Panbo

Smartgauge battery monitor, RC proclaims "paradigm shift"!

... written for Panbo by Ben Ellison and posted on Apr 23, 2014

Smartgauge_testing_courtesy_Compass_Marine.jpgIf you like sailing with some electronics running, or just anchoring without a generator, you're probably very interested in the State of Charge (SoC) of your battery banks. Voltmeters, however, only hint at what's going on, and true battery monitors require careful calibration and the installation of shunts, but still tend to get out of whack over time. Well, darned if the great RC Collins of Compass Marine didn't go to extraordinary lengths to prove that the Smartgauge -- a little known product that's been around for almost a decade -- can somehow accurately measure SoC without calibration and without a shunt, and yet still get even more accurate over time. Apparently when it shows your Charge at 92%, as above, your battery bank really is at 92% capacity...

SmartGauge_test_lab_courtesy_Compass_Marine.jpg

RC Collins is knowledgable on many boating subjects, but I first wrote about his Compass Marine site in regard to a highly detailed 2011 article titled Installing a Battery Monitor. You can learn a lot about traditional amp hour counters in that article, particularly the popular Victron BMV-602 which Collins also used as one reference in the elaborate Smartgauge test setup you see in these photos. Let's note that unlike the Smartgauge the BMV and other monitors with shunts can display realtime current use -- which I found useful for understanding Gizmo's power needs when I first got her -- and can also calculate remaining amp hours in addition to charge percentage, at least theoretically. The Link 1000 Gizmo came with seemed very accurate about actual current use, but nearly drove me bonkers with inaccurate amp hour info (until I was calmed down by battery-smart readers).

SmartGauge_test_lab2_courtesy_Compass_Marine.jpg

A Smartgauge only delivers voltage and SoC (and voltage for a second battery bank) but, wow, it seems to do that amazingly well. While I highly encourage you to read the full Compass Marine Smartgauge review, know that after four months of testing with all four major battery types, RC wrote "If I had three thumbs this product would get all three!" Reading the full test is also a lesson in how difficult battery monitoring really is. In order to calibrate the testing, Collins built a system able to properly measure batteries against their 20 hour capacity rating. Gizmo's "245 A.H. @ 20Hr" AGM 8D batteries, for instance, should each be able to deliver 12.25 amps per hour for 20 hours while going from full charge to 10.5 volts at about 77 degrees. But these batteries -- now at least five years old -- no longer have such capacity (if they ever did) and the difficulty of making an accurate 20 hour test would be strike number one against the accuracy of a traditional amp counting monitor (followed by Peukert's law and other nuances that make this subject so complex).

Smartgauge_wiring_diagram_aPanbo.jpg

In fact, RC sometimes had to run three complete 20 hour discharge/recharge tests to determine the capacity of used batteries, and that's when the Smartgauge really started blowing his mind. Even though it only sees voltage, via a pair of 14 gauge wires going directly to the battery terminals, the Smartgauge usually came up with an accurate State of Charge percentage before Collins and all his precision equipment could even determine accurate capacity. He notes that the device can't handle Lithium batteries and is less accurate when a bank is being charged, but in discharge mode he believes it's more accurate than the two traditional shunt sensors he compared it to even though they were painstakingly calibrated (and won't stay that way on their own). And on top of it all, RC manages to includes some battery geek humor; for cripes sake, if you do act on the results of his hard work, buy your Smartgauge at Compass Marine.

As much as I appreciate RC's testing and writing, I do feel like an idiot for having had a sample Smartgauge for years but never getting around to testing it. I even installed it on Gizmo's power panel last spring but then got sidetracked when I realized that I could use the already wired Link shunt with the wonderful CZone Signal Interface (so SoC and live current load are on NMEA 2000, though of course that setup needs better calibration). Another possible test this season is the Victron Color Control QX discussed here, which could conceivably monitor Gizmo's inverter/charger, batteries (via the latest BMV 700 series) and even the solar panels, with N2K output too.

Finally, I'd like to apologize for an extremely delayed product test to Smartgauge designer Chris Gibson (though I certainly won't come close to RC's super thorough review). The SmartGauge Electronics website is an interesting place to visit but I'm not sure much has happened there since Merlin Power Systems acquired Smartgauge in 2009. At Merlin you can see some of Gibson's other power products like SmartBank split charging, and note that Balmar now distributes Smartgauge in the U.S. (reportedly with excellent technical support). But it's at the original site where you can learn about the Smartgauge test boat -- a "trad style" 70-foot narrow boat -- along with neat details like how the gauge's alarm relay outputs can be used with SmartBank to automate a battery locker venting fan. "Smart" is a badly overused word in modern electronic products, but maybe not this time.

Smartgauge_test_boat_courtesy_Smartgauge.jpg

Comments

Ben, How much power is consumed by this product while monitoring? Is this product left on 24 hrs a day as many products are, or since it does not measure current, can it be entirely turned off when not in use ?

Posted by: Dan Corcoran (b393capt) at April 23, 2014 4:51 PM | Reply

Hi interesting article about the Smartguage and thanks for the link to the test.

I hate to do this, but you're doing the amps per hour thing again.

My gizmo draws 5 amps
It draws 5 amps now, it draws 5 amps tomorrow. Anytime it's on the current is 5 amps
If I leave it on for 2 hours it uses 10 Amp-hours
If I leave it on for 1 hour it uses 5 Amp-hours

Ah, is a product, amps times hours

Amps per hour kind of implies that as each hour goes by the current is getting greater.

"should each be able to deliver 12.25 amps per hour for 20 hours "

I suppose if you really had to put in an extra word you could go with
"should each be able to deliver 12.25 amps continuously for 20 hours "

Sorry, I'll go back to sleep!

Cheers,
Derek

Posted by: Derek Hodge at April 23, 2014 4:59 PM | Reply

Dan, please read RC's full review; all your answers are there in detail.

And sorry back at you, Derek! I still don't get it. For the life of me I do not understand why my use of the phrase "amps per hour" implies that current is getting greater by the hour. Why would it?

If I'm running at 10 miles per hour for two hours (while using 10 amps per hour), I didn't speed up, did I? No, and I covered 20 miles of distance and consumed 20 amp-hours. And isn't running at 10 miles an hour for two hours very similar to saying you maintained a speed of 10 mph consistently for two hours?

I did Google "amps per hour" and can see that it's not used very much, but I also don't see any references to it being a misnomer.

Posted by: Ben at April 23, 2014 7:06 PM | Reply

It's just that it's not "amps per hour", but simply "amp-hours", or as Derek tells it "amps times hours".

"Amps per hours" would instead mean "amps divided by hours", which would mathematically go down as the considered duration goes up, for a constant amperes use (take your 10 miles covered in one hour ; now take two hours to cover the same distance : it's now equivalent to only 5 miles covered in one hour, which is of course less than 10 miles per hour), which would not make any sense with what we consider here, ie energy use...

... as amp-hours are not a measure of a rate, they're instead homogeneous to an amount of energy (eg joules) divided by volts (1Ah=3600J/V).

You may see that as nitpicking, but science asks for precision.

Posted by: norman at April 23, 2014 8:55 PM | Reply

A new mouse-trap?? Have the laws of pysics changed? Sorry, I don't see how anyone could measure battery capacity by only measuring volts. The only way to determine the capacity remaining (or used) in a bettery is to measure the energy used (or replaced). Energy is expressed as 'watts'. Wattmeters need to measure voltage and current. But, watt/hours is what we really want to know, same as your power company measures your household electricity use. However, watts measurements on a boat with relatively constant voltage would be a bit tedious. So we measure energy by measuring current (amps). You have to measure amps going in and amps going out. When compared to time, you can then display an amp-hour capacity used or remaining (depending on how you set the system up). Volts actually has very little to do with it, assuming there's enough to run your device. Currently there are only two reasonable ways of measuring current: use of a shunt, or magnetic field measurement (such as used on clamp-on meters).
For nearly 10 years, I have been using a BEP DCM600 battery monitor, which I have set up to contunously displays amp-hours remaining. This device has been flawless. It is a shunt type of device, and it also monitors 3 battery banks (2 engine and one house, though the amp-hours is only for the house bank). The initial set-up is not hard (using some rules-of-thumb for such settings as Peukert's constant 'k') and have not touched it since.
Of course, if your boat sits for a long time without use, no battery monitoring system is going to acurately cater for self discharge of batteries - short of having some sort of electronic hydrometers on each battery cell. In my case, the boat is used weekly on average and otherwise sits connected to shore-power and on float-charge, so I can assume that when I start my voyage, the batteries are fully topped up.

Posted by: Dainis at April 23, 2014 9:17 PM | Reply

Ben,

I agree that your usage doesn't imply acceleration but I don't think that's the point Dan is trying to make.

An ampere is not a quantity of stuff, it's a rate of consumption of stuff. Specifically, it's the rate of the passage of charge (measured in coulombs) per second through a circuit. The analogy would be a quantity of fuel (measured in gallons) versus the rate of fuel burn (measured in gallons per hour). When you say "while using 10 amps per hour" it's exactly like saying "my engine is using 10 gallons per hour per hour." It just doesn't make sense.

The practical issue is that batteries store charge (coulombs), not amps (coulombs per second). When you say you've used 10 amp hours, what you're saying is that you've been using 10 coulombs per second for an hour, which allows you to calculate exactly how much charge has been drawn from the batteries (10 coulombs/sec * 3600 sec = 36,000 coulombs). That's meaningful and useful. The only use for talking about amperes in the absence of the duration you've been using them (amp hours) is in the context of wire capacity, fuse sizes, etc, where the rate of passage of charge is a concern because you could overheat the wire, blow the fuse, or the like. Going back to the fuel analogy, talking about engine fuel consumption rates in the absence of a time multiplier is only useful for determining fuel line diameters, pump size, fuel filter size, etc.

Just to flog you further on this, if you are using amperes as a measure of electrical efficiency (e.g. this instrument draws 0.5 amps when in operation), you should always be thinking about it in the context of time. That is, if my new boat toy draws 0.5 amps when it's on, it's going to pull 12 amp hours per day and that's 5% of my 240 amp-hour battery capacity.

Hope this helps clarify rather than confuse...

Posted by: SaffyThePook at April 23, 2014 10:08 PM | Reply

Dainis, did you read RC's full review? He was skeptical too.

Norman, I think that most people know that amps per hour means amps used per hour, not amps divided by hours. But, trust me, I regret using the phrase and will try not to use it again. Sorry!

Now, let's please focus on the subject, which is the Smartgauge and how it and other battery monitors can help boaters manage their power and preserve their batteries. Thanks.

Posted by: Ben at April 23, 2014 10:19 PM | Reply

What happened to the TypePad sign-in?

- - -

Short aside regarding Amp-hours:

miles ~ Coulomb (a certain number of electrons, not energy)
miles per hour ~ 1 Coulomb per second = 1 Ampere

Ampere is already like mph, so Amps per hour is redundant and confuses the meaning.

Amps actually is a current, so you would not say Amps per hour any more than you would say knots per hour.

10 mph * 2 hours => 20 miles

Amp-hour means Amps times hours, so it would be like knot-hours, although we just say miles whereas few say Coulombs.

1 A * 1 hour => 1 Ah = 3600 Coulomb
5 A * 2 hours => 10 Ah

- - - -

Back to the story, the shunt type battery monitors should always correctly indicate Amp-hours used or added (a measurement). Where they can be incorrect is relating that to the state of charge of the battery (a calculation).

Did anyone ever test the Battery Bug mentioned on Panbo in 2009?
http://www.panbo.com/archives/2009/03/battery_bugs_new_monitoring_tech.html

Posted by: norse at April 23, 2014 10:38 PM | Reply

In the time it took to get registered again, others jumped in about Amp-hours. Sorry to add to the overkill.

Like Dan, I wondered how much this thing draws itself. It is always on. From Balmar, there is a sleep mode. From RC, 5 mA, which is less than 1 Ah per week (since this was explained above, everyone knows this one is real).

RC thinks it just measures the voltage discharge curve very accurately. Awesome.

Posted by: norse at April 23, 2014 11:07 PM | Reply

"the shunt type battery monitors should always correctly indicate Amp-hours used or added..."

Maybe they should, Norse, but they often don't. RC has examples and there's also a link in the entry to my experience with the Link 1000.

Posted by: Ben at April 24, 2014 12:09 AM | Reply

Pity Balmar didn't put some effort into the DESIGN of the gauge: incredibly ugly, only an engineer could love this look.

Posted by: paul g at April 24, 2014 3:49 AM | Reply

Right, what I meant to say was: "the shunt type battery monitors should always correctly *know* Amp-hours used or added..."

They can measure incoming and outgoing. Whether they show it to you, either raw or corrected, depends. The problem is that the high-water and low-water marks drift and after a while they are just guesses, especially with Peukert and all the other details. So the Ah remaining indication and SOC indications are not always good.

If Smartgauge works by accurate rest voltage measurements, I am surprised it only displays voltage to the nearest 0.05 V

Posted by: norse in reply to Ben at April 24, 2014 3:52 AM | Reply

Dainis -- would love to know your trick with the DCM-600. We have the same, and while it works pretty well for a month or maybe even 2 at a time, it then usually requires a reset (hard unless we are motoring for 20+ hours)...unless of course the boat is at a dock and gets fully reset to 100% while at rest. I have found otherwise, the capacity remaining slowly drifts no matter what settings I use.

Ben -- have you sorted your velocity vs. acceleration thing. I think it has been explained above, but would only add, we are really measuring coulombs, which we express as a flowrate (velocity) in amps (similar to knots). To say amps per hour you are effectually saying knots per hour which would imply an acceleration (I increased the speed 1 knot per hour so where I started at 3 knots, I had accelerated to 5 knots 2 hours later). We then add to the complication by measuring energy in amp-hours (i.e. x amps for y hours) instead of talking in coulombs. This would be like saying I traveled 200 knot-hours instead of saying I traveled 200 nautical miles. Confusing....yes!

I still need to read the full review, but I would imagine this product would not work if the boat had some sort low amp/long duration charging system like solar as the batteries would be charging for much of the day, meaning they were not at rest very often. Maybe we just need battery manufacturers to start developing and installing electronic hygrometers into the batteries to report battery state and health (sort of like laptop batteries that have internal circuitry to report charge state and health).

Posted by: Gram Schweikert at April 24, 2014 7:37 AM | Reply

Actually, the super-cheap Nasa Marine BM1 ( http://www.nasamarine.com/proddetail.php?prod=BMN-1_Battery ) battery monitor is able to quite well gage the state of charge by looking at the battery during discharge. You only have to switch on some load to have a couple of Amps running, and in a couple of minutes it will show you the charge level, which is quite accurate to my experiences.
Of course, it does need a shunt, so it is also able to accurately measure Amp-hours.
And I have to admit, it may not be as good as this instrument, but sure looks way better.

Posted by: Jack Lazy at April 24, 2014 7:43 AM | Reply

Balmar did not design the Smartgauge; Balmar distributes the Smartgauge.

Interestingly, a prime market for Smartgauge technology -- which Merlin calls the "Merlin Model Based Method of monitoring" -- is the military and they claim accuracy validation from EnerSys, the "world's largest battery manufacturer."

RC Collins quotes some the EnerSys testing results in his review and here's another mention here: http://goo.gl/PJlIsf

Incidentally, Merlin's shuntless Datacell monitor seems to be a Smartgauge on steroids, but the Datacell II adds shunts and temperature sensors and thus can report on battery State of Health, Time Remaining, etc.

http://www.merlinequipment.com/markets/group.asp?groupid=19

Posted by: Ben in reply to paul g at April 24, 2014 11:12 AM | Reply

OK, I admit I'm stumped.. I figured out how big a hole it needs (3.75" X 2.55") but even the Balmar manual doesn't tell me how it's fastened into that hole. Are there screw holes in the front plate, or does it fasten in some other way? (please don't tell me they're using double-stick tape!)

Posted by: Hartley at April 24, 2014 12:35 PM | Reply

It's light enough that tape might work, but there are holes at each corner for tiny black screws.

Posted by: Ben in reply to Hartley at April 24, 2014 12:44 PM | Reply

Hartley,

The screw holes are pre-spotted (partially drilled) on the back-side of the mounting flange/lip. If you want to mount with screws just finish drilling the holes all the way through to the front. This is so if you wanted to use double sided tape you could, or just continue drilling through for #6 screws.

Ben,

Thanks for the kind words!!

-RC Collins
Compass Marine Inc.

Posted by: Anonymous in reply to Hartley at April 24, 2014 1:08 PM | Reply

Re: A new Mousetrap

Daine,

One thing you can do by just monitoring voltage is to use the internal resistance of the battery as a shunt.
Since you can't directly measure the current you can approximate the current as a percentage of the total capacity by measuring the change in voltage over time.
This can work very well for discharging but tends to get a bit funky when you intermix it with small charge cycles.
I suspect they have other tricks up their sleeves to fine tune their algorithm.
BTW, this is why it doesn't work with Li-ion batteries: the discharge profile is almost flat.

Posted by: Andy Brown in reply to Dainis at April 24, 2014 1:14 PM | Reply

Thanks, RC! (and Ben) - I figured it had to be simple, but I just couldn't see it (and now I know why!)

Posted by: Hartley in reply to Anonymous at April 24, 2014 2:16 PM | Reply

Hartley,

I will take some measurements and pics of the back side of the SGBMU later and add them to the article. Sorry about the oversight..

-RC Collins

Posted by: RC in reply to Hartley at April 24, 2014 2:24 PM | Reply

I can't find the link to the full Compass Marine article, just the "preface." Anyone have the URL to the "read on" part?

Posted by: Kevin at April 24, 2014 7:09 PM | Reply

Kevin, that is the right link, repeated here:

http://www.pbase.com/mainecruising/smart_gauge

But Compass Marine is formatted in an unusual way. What you should see is the Preface in a large block with the Balmar Smartgauge photo at upper left. If you scroll down you should see the next block of text and photo (titled "Traditional Ah Counter") and then the next block and so forth.

BUT if you click on a photo you'll go to a separate page with a larger photo and all the text that was in the block in a different format. To go from there to the next block you either have to go back to the main entry page or advance blocks with the Next/Previous buttons upper right.

So there's two ways to browse through the whole entry but it is a little confusing at first. Good luck!

Posted by: Ben in reply to Kevin at April 24, 2014 7:54 PM | Reply

Thanks...I see now.

Posted by: Kevin at April 24, 2014 8:35 PM | Reply

If one already had a "coulomb counter" style battery monitor, and then installed a SmartGuage, would it then be possible to use the SmartGuage reading to somehow calibrate the coulomb counter?

It seems that it would be technically possible to combine the best features of the two types of monitors together, thereby obtaining the best of both worlds. I suppose the market just isn't deep enough to justify the R&D.

Posted by: Russ at April 25, 2014 3:27 PM | Reply

Russ, the "Datacell II" monitor by Merlin that Ben referred to in his reply of 4/24 @ 11:12 AM appears to combine both technologies. Merlin appears to be the actual manufacturer of the SmartGuage, btw..:-)

Posted by: Hartley at April 25, 2014 4:47 PM | Reply

Very interesting article, as alway here in Panbo!
At the same time great job by Compass Marine!
A question:
Is SGBMU usefull for a sailing boat that has solar panels and wind gen, almost allways charging the house bank, at the same time that there is consumption ?
I think that maybe there is a kind of problem when charging current is present, but I didn't had the time yet to read all the test by CM.

Posted by: Geo at April 25, 2014 5:50 PM | Reply

Thanks Hartley, I overlooked that.

Incidentally, here is a link to a forum comment regarding the SmartGauge by Chris Gibson, who invented it. It gives some intriguing clues to how the SmartGauge works.

http://www.ybw.com/forums/showthread.php?73924-Smartbank-battery-management/page5

(And some of the preceding comments in the thread are remarkable similar to some of the more critical comments in this thread).

Posted by: Russ in reply to Hartley at April 25, 2014 6:35 PM | Reply

So let's say I have a 400 AH (nominal) size bank, but it's a couple years old. If my coulomb counter reads "-150 AH", and my SmartGauge says "50% SOC", could I then deduce that my actual capacity is only 300 AH, and reprogram my coulomb counter accordingly?

If so, doing this once or twice a year might be a way to keep things in sync, assuming all the other nuances of the coulomb counter are programmed correctly.

Posted by: Russ at April 25, 2014 7:09 PM | Reply

Ben I bought one to try in a new to me boat that I'm currently building for my wife and I. It's a 36 Gulfstar trawler. My house bank will be 6 - 8 (TBD) DEKA UNIGY-1 12AVR100ET batteries . I read the reviews and the testing done and I feel this is worth trying.
I'll let everyone know how it performs vs the standard for me which is my Link2000R controller/ monitor.
Bill
MV: Wireless One
Currently Building MV; Wireless Too
1976 36 Gulfstar trawler

Posted by: Bill Lentz at April 26, 2014 7:59 AM | Reply

I was thinking along the same lines, Russ, and hope to try it. And thanks for finding that interesting old YBW forum thread.

Gibson does hint at several ways a shuntless monitor can analyze battery activity, but I'm struck by the emphasis he puts on simplicity for the average boater. He's obviously a complete power geek BUT he spent 10 years providing technical support for Heart Interface and Cruising Equipment. He's comfortable using an amp counter himself -- and can talk about coulombs and joules til the cows come home -- but, like RC, he's had a lot direct experience with how opaque and frustrating battery monitors can be for civilians.

Posted by: Ben in reply to Russ at April 26, 2014 8:04 AM | Reply

Only 5 years to get round to testing it? Glad you like it.

PS. It is indeed Gibson not Gibbons :)

Posted by: Chris Gibson at April 27, 2014 8:24 AM | Reply

Doh! I apologize again, Chris, and should also mention that I actually haven't yet tested the Smartgauge. This entry is about the detailed analysis conducted by RC Collins for his Compass Marine site:

http://www.pbase.com/mainecruising/smart_gauge

But I do plan to use the Smartgauge on my boat this season and hopefully compare it to a Victron BMV or similar. And I can fix blog entries quickly, so the further "Gibbons" embarrassment is erased ;-)

At any rate, thank you for posting here, Chris, and I hope you'll field a question. Have you and Merlin considered a NMEA 2000 version of Smartgauge? I believe it could be a black box that simply outputs State of Charge to PGN 127506 (DC Detailed Status) and the two voltages to 127508 (Battery Status). Most multifunction and instrument displays can show such values these days and some can alarm on them.

Posted by: Ben at April 27, 2014 10:38 AM | Reply

Now you’re talking! I’m all for a black box NMEA 2000 Smartguage or a Smartguage that looks up to date.

My 1996 installed Ample Power EMON looks more sophisticated then the Smartguage. However, I only understood how to use the Ample Power monitor for about twenty minutes after it was installed and then used it only to check battery voltage. It is wildly inaccurate without constant battery capacity updates - something I only did once in 18 years. I’m willing to make the switch to Smartguage, but would like to see the data on N2K. Chris, any possibility?

Posted by: Richard C at April 27, 2014 11:26 AM | Reply

No worries Ben.

I can field certain questions (including yours), but I have to be a bit careful what I say as I no longer own it. As is probably apparent, Merlin Equipment bought the company off me.

NMEA2000 is something we are considering, not for SmartGauge, but for Datacell, but I wouldn't hold your breath as it's relatively low down the priorities list. Having said that, priorities often change.

The thing I find quite amazing is that after all these years, after all the independent testing that has been done (Enersys, UK MoD, US DoD, TARDEC, TACOM, Balmar, Compass Marine...) I still see the odd comment that (paraphrased) reads "I don't understand (or "I don't see") how it works, therefore it can't". "I don't understand" means precisly that and nothing more.

It isn't perfect, nothing is, but it does meet the original design spec which was to solve the major problems with amp hour counters. Those problems really are a major issue to many people.

Posted by: Chris Gbson at April 28, 2014 3:35 AM | Reply

Can SmartGauge be used to measure three battery banks (2 house and 1 starting)?

Can I add a simple switch to turn the battery bank on or off (I presume this would screw the SmartGauge's logic in CONSTANTLY monmitoring batteries?

Let me know

V

Posted by: Vladimir at April 28, 2014 11:26 AM | Reply

Vladimir - I doubt it would work, as it would appear to need to monitor the battery pretty much continuously to ascertain the true state of charge - if you switched it, all it would know is the instantaneous voltage, and there are much cheaper & easier ways to measure that.

Chris - I understand your pain - the "I don't understand it, so it must not work" reaction is awfully common these days - in pretty much every field of endeavor. Ask the folks at Hydrovane, for example..:-)

RC - while you're taking pictures of the backside, run your ruler around it, too - those of us trying to figure out if it will fit will appreciate it! The "installation manual" only tells us how big of a hole it needs..:-(

Posted by: Hartley at April 28, 2014 11:53 AM | Reply

Dimensions are here...

http://smartgauge.co.uk/fitting.html

Battery bank 1 (the house bank) has to remain permanently connected. If it is disconnected, the whole thing falls apart. You can, however, switch battery 2 between different batteries. Battery 2 only reads voltage, and you can't do this if a SmartBank Advanced is installed.

Posted by: Chris Gibson at April 28, 2014 1:14 PM | Reply

Thank You, Chris (and Smartguage!) - Balmar needs to update their manual.

So if I have two house banks - combined for charging, separated (using voltage-driven combiner relays) for discharge, I would really need two of these units - one for each "house bank"? The start battery really shouldn't be discharged except the very short-term load of starting the engine or generator, so it's state of charge isn't as critical (to me).

Posted by: Hartley at April 28, 2014 2:15 PM | Reply

OK - and having spent the last 30 minutes or so browsing - anyone with an interest in the subject of boat electrical systems should cruise on over to http://smartgauge.co.uk/technical1.html and read what they've got there. Yes, it's written for Europeans, with a few words that us Yanks might have to think about, but they have a LOT of very useful info there - including answering a lot of "why?" questions I bet all of us have had.

So - Thanks Again, Chris!

Posted by: Hartley at April 28, 2014 2:47 PM | Reply

We looked at the SmartGauge when we compared 15 different battery monitors for our internal benchmarking. The SmartGauge's marketing materials sounded great, but I couldn't find a single sales guy at Balmar or Merlin who could explain (even basically) how it worked without a shunt. While I suspect it is some sort of ESR meter, no one could confirm or deny that. As a dealer, we like to have a solid footing regarding the technology, so we can then explain it. After an exhaustive search, we ended up with the BM Pro, because we really liked the idea that it has a shunt (so we could see incoming Amps from the wind generator), but is also self-learning and gives SOC, and not Amp-Hours, which are meaningless. There is no synchronizing necessary, as its accuracy improves as it learns. It also has a fantastic and large back lit display, along with a USB charger port, and now comes with a Bluetooth link to your iPhone/Android.

Posted by: Jeff Field at April 28, 2014 4:29 PM | Reply

Thanks, Jeff, the BM Pro sounds interesting. I found some information about it at your site and also at Sentec's store:

http://store.marinebeam.com/battery-system-monitor.html

https://teambmpro.com/au/drifter-bm.html

Can you point us to a manual that shows more about how it installs and works?

Posted by: Ben in reply to Jeff Field at April 28, 2014 5:54 PM | Reply

Ben, thank you for bringing up the NMEA issue. But don't Maretron have a product that already bring these goodies? It even have a more sdvanced method of counting the coulombs than through a shunt. They use a coil so that the installer don't have to break into the wiring.

I wonder if they already have figured out hove to combine the counting of coulombs with the information provided by the voltage measurement?

Keep up the good work!

Posted by: Jorgen at April 29, 2014 1:30 PM | Reply

Great report by RC. Thank you Ben for posting it.

Posted by: Ed K at April 30, 2014 10:09 PM | Reply

Does anyone can figure out if the continuous charging that comes from solar panels, during day time, confuses S.G.B.M readings?
RC is writing '...I did see the Smart Gauge get a bit confused when the bank was being charged. It can't really track the capacity of a battery charger now can it...?'
But what's that means when charging is always present from solar panels?

Posted by: geo at May 1, 2014 5:11 AM | Reply

Unless you live on the sun, a bit warm for my liking, solar is daylight hours only. A once or twice per day SOC reading should be more than enough for adequately monitoring SOC and managing your bank in the healthy range.. You can simply check before bed and when you wake... After the SGBMU has learned the bank it is pretty quick to refind accurate SOC once the bank is back under load.

Even during charging the most I saw it off by was approx 10-12%, which is still far better than perhaps 95% of the Ah counters in use out there, in regards to the SOC screen.. I offer battery capacity testing for my customers, as a service, and I have yet to see more than 1 or 2 used batteries line up with "rated" capacity better than a 5% variation...

Posted by: Compass Marine in reply to geo at May 1, 2014 8:51 AM | Reply

I wasn't aware of any solar system that charged when the sun is out of view? Perhaps if you combine that with wind then there might be an issue? I hate to state the obvious.
Bill Lentz
1993 Mainship 40 Sedan Bridge
Currently building a 1976 Gulfstar Trawler.
Home ports Little Egg, NJ
Middle River, Maryland

Posted by: Bill Lentz in reply to geo at May 1, 2014 9:07 AM | Reply

Jeff, I would argue with your assertion that Amp-hours are meaningless. In fact, I would argue that they're generally more useful than SOC.

Monitoring Amps and Amp-hours gives you a quantifiable sense of how much your various electrical consumers on board are using. For instance, I've retrofitted your LED bulbs throughout my boat and measured what my consumption is with all my lights burning, so I now know how long I can stay on the hook and lit. Is it time to re-insulate the old ice-box or is my time and money better spent elsewhere? All I have to do is take a look at my refrigerator's electrical draw to see how much it's using. Etc, etc.

Some of us are very conservative about how much power we use before charging (e.g. no more than 30% SOC, even if you downrate your bank's capacity with age) and/or have wind/solar quietly helping out while drawing the banks down. For folks like that, it can be more useful to have the Amp & Amp-hour info.

Posted by: SaffyThePook at May 1, 2014 9:41 AM | Reply

We will correct the information in the manual.

Posted by: BALMAR in reply to Hartley at May 2, 2014 1:42 PM | Reply

I used Smartgauge and Smartbank on my canal boat in England for years. The accuracy of the algorithm that Chris developed is astonishing. The only exception is during charging as Smartgauge can only read the voltage being put into the battery. At that point a weaker algorithm is used, but when the charging is done, the full accuracy comes back. I highly recommend it. Realize that this is a tool to measure SOC, not a tool to measure amphours. I've met Chris and he's a brilliant fellow (and a nice guy too!)

Posted by: Dov Lazar at May 15, 2014 11:57 AM | Reply

Dov nice to see you posting here. I finally closed on the Gulfstar and expect to install the BalMar SOC meter within the next 2 weeks. In addition I will have a conventional negative buss AH's counter. I think they both have a place for those who are so inclined.
My main concern is I will more than likely always have some sort of charging source be it the AIR-X wind generator or the 600 to 800 watts of solar. In addition the ProMariner 2000 PS inverter / 70 amp charger. My Westerbeke can also output up to 50 amps of DC in addition to it's AC output. When I'm on the hook at night and stop the wind generator will probably give me the true benefit of checking the house banks state of charge.
Bill Lentz
Wireless Too
1976 Gulfstar 36 MKII trawler
currently in the Middle River, Maryland eventually it will be in Little Egg, NJ.

Posted by: Bill Lentz at May 21, 2014 4:10 PM | Reply

After installing and using the Smartgauge I have to say that everything said in this article about accuracy and function is true. I don’t disagree with the outcome of testing by Compass Marine, however, for those thinking about buying the Smartgauge you should be aware of or at least look closely at a few other considerations. For one the programing regimen is frozen in early 1990’s technology with cryptic display abbreviations and button pushes to do the setup. If you want to take full advantage of the very useful features, this can get frustrating quickly. Not making this programing task any easier is the poor quality of the button pad. Maybe I have a defective unit, but the amount of pressure needed to register a selection is out of the ordinary, so much so that I almost gave up pressing my thumb against the pad thinking I was about to break the surface when low and behold it did register the button push. Getting back into programing mode is hit or miss trying to push two buttons simultaneously.

There is no reason, with todays technology, that a front mount USB connection can’t be implemented allowing all programing to be input or modified on a laptop. There is no reason why a reasonable size display can’t be used so the cryptic abbreviations can be made into real words so anyone onboard can read error messages or programing without a cheat sheet.

This is a terrific device and works great, but think primitive before you decide to buy.

Posted by: Richard C at June 30, 2014 7:29 AM | Reply

This device is certainly good for the battery makers! I installed one and it consistently shows a SOC at least 20% less than my old meter, which is clearly calibrated to the battery's specifications rather than their current capacity.

It looks like I need a new bank, which surely does not make me happy. The good news is that it will keep the new bank in good shape.

Posted by: Steve at August 23, 2014 2:51 PM | Reply

As far as I know you need SOC info to know when to stop discharging or to stop charging. The middle ground isn't 'that' interesting or even really critical. The manual says don't use it to determine when the batteries are full (!?) and low ish states of charge can be determined with simple battery voltage observations - e.g. 11V.

How does one know how accurate it is? What do you compare it to?

Posted by: Bob at August 26, 2014 10:15 AM | Reply

I don't understand, Bob. Isn't RC's incredibly detailed testing and write-up all about how accurate and useful the Smart Gauge SOC is?

http://www.pbase.com/mainecruising/smart_gauge

Posted by: Ben in reply to Bob at August 26, 2014 10:52 AM | Reply

Hi Ben,

Yes, the test write up is very informative. I'm really impressed with the Smartgauge actually and I'd like to get one of these for my significant other to look at :) I'm a bit of a techno-geek-engineer so of course I don't mind the drifty Ah counters because I understand enough about them to kind of know how they work and where the limitations are - hence my comments above.

Battery SOC is an interesting topic. So is Peukert's. Check out figure one in the following:

http://www.absoluteefficiency.com/LEAF/Peukert_capacity_leadacid_lithium.pdf

In red line shows the Peukert reduced battery capacity at a 50A discharge current. The green line shows capacity at 5A. The redline drops to 10.0V more rapidly as one might expect from the so-called Peukert effect thus defining an empty battery at 0% SOC. The green 5A discharge line gets all the way up to 67Ah or so.

But look at the reality of the so-called discharged battery. The 50A discharge results in 45Ah removed before a drop to 10V but when the tester waits 6 hours all allows the battery to settle there is ANOTHER 20Ah that can be removed from the battery at the same 50A!!! So what was the prior SOC? It certainly wasn't empty. I'd be interested to know how Smartgauge would deal with that test - hence my question above.

With a C/20 Ah rated battery it's *difficult* to get rated capacity out at higher currents due to chemical reaction speed limitations. But the C/20 rating does a reasonable job of describing the absolute ability of the battery to store energy - even if you can't get it 'all at once' with a high-rate discharge. In this way some of these Peukert compensated Ah meters are a bit misleading. That 65Ah battery in Figure one ultimately supplied 65Ah at C/~1 and 67Ah at C/~13. This is not what 'Peukert's' would say as far as traditional Ah meters go.

Everything is 'easy' if one only considers repeated constant current charge and discharges. But a) this isn't how it is with real load in real life b) batteries behave very differently with real load profiles. This is a major challenge for battery monitors.

The test you linked to doesn't obviously evaluate Smartgauge in any kind of real life variable or stepped loading conditions. The interesting and potentially challenging results are not shown so it's very hard to know how accurate it is. If the loads are constant it's easy to be really accurate but if the loads change, as they do in real life, it's really difficult to be accurate. The test doesn't provide insight on this. In my limited opinion that is.

Posted by: Bob at August 27, 2014 2:46 PM | Reply

Leave a comment