Battery Monitors

Makana came with a Victron battery-monitor on battery #1. Works great. But I have nothing (except the factory-original volt-meter on the panel) to indicate what's going on with battery #2.

As long as we're still in the it's-too-cold-and-snotty-out-there-to-play-on-the-boat part of the year, that's on my list of things to ponder.

The "easy" thing to do would be to stay with Victron. I could get a second single-battery monitor and mount it next to the one I have. Wiring would be straightforward, and I think I understand it. Or I could replace the existing with a Victron 702-series which has the ability to monitor two batteries. I think I'd need to replace the existing shunt, too, but Victron seems to have made that easy.

Two good options.

But Maine Sail seems to be viewing the BalMar SmartGuage very favorably. In his reviews, he talks about how Ah-counters (like the Victron) get *less* accurate as the batteries age - they may be showing 100% SOC (state of charge) when in fact the battery is no longer capable of storing the 100% of rated amp-hours compared to when it was new.... but because it tracks amps-in and amps-out an Ah-counter isn't aware of that. That means that when the Ah-counter displays 70% (or whatever) SOC, you might actually be below 50% at the battery.

Maine Sail says that the BalMar, by measuring voltage rather than amps, gets *more* accurate over time.

Plus, the BalMar doesn't need to use a shunt, so the wiring is even easier.

So... before I get too far in my thinking on this... anyone using a BalMar and have feedback on it?

And, any electronics-geeks want to chime in on the relative merits of measuring voltage vs. measuring current? (please use short words when explaining)

I'd like to know the SOC of each of my batteries, and have it be a metric I can rely on. The Victron has been very good, but... I'm open to newer and better, if it exists.

Thanks,
Bruce

Smartgauge

http://www.ericsonyachts.org/infoexchange/entry.php?485-2016-New-Batteries-New-Smartgauge

My experience, and it's still working fine.

Loren

Loren Beach said:

http://www.ericsonyachts.org/infoexchange/entry.php?485-2016-New-Batteries-New-Smartgauge

My experience, and it's still working fine.

Loren

Click to expand...

A++ to the Balmar and Loren's blog. I get inconsistent charge percentages off the solar mmpt, but consistent and seemingly accurate readings from the Balmar. Terrific discussion on Marine How To dot com: https://marinehowto.com/smart-gauge-battery-monitoring-unit/. Scroll down a bit to get to Rod's testing techniques.

Cheers,

Max

Bruce, all the write-ups for the Balmar Smartguage are very good. Just note, it will only read "State of Charge" (and voltage) for 1 battery (or bank). For the second battery (bank) it only reads voltage.

Kenneth K said:

it will only read "State of Charge" (and voltage) for 1 battery (or bank). For the second battery (bank) it only reads voltage.

Click to expand...

Huh.

Well, that's something to think about.

Thanks...

Usually second battery is either reserve or starter and as such should be at nearly 100% charge at all times. IMHO in this case it is not worth adding second battery monitor. Simple voltmeter should suffice and you get that for free with some battery monitors. I think smartgauge can do it.

I too have the Balmar and am quite satisfied, no shunt required and seemingly accurate SOC readings. You cannot get SOC from just a no load voltage measurement, would be wonderful if you could but not possible. The current counters measure the charge leaving the battery very accurately but cannot account for the changes in the fully charged mode so since the SOC is the difference between the fully charged value and the charge leaving the battery, the SOC is not accurate after a period of time.

According the web site, the algorithm that measures the SOC from just a voltage measurement was developed by the British Air Force and is proprietary to Balmar. No clue how it works but apparently it gets more accurate over time, just the opposite of the current counters.

I talked to a guy at Balmar about the Smartgauge. Here's what he told me. It's programmed with the charging curves for 400-something (440 maybe) common batteries. The first thing you do when you set it up is select the battery type. I suppose that narrows it down to roughly one fourth of the 440 or so graphs. It then needs to look at at least three charge-discharge cycles from your battery. It records the data from these three Cycles. The three points dictate a rough curve. It then matches that curve to the "best fit" for the type of battery that you've selected and the graphs it has preprogrammed. That's why it gets more and more accurate the more you use it. With more data poinits, it zeros in more on a particular graph, or jumps to another one if that becomes a better fit.

I should have asked, but dididn't, what exactly the graphs are of? I don't know if it's simply state of charge versus voltage for any of the given batteries. That might be too easy. I read somewhere else that the gauge uses a measure of capacitance. Don't know anything about that.

The guy was very helpful on the phone, maybe with enough people we can get them to divulge their secrets.

With a Victron and older, questionable batteries, you can set battery capacity to 80 percent of listed for a more accurate guess.

Question: a one-battery monitor hooked to a two-battery bank with battery combiner: does it read both?

Ken, thanks for the info. Sounds like a reasonable explanation, there is much more information available by measuring the voltage curve during the charging cycle than just a simple one shot voltage measurement.

Christian Williams said:

Question: a one-battery monitor hooked to a two-battery bank with battery combiner: does it read both?

Click to expand...

I presume we are talking about coulomb counting monitor, e.g. not smartgauge. If the the negative terminals of all batteries are wired together and then go through the shunt (as they should) the shunt will capture current from all batteries and thus will "read" both. However your starter or reserve battery is not cycled much and should not affect reading much either. Now, what did you enter as battery capacity? If just house battery then % left should be correct for house battery. Ah reading will always be correct as it does not depend on battery capacity.

There were some speculations floating on the net about smartgauge algorithm and ppl suspected that it measures battery DC and/or AC impedance in addition to voltage. 14 gauge wiring requirement kind of supports this.

Christian Williams said:

Question: a one-battery monitor hooked to a two-battery bank with battery combiner: does it read both?

Click to expand...

"No" is the simple answer. The two banks are not combined during discharge. If both battery's negatives go through the shunt (they shouldn't) you would read both, but wouldn't get meaningful information because the two have different load cycles and voltages while discharging.

sharonov said:

There were some speculations floating on the net about smartgauge algorithm and ppl suspected that it measures battery DC and/or AC impedance in addition to voltage. 14 gauge wiring requirement kind of supports this.

Click to expand...

How do you measure impedance with only a voltage measurement? BTW, the manual calls for #16 wire, but I see that only as a minimum for mechanical strength, and to avoid voltage drop. No other electrical significance.

Tom Metzger said:

"No" is the simple answer. The two banks are not combined during discharge. If both battery's negatives go through the shunt (they shouldn't) you would read both, but wouldn't get meaningful information because the two have different load cycles and voltages while discharging.

Click to expand...

Good point. However in my case, it does not matter much as reserve battery does not cycle, it sits at 100% all the time. Even engine normally starts from the house bank. Nevertheless I'll look into excluding reserve battery current from the shunt next time I am messing with it.

Tom Metzger said:

How do you measure impedance with only a voltage measurement?

Click to expand...

It is not a voltage measurement, it is two wires going to the battery and we have no idea what they do. If you follow this argument, a multimeter would not be able to measure resistance with just two wires. Of course, things are not that simple when dealing with a (very) low output impedance of the battery. One can inject AC signal through these two wires, possibly at multiple frequencies and measure voltage and current (both magnitude and phase) and thus complex impedance. Measuring at low frequency would be problematic as battery impedance (and AC voltage) would be quite low but I would imagine it will increase at higher frequencies (lazy ions ;-). Can one derive state of charge from this? I do not know but seems plausible. Oh, and the impedance of the wires can potentially be corrected for if we measure at two frequencies since frequency response of the wire would be different from that of the battery. At least this is where I would start if I was tasked to design a battery SoC monitor that uses only two wires.

Tom Metzger said:

BTW, the manual calls for #16 wire, but I see that only as a minimum for mechanical strength, and to avoid voltage drop. No other electrical significance.

Click to expand...

My (pretty recent) manual says 14 gauge. Of course, an even simpler explanation for a heavy gauge wire could be that they did not want to see voltage drop from self power draw (15mA).

sharonov said:

Good point. However in my case, it does not matter much as reserve battery does not cycle, it sits at 100% all the time. Even engine normally starts from the house bank. Nevertheless I'll look into excluding reserve battery current from the shunt next time I am messing with it.

Click to expand...

It does matter because you are charging through both batteries and discharging through only one. If using a Smart Gauge you are screwing with the DC impedance, the AC impedance, and the voltage per your explanation. No way could the meter pick the correct curve per Ken's explanation. If you are building in inaccuracies you may as well go back to just using a voltmeter.

And speaking of voltmeters, a voltmeter is an ammeter with a relatively large constant impedance. An ohmmeter is a voltmeter with a constant voltage applied. Both ammeters with two wires, regardless. Generally, on a multimeter they are different two wires.

It is not a voltage measurement, it is two wires going to the battery and we have no idea what they do. If you follow this argument, a multimeter would not be able to measure resistance with just two wires. Of course, things are not that simple when dealing with a (very) low output impedance of the battery. One can inject AC signal through these two wires, possibly at multiple frequencies and measure voltage and current (both magnitude and phase) and thus complex impedance. Measuring at low frequency would be problematic as battery impedance (and AC voltage) would be quite low but I would imagine it will increase at higher frequencies (lazy ions ;-). Can one derive state of charge from this? I do not know but seems plausible. Oh, and the impedance of the wires can potentially be corrected for if we measure at two frequencies since frequency response of the wire would be different from that of the battery. At least this is where I would start if I was tasked to design a battery SoC monitor that uses only two wires.

Click to expand...

I'm an old guy - think big glass bottles - and think of wires doing only one thing in a circuit. I also took Ken's report at face value, a very smart voltmeter. Your idea is certainly plausible. I would expect, though, if your idea is correct they would have used coax connections to the battery.

My (pretty recent) manual says 14 gauge. Of course, an even simpler explanation for a heavy gauge wire could be that they did not want to see voltage drop from self power draw (15mA).

Click to expand...

The manual on line at Balmar today says 16 AWG. http://www.balmar.net/operation-manuals/ The length of the wire would probably have more effect than changing one wire size, but we are talking about several millivolts drop. At higher frequencies the relationship between the positive and negative wires would have more effect, not being coax.

Battery meters

If you have an Amp meter you have an indication of the current charging or discharging the battery. A Volt meter connected across the battery will tell you the voltage at the terminals. A hydrometer will measure the specific gravity of a lead acid wet cell battery, and give a better indication of condition. Gel Cells, and sealed Lead Acid, the only data accessible is voltage and Amperage.
As a battery is used the internal resistance increases, charging voltage will appear higher, and load voltage will be lower.
With the smart monitors, you get what the programmer thinks. The truth in hidden in the software. Newer may be better, as they have the shortcomings of others to learn from.
So, know the age, and usage of your batteries, and consider the approximation of any monitor.

In reference to Tom's comments about the possible Smart Gauge inaccuracy when connected to a bank of more than one battery (connected in parallel), please see http://www.balmar.net/products/smartgauge-battery-monitor/. While I agree with Toms analysis, Balmar indicates in the diagram on that website that there can be as many as three batteries in parallel (and probably more, who knows). Until we know exactly what Balmar is comparing in those curves, multiple batteries wired in parallel in a bank may be ok.

HerbertFriedman said:

In reference to Tom's comments about the possible Smart Gauge inaccuracy when connected to a bank of more than one battery (connected in parallel), please see http://www.balmar.net/products/smartgauge-battery-monitor/. While I agree with Toms analysis, Balmar indicates in the diagram on that website that there can be as many as three batteries in parallel (and probably more, who knows). Until we know exactly what Balmar is comparing in those curves, multiple batteries wired in parallel in a bank may be ok.

Click to expand...

I think potential problems arise not when you have parallel batteries but when you have two banks with an ACR switch. In this case smartgauge will "see" both batteries while charging but only house battery while discharging. I recall somebody saying that smartgauge is not that accurate while charging.. could second battery via ACR be screwing with it?

I understand your comments regarding the ACR. I thought that the ACR that I installed was Balmar and that Balmar would have mentioned any problems with the ACR in place, but looking back at my records, I see that the ACR was made by Blue Seas. So maybe there is a problem with the ACR. Perhaps a call to Balmar will settle the issue.

HerbertFriedman said:

I understand your comments regarding the ACR. I thought that the ACR that I installed was Balmar and that Balmar would have mentioned any problems with the ACR in place, but looking back at my records, I see that the ACR was made by Blue Seas. So maybe there is a problem with the ACR. Perhaps a call to Balmar will settle the issue.

Click to expand...

Batteries connected through an ACR are NOT paralleled in the sense of being one bank. A bank is made up of multiple batteries with both pos & neg connected directly together.

The monitors monitor one bank.