Accurate(ish) battery capacity test using a smartguage?

[GHA]

A few things first - very few liveaboard cruisers have any idea about the health of their battery bank. It's either new, seems OK.... or dead..

Must be next to no-one who has actually done an accurate capacity test. (Can think of one )
Smartgages are very accurate on discharge, amp counting battery meters generally are good as well, it's when they don't get to 100% that they drift.

So.... with a smartguage and a battery meter get the batts to really full charge and make sure the meter has reset to 100%, thn overnight turn off any charging sources and next morning look at the values. Say SG read 75% and AH used = 50Ah then your batts are about 200Ah.

Anyone tried it.? (Guessing the answer will be no )

But for 99quid just ordered a SG so will be fun to try and find out

 

[Poey50]

A few things first - very few liveaboard cruisers have any idea about the health of their battery bank. It's either new, seems OK.... or dead..

Must be next to no-one who has actually done an accurate capacity test. (Can think of one )
Smartgages are very accurate on discharge, amp counting battery meters generally are good as well, it's when they don't get to 100% that they drift.

So.... with a smartguage and a battery meter get the batts to really full charge and make sure the meter has reset to 100%, thn overnight turn off any charging sources and next morning look at the values. Say SG read 75% and AH used = 50Ah then your batts are about 200Ah.

Anyone tried it.? (Guessing the answer will be no )

But for 99quid just ordered a SG so will be fun to try and find out

That sounds a good idea, especially as I have a Smartgauge and a Victron BMV-700. I take your point about getting the batteries fully charged to start with. As others have found the Smartgauge does tend to show 100% a bit too ambitiously.

 

[pcatterall]

Dont know about an 'accurate capacity test' but we do an anuall practical test by loading our batteries with a known discharge ( 10amps) and observing voltage deplation and the readings on our monitor.
It may not be 'scientific' but it gives us confidence that the batteries will provide the capacity we need when we need it and that our monitor is giving us a fair interpretation of the battery state.
We would do this if setting off on a long ( several day) passage.

 

[vas]

A few things first - very few liveaboard cruisers have any idea about the health of their battery bank. It's either new, seems OK.... or dead..

Must be next to no-one who has actually done an accurate capacity test. (Can think of one )
Smartgages are very accurate on discharge, amp counting battery meters generally are good as well, it's when they don't get to 100% that they drift.

So.... with a smartguage and a battery meter get the batts to really full charge and make sure the meter has reset to 100%, thn overnight turn off any charging sources and next morning look at the values. Say SG read 75% and AH used = 50Ah then your batts are about 200Ah.

Anyone tried it.? (Guessing the answer will be no )

But for 99quid just ordered a SG so will be fun to try and find out

what is a smartgauge then GHA? any links?
I also do have a BMV-700 installed and do follow what it says (considering all values as estimates)

V.

 

[Mistroma]

what is a smartgauge then GHA? any links?
I also do have a BMV-700 installed and do follow what it says (considering all values as estimates)

V.

Completely different beast from the normal Coulomb counters which monitor Amps & Watts in and out.

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

Plenty of places stocking it and it used to be around £170 so £99 is a good price.

Nothing is perfect and it does have drawbacks but simple to install with no shunt (MUST have direct connection to battery posts). I've had one since 2012 and it works very well, at least until a power surge threw it off in 2017. It had it replaced in 2018 and it is back to normal.

It doesn't give an accurate state of charge (SOC) when charging and uses a pretty simple standard Voltmeter estimate of SOC. This tends to make it optimistic and read 100% too soon. The main effect is to show a pretty sudden drop within an hour or so on discharge. You get used to this and expect 100% to suddenly become 90% when the sun goes down. It is accurate on discharge and is a good indicator of battery state until solar, alternator etc. kicks in again. My solar usually gets the battery back to 100% each day in mid-summer so that I don't get the sudden drop. That only happens when batteries haven't been fully charged.

 

[GHA]

what is a smartgauge then GHA? any links?
I also do have a BMV-700 installed and do follow what it says (considering all values as estimates)

V.

Review here =>
https://marinehowto.com/smartgauge-battery-monitoring-unit/
Balmor have a new unit out which is maybe why the old one is so cheap
http://www.balmar.net/sg200-battery-monitor/

Ebay listing -
https://www.ebay.co.uk/itm/Merlin-E...285e7ce7ef:g:CxAAAOSwJtJcSenx&redirect=mobile

 

[SteveNotlob]

I don't use a smart meter but i do check the batteries on our cars and boat every few months with one of these :-
https://www.ebay.co.uk/itm/Sealey-D...m33f870a0f2:g:RwsAAOSwQZdb7XWG&frcectupt=true
The battery on my Kia starting to show below 60% consistently, even after a long run with no lights, A/C, etc so I took it to the local dealers to find if it was battery or charging issue.
They put their expensive all-singing-and-dancing tester on to the battery and the difference between their readings and mine with this device was just 3% - I can live with that

 

[laika]

Isn''t the problem with the smartgauge approach that it's measuring the battery *bank*? When you've got 4 or 5 of them and you suspect one of them is duff that's a lot of moving batteries and cables around to test them all individually in turn, or have I missed the point?

I ask because I think I've got a duff battery and was considering asking for recommendations for battery load testers, so interested to see SteveNotlob's post

 

[lw395]

I think, if you want to test a battery, it's best to stick to really basic methods where you can understand exactly what the meter is telling you.
But you need to be clear what you are trying to find out.
For a general health test, I suggest:


Charge fully with a known good charger. In my case that means leave it overnight at least on the Ctek.
For a start battery, either start the engine and check the minimum that the battery dips to, or use a high current test. For my motorbike battery I would use a shunt that gives 100A for 20seconds and if it stays above 10V the battery is deffo good.
Good for starting that is, I've known batteries pass this test despite the capacity having dropped a lot from new.

For a house battery, run a (e.g.) 10A load for an hour and see what the voltage drops to. Compare with your favourite graph. If one battery in a bank is really sick, a 10 minute test might show it!

When a test box is telling me 55%, what exactly does it mean by that? and how can I know whether that battery will do what I ask of it?

With modern cars, smart alternators are doing all sorts of wierd things like not charging when you accelerate or go up hill. They may almost never fully charge the battery. Cars are getting less and less like boats, so be wary of comparing test methods.

 

[GHA]

Isn''t the problem with the smartgauge approach that it's measuring the battery *bank*? When you've got 4 or 5 of them and you suspect one of them is duff that's a lot of moving batteries and cables around to test them all individually in turn, or have I missed the point?

I ask because I think I've got a duff battery and was considering asking for recommendations for battery load testers, so interested to see SteveNotlob's post

AFAIU they will give an accurate indication of SOC on discharge within a few cycles so probably not much use to spot an individual duffer in a bank without lots of logging, which wouldn't be too tricky with a couple of jumper cables and lots of time, just do one at a time. Lots of time Anyway it will be interesting to have a play and in the boat rules I'm allowed to spend under 100 quid on toys now and again Thinking more of cruisers where any kind of accurate testing never really happens beyond seeing if the lights go dim overnight with everything left on And getting a decent handle on health would be gold dust info before another ocean passage, when they die they seem to die quick. Been there ..


Might be interesting trying to somehow get a better guess if the puerkert number from the datasheet seems still valid though that might be asking a bit much.... fun to try though

 

[vas]

Review here =>
https://marinehowto.com/smartgauge-battery-monitoring-unit/
Balmor have a new unit out which is maybe why the old one is so cheap
http://www.balmar.net/sg200-battery-monitor/

Ebay listing -
https://www.ebay.co.uk/itm/Merlin-E...285e7ce7ef:g:CxAAAOSwJtJcSenx&redirect=mobile

sorry I'll pass,

reading the review and slowly loosing any will to live...
600W solar unobstructed from mast/boom/whatnot, and 4X T105RE under the Med sun with low consumption, I'll keep on looking at the stats given from the victron mppt 30/100 and the correctly wired BMV-700 shunt and be done with I think.

cheers

V.

PS. wouldn't mind a way to route the BMV data to the N2K bus, but I'll probably mark it as next winter project...

 

[LadyInBed]

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

Plenty of places stocking it and it used to be around £170 so £99 is a good price.

Quote from website

Web site and all contents Copyright SmartGauge Electronics 2005, 2006, 2007, 2008, 2009. All rights reserved.
Page last updated 17/02/2009.
Website best viewed on a computer of some sort.

 

[GHA]

PS. wouldn't mind a way to route the BMV data to the N2K bus, but I'll probably mark it as next winter project...

Signalk has an app to talk to victron venus so should be able to do that on a raspberry pi. Then send to N2k, to the web, save the data and plot, whatever you want to do with it

 

[Mistroma]

Quote from website

Web site and all contents Copyright SmartGauge Electronics 2005, 2006, 2007, 2008, 2009. All rights reserved.
Page last updated 17/02/2009.
Website best viewed on a computer of some sort.

Hilarious. I had that link at hand because it had a lot of useful technical information. I also found the the tips and tricks section very useful for a few very low cost mods. (around 50p to a few pounds).

i.e.
1) I have variable speed forced battery ventilation for cost of some ducting (I already had a computer fan and a diode).

2) I can disconnect engine & bowthruster AGMs at a lower voltage and fully charge T105's at 14.9V. All three banks are disconnected during discharge as usual.

3) Slight mod. to improve alternator charging (some time and a diode required).

I had forgotten that SmartGauge had been sold on quite a few years ago and they wouldn't have been updating the original site afterwards.

 

[sailinglegend420]

....Say SG read 75% and AH used = 50Ah then your batts are about 200Ah.....

Thanks to GHA for posting this. For the last 4 years I have also been estimating how much actual capacity is left in my battery bank, but I fear the SG has gone way over the heads of most people who don't seem to understand what it is doing. It is like a car fuel gauge measuring how full your tank is but it tells you how full your batteries have been charged.

I have been using one for 13 years and - yes it's 13 years old now and still not used by that many people because it appears it can't possibly measure accurately you battery State of Charge (SoC.)

......It doesn't give an accurate state of charge (SOC) when charging and uses a pretty simple standard Voltmeter estimate of SOC. .....

This proves my point! Yes it measures voltage - but over 1000 times a second - so this way you can estimate the current without a shunt by clever mathematics, and so calculate the SoC.

It does a lot more than just measuring voltage and uses - as far as we can guess - active impedance compensation. Feed this into a complex algorithm and you can come up with very accurate SoC readings that gets more accurate as it learns about your battery. That is why the military and service vehicles changed to this because of errors with their shunt based systems that get more and more inaccurate as the batteries get older.

I don't use a smart meter but i do check the batteries on our cars and boat every few months with one of these :-
https://www.ebay.co.uk/itm/Sealey-D...m33f870a0f2:g:RwsAAOSwQZdb7XWG&frcectupt=true
....

These 'load testers' are just designed for starter batteries not house batteries - don't waste your money on them for your boat.

*************************

My original Smartgauge and the Ah lost over night did suggest that my battery State of Health (SoH), or actual battery capacity, was getting less. We noticed that we couldn't spend as much time at anchor as we used to before the battery volts got down to 12.2v - approximately 50% discharged.

My calculations using the Smartgauge, at a known accurate 100% SoC after a couple of nights on shorepower, showed the Ah withdrawn overnight suggested my battery SoH was about 60% - that meant my bank had lost 40% of its original capacity. Because I had a very large bank of Lifeline AGMs I(1050Ah) I was still able to keep using them for another two years before I changed them at 52% SoH - and yes I did a full capacity test on each of them with a load of 10 amps for 20 hours and after about 11 hours they dropped quickly down to 10.5v.

Batteries should normally be replaced when their SoH gets below 80%, but Lifeline say you can take their AGMs down to 50% - another reason my bank lasted 14 years.

Balmar have just produced the SG200 which contains their version of Smartgauge and a shunt based Ah counter which will give you very accurate SoC and SoH readings.

 

[Mistroma]

This proves my point! Yes it measures voltage - but over 1000 times a second - so this way you can estimate the current without a shunt by clever mathematics, and so calculate the SoC.

It does a lot more than just measuring voltage and uses - as far as we can guess - active impedance compensation. Feed this into a complex algorithm and you can come up with very accurate SoC readings that gets more accurate as it learns about your battery. That is why the military and service vehicles changed to this because of errors with their shunt based systems that get more and more inaccurate as the batteries get older.

That is correct when discharging. I'm certain that I was told that it uses a much simpler method when charging and is pretty much related to a lookup table of Voltage. Not quite that simple but their technical support people indicated as much and that it wouldn't be accurate when charging. This tied up with the data I'd collected and explained the sudden drop from 100% when my solar array was much smaller. The batteries weren't getting to 100% but SmartGauge showed 100% (SG said more like 90%-95% when I checked).

I think that their technical people will confirm this behaviour if you ask that specific question. Smartgauge isn't as good at giving accurate readings during some stages of charging (i.e. top end of scale).

Accuracy improves once discharging begins and SmartGauge uses a much more complex method of estimating state of charge. I've never considered this to be a big problem. I get accurate readings on most days as my batteries do really reach 100%. However, I am not surprised by an initial sudden drop from 100% on cloudy days.

The only real problem I had was probably caused by a Voltage spike. SmartGauge technical support couldn't make sense of the fault initially and said:

"In general, if the Smartgauge had developed a fault then there would be discrepancies in the voltage reading also, not just the SoC reading. So in essence, you have provided me with a very difficult scenario to try and figure out!

I was able to supply quite a lot of detailed data which showed a step change in accuracy of state of charge (SOC) under discharge. They were keen to get hold of the unit to confirm my SG, Voltage and SOC readings as they had never seen that particular failure mode.

The fault was confirmed and I think that some internal registers had been corrupted. I now have the latest version PCB in the original casing at modest cost.

SmartGauge were always tight-lipped about exactly how it works. Many people claimed SmartGauge was snake oil and couldn't work at the time I was thinking of buying one. However, I did find a whitepaper about 9-10 years ago and I seem to remember it covering HF pulsed measurements and use of Nyquist plots to develop data on battery state. It was enough to persuade me that it was worth taking seriously.

 

[sailinglegend420]

.....I think that their technical people will confirm this behaviour if you ask that specific question. Smartgauge isn't as good at giving accurate readings during some stages of charging (i.e. top end of scale).....

I agree that at the top end it is not as accurate when charging, it tells you this in the manual, but it falls from an accuracy of 1% to about 5%, which is very much more accurate that any shunt-based unit on the market. How accurate do you want it to be anyway?

This was all confirmed to me many years ago by the MD James Hortop who I know very well He used to be Technical Editor at PBO. He bought the product from Gibbo the designer because his service clients (Fire, police, ambulance etc.,) wanted a more accurate Battery Monitor that was easier to fit than shunt- based monitors, and one that would never give incorrect readings because the shunt had been wrongly installed or other devices were wrongly installed, which bypassed the shunt. This happens on a very large number of shunt-based monitors that I have seen as a live aboard in the Med for 13 years. This means they will never be accurate at any SoC or charging situation.

All batteries should be bought back to a real 100% charge at least every 10 days to avoid a Partial State of Charge situation (PSOC) that causes permanent sulfation which cannot be removed by equalising. This is the main cause of a reduction in battery capacity over time. 100% full doesn't mean when a little green light has come - that may only tell you the batteries have gone to float. Float does NOT mean the batteries are fully charged.

There is only one simple way to accurately measure when batteries are 100% charged and that is when the current going in is 0.5% of their Ah capacity C - at their absorption voltage. For a 400 AH bank this would be 2 amps at the absorption voltage of at least 14.4v. The problem is this voltage will never be seen as the regulator will already have dropped to a float mode and be charging at a much lower rate. So a simple trick is to switch off the regulator and then switch it back on again. If the current going into the battery increases as the regulator goes back to the higher absorption voltage then the batteries were not "fully charged". Leave the charger on and watch if the current falls to 0.5%C. It might again fall back too early to float so this might have to be done several times before the current going in gets closer to 0.5% of C.

 

[GHA]

Smartgauge arrived & I've just started having a play around. So Peukerts, I'm assuming battery monitors (ancient BEP in my case) display actual Ah used then use peukets number/equation to calculate percentage capacity used from the battery capacity programmed in?

Once the LX is put back together I could probably check this as a trusty Raspberry Pi saves volts/amps but not for a little while. Just messing about on a little 12v battery at the moment.

And anyone know, does the peukert number change with battery age?

mA out of a little battery with an ESP32 load which also measures current in/out. SG said 59% at end of discharge but that's just first cycle so prob hasn't had time to sync yet.