State of Charge Determined by Under Load Voltage

[Dockhead]

Originally Posted by conachair

Best mantra IMHO is "I'm certain to be wrong about everything, it's just how wrong and does it matter?"

The earth is almost but not quite flat..

That's a very subtle point, made exceedingly well with the last sentence Bravo.

[john61ct]

Originally Posted by Dockhead

The Smart Gauge attempts to guess at SOC achieved during charging, something you can't do with voltage, but these guesses are very weak.

As the docs state, off by 4% while charging, a bit less maybe under load.

Accuracy better than 2% at rest.

In either case more accurate than shunt-based coulomb counters, so I wouldn't use the term "guessing".

Many people use both, since a good BM keeps a log you can export for analysis.

Setting up initial benchmarks when the bank is new, cross-referencing with other methods is very helpful to get a handle on capacity losses and other issues over time.

But only the SG keeps abreast of the AH walkdown automatically, BMs require you to revise the manual capacity entry over time or slowly the SoC reading will get unrealistically optimistic. More quickly in a PSOC or other abusive scenario.

I'm sure the voltage vs SoC curve changes with the walkdown too, likely differently for each battery type. Which is why having the initial benchmarks and cross-references would be useful, no substitute for getting familiar with the behavior patterns of your unique setup.

[john61ct]

Originally Posted by conachair

Best mantra IMHO is "I'm certain to be wrong about everything, it's just how wrong and does it matter?"

The earth is almost but not quite flat..

Bit like Eisenhower about plans being useless, but the planning process essential.

[S/V Alchemy]

Is this MaineSail's new, stand-alone site, or is the Compass Marine site still active?

[Oceanride007]

I recommend casual readers interested in this method to first get a Digital Voltmeter capable of reading to hundreds of a volt, mV would be even better but as far as I know only a Hand held VM can do that, as I also agree "Data is lovely"
Making a graph Voltage against SOC is great, I haven't got mine yet as at anchor, hard to figure where 100% SOC is, but I got useful data and downloading amps consumption and charging has also given useful data from a recording devise, but the instrument I used was delicate, didn't like the Marine Environment, and I've yet to find something more robust, any suggestions?
Anyway I digress, the purpose of this post is to suggest to the converted that Volts against SOC at 0 amps is great, but I also think a few more consumption curves at say 2.5 Amps, 5 amps, 7.5 amps, these amps for me would coincide with rough at anchor, day sailing, night sailing consumptions. Now there is nothing new in this, many years ago, I have seen typical consumption and charge curves for FLA in text books. And of course interpolation in between values is easy.

[conachair]

Originally Posted by Dockhead

That's a very subtle point, made exceedingly well with the last sentence Bravo.

Cannot take credit for the flat earth example,, that was Carl Sagon - might have been this video?, todays wifi too shakey to stream youtube.

And very relevant, IMHO, to pretty much any technical post here.

Our western cultural history and modern view of technology tends towards a very black and white view of the universe where reality can be defined exactly, something the universe seems determined not to allow

All we can ever get (especially with batts) is an approximation, better to take this onboard then determine (means guess or ignore for most of us ) if it's accurate enough.

https://www.youtube.com/watch?v=IFWpVIF5g8g

[Dockhead]

Originally Posted by Oceanride007

I recommend casual readers interested in this method to first get a Digital Voltmeter capable of reading to hundreds of a volt, mV would be even better but as far as I know only a Hand held VM can do that, as I also agree "Data is lovely"
Making a graph Voltage against SOC is great, I haven't got mine yet as at anchor, hard to figure where 100% SOC is, but I got useful data and downloading amps consumption and charging has also given useful data from a recording devise, but the instrument I used was delicate, didn't like the Marine Environment, and I've yet to find something more robust, any suggestions?
Anyway I digress, the purpose of this post is to suggest to the converted that Volts against SOC at 0 amps is great, but I also think a few more consumption curves at say 2.5 Amps, 5 amps, 7.5 amps, these amps for me would coincide with rough at anchor, day sailing, night sailing consumptions. Now there is nothing new in this, many years ago, I have seen typical consumption and charge curves for FLA in text books. And of course interpolation in between values is easy.

I actually did this, and in fact spent a couple of years doing it. I took frequent specific gravity measurements to establish real state of charge.

I found that with my bank and with my loads -- all voltage measurements done with no recent charging and no recent heavy loads -- there is very little difference between open circuit voltages and loaded voltages, at any given state of charge. To put it another way -- if I switch on a load of a couple hundred watts, there is no measurable voltage sag in my system. It takes some tens of amps to pull the voltage down. YMMV of course depending on the relationship between your loads and your bank capacity. I have 420 A/h @ 24v.

What is most important, and what applies in every case, provided only that there is no charging going on and no surface charge, is that if you don't know the load, you do always know that the real state of charge, will be higher than what is indicated by reading the open circuit chart. So it means that you can't go wrong by using this chart. There is no need to interpolate anything unless you're trying to understand the SOC to a single digit percentage, but I never saw any point in that.


I agree about being careful with voltage measurement. I have two sources for this data -- my SmartGauge, wired up directly to the batteries with heavy cables, and a panel meter, similarly wired.

The SmartGauge has digits for hundredths of a volt, but does not actually give such data (last digit is always "0"). My panel meter reads to tenths. They always agree exactly.