Posts: 312 ✭✭✭
    #4
    Have you looked at this?


    · Share on Twitter
  • EstragonEstragon Registered Users Posts: 4,484 ✭✭✭✭✭
    #5
    What did you use for SG sensors?
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
    · Share on Twitter
  • cptdondocptdondo Solar Expert Posts: 96 ✭✭✭
    edited February 2019 #6
    Estragon said:
    What did you use for SG sensors?
    These guys: https://www.sensorsmag.com/components/fiber-sensing-to-measure-battery-state-charge - the company name is at the bottom of that paper.

    The sensors were fragile and pretty reliable.  I never moved forward with them as in a mobile environment AGMs make far more sense.

    UPDATE: Just checked the website; the sensors are no longer available.  
    · Share on Twitter
  • kelperiskelperis Registered Users Posts: 45 ✭✭
    #7

    VictronConnect is what I use

    · Share on Twitter
  • petertearaipetertearai Solar Expert Posts: 454 ✭✭✭✭
    #8

    I use the Victron , seems to be good . Input the % efficiency and Purcurt of the battery . It then makes the required calculations . Still as previously said , it is still only a guide . Not an accurate number .

    2225 wattts pv . Outback 2kw  fxr pure sine inverter . fm80 charge controller . Mate 3. victron battery monitor . 24 volts  in 2 volt Shoto lead carbon extreme batterys. off grid  holiday home 
    · Share on Twitter
  • jonrjonr Solar Expert Posts: 1,384 ✭✭✭✭
    #9

    IMO, the key to accurate SOC via coulomb counting is to have a daily reset to a known state. With a system that adaptively corrects out it's errors for your typical usage.

    I am available for custom hardware/firmware development

    · Share on Twitter
Sign In or Register to comment.