Just read this whole thread. As a newbie I have to admit it is mostly confusing. Seems like, in this case, we've defeated the purpose of using a forum to gather accurate and meaningful data.
I wish there was a way to separate the wheat from the chaff as you say. Guess I'll have to just keep reading reading reading, experimenting and measuring. I wish I understood this stuff better.
Thanks everyone I know you are trying to help.
Unfortunately, voltage curves for lifepo4 batteries are either really complicated or only semi-useful.
I think the quick answer is "don't do it" -- but no one likes that answer
.
Depending on the use case: do you really care exactly what the SOC is? If you just want a rough hand-wavy guide, then this table works well for a 4s battery: (from
http://nordkyndesign.com/practical-characteristics-of-lithium-iron-phosphate-battery-cells/)
Voltage reading | Assessment | Capacity |
13.3V or more | Near full | Over 80% |
Above 13.2V | Plenty of reserve | At least 70% |
Below 13.15V | Getting on the low side | Less than 40% |
Below 13.0V | Definitely getting low | Less than 25% |
To get something more accurate, it depends on temperature, how long the cells have been resting, whether they were last charging or discharging -- and even after you take all that into account you can hardly tell the difference between 40% to 70%.
And even THEN for just the hand-wavy chart, if your meter is off by 0.05v the guestimate is way off.
I looked for the first result I got for a
search for "DMM" on amazon, an inexpensive $35 unit.
This has:
- Digital Display: 5 ¾ , 6000
- (note: the "5 3/4" is bogus -- I'm going to believe the "6000", which is 3 3/4 digits)
- DC Voltage: 600mV/6V/60V/600V ±(0.8%+3), 1000V (1%+5)
So if you're trying to measure 13.2v (and you believe their specifications), you can get an error of up to:
13.2v * 0.008 + 00.03v --> 0.1056v + .03v --> 0.14v
So you'd see something between 13.34 and 13.06. This makes even the hand-wavy chart a bad guess.
Or you can shell out for something a bit ridiculous like the $250 121GW (
https://www.eevblog.com/product/121gw/):
- Dual 50,000 count display with bargraph
- 0.05% + 5 Basic DCV Accuracy
For the same 13.2v:
13.2 * 0.0005 + 0.0005v --> 0.0066v + 00.005v --> 0.012v
So you'd see something between 13.212 and 13.188.
FWIW: if you combine the el-cheapo meter with a decent reference voltage, you can figure out a calibration factor and get pretty good results. Something like this thing:
https://www.amazon.com/gp/product/B087Q541YH
Stevelk said:
So then the correct way to check SOC would be.. what?
My guess: A shunt and a coulomb meter. But used how and when?
And as a final note: if you want/need an accurate SOC meter (and can't deal with the handwavy voltage meter), a shunt & coulomb counter is the way to go.
You can play around with cheaper things like
two of these. You need two, because they only measure current one way -- you need one for charge, and one for discharge. IMO it's just not worth it if you want something easy to use.
Or you can get more expensive ones from
Renogy or
Victron.