Battery Monitor Reviews Wanted

[endoftheroad]

noelex,
I was on the wrong track thanks for straightening me out.

[Maine Sail]

Originally Posted by endoftheroad

noelex,
I was on the wrong track thanks for straightening me out.

It is simple to do a manual re-set:

#1 know your banks 20 hour Ah capacity, eg: 400Ah

#2 know what 2% of that is, eg: 8 amps

#3 Run the motor for about 5 minutes to clear any "in-rush" current.

#4 Make sure all DC/HOUSE loads are turned OFF

#5 Check the voltage screen to be sure your batteries are at 14.2 - 14.6V or so.

#6 Read the AMPS/CURRENT screen on your battery monitor

#7 If bank is taking less than or equal to 8A of charge current it is actually full enough to do a manual reset.

#8 It is important that this be done at your absorption voltage or above 14.2V when you make the reading. Float voltages can result in you seeing equal to or less than 2% before the bank is actually full...


Please keep in mind that just because a solar, wind, battery charger or alt regulator has gone into "float" mode that the batteries may not be full at all. Many of these charge algorithms are simply timed and do not know the size of your bank. Ideally your charge devices would not go into float until the bank was taking less than 2%, but this rarely happens.....

So, simply check your "accepted current" at absorption voltage, with no loads on, by using the engine alt... You could use a battery charger too...


Please keep in mind that 2% is not technically "full" but full enough on a cruising boat that you'll usually never get much beyond there. A full healthy battery bank will take well under 2% in charge current when truly 100% full..

[goboatingnow]

As of yet my emails to Smart Gauge have gone unanswered as to whether this technology will work on LiFePO4. I would expect that it could but it instills little confidence when the company won't even answer an email with a simple yes or no answer.. Balmar, the US Importer, does not know the answer either...

SmartGauge as I understand it, ( and its the only way it can work) attempts to apply the dynamic voltage model versus SOC to a LA battery , The common wisdom is that battery voltage in a dynamic LA situation does not represent SOC, Thats only true at the simplest level, ie that there isnt a linear relationship

ie 2xSOC <> 2x f(V), where F(V) is a simple linear or near linear transfer function

however there is complex transfer function that does map battery voltage onto SOC and smartgauge attempts to model this, Now accurate it is nobody knows, but in theory that method is technically much better then current shunts.


Given the linear relationship in Li , It is entirely appropriate to use voltage as an indication of charge state. A current counter Monitor isnt needed other then to record charging parameters) in Li systems, as is witness by the many Li management ICs that just report charge vis cell voltage.

dave

[Paul L]

Originally Posted by goboatingnow

SmartGauge as I understand it,......
ie 2xSOC <> 2x f(V), where F(V) is a simple linear or near linear transfer function

however there is complex transfer function that does map battery voltage onto SOC and smartgauge attempts to model this, Now accurate it is nobody knows, but in theory that method is technically much better then current shunts.
.....

Dave,
So how is that going to work when you have a continuing variable load(boat load), along with continuous charge (solar) on a cruising boat?

[goboatingnow]

So how is that going to work when you have a continuing variable load(boat load), along with continuous charge (solar) on a cruising boat?

The terminal voltage of a large battery bank is almost totally determined by the battery voltage rather then any charge sources.

If you look at the research papers http://facta.junis.ni.ac.rs/eae/fu2k72/5mischie.pdf

Youll see that for a given load ( ie discharge) or charge source , there is a voltage relationship.

The trick is devising a software algorithm that models the curves with sufficient accuracy and also that the model copes with different battery capacities, etc

Hence the model accommodates a variable discharge load and variable re charge sources.

dave

[ebaugh]

YMMV, but SOC on FLA can be pretty accurately measured by bank voltage when the bank is at rest, or close to at rest. 12.8V is 100% and 12.0 is 50%. It's almost linear. You do need to top off weekly and check the specific gravity once a month and equalize if the cells get out of whack. Old worn out batteries seem to be about .1V less than this.

For LFP, I'm planning to try and create both a voltage and current counter meter in conjunction with a battery monitor targeted for lithium. But there is very little voltage change in the 20-80% SOC range, so you need a very accurate voltmeter. Plus I don't know yet if a "new" cell will have different characteristics than an older cell. Only a slight change will make voltage alone pretty useless. I think LFP benefits from, and needs some sort of Ah counting meter as a guide.

But no monitor on any chemistry is really going to be able to have absolute accuracy if not "reset" occasionally at "full" charge since any tiny error, and they are impossible to eliminate, gets amplified over time.

[Maine Sail]

Originally Posted by ebaugh

But no monitor on any chemistry is really going to be able to have absolute accuracy if not "reset" occasionally at "full" charge since any tiny error, and they are impossible to eliminate, gets amplified over time.

Well said..! Even with our LiFePO4 bank and a Peukert programmed for 1.01 I still do manual resets when the bank is "full". Force of habit I guess... Thus far, at a 1.01 Peukert, it has been spot on every time I have re-set it. Even the automatic re-sets have been accurate.

Having used battery monitors for over 20 years I have never relied on them automatically re-setting. In my experience they do a very poor job of this especially when you introduce solar or wind (with our LiFePO4 bank solar is turned off when we are not there).... Manual re-sets are easy...

I have one customer who was recently reading -162 Ah's on a 80Ah starting battery. Since that can't physically happen I still stand by; "battery monitors are only as smart as the people wiring, using and programming them."

[Sandero]

Originally Posted by Maine Sail

Since that can't physically happen I still stand by; "battery monitors are only as smart as the people wiring, using and programming them."

My problem in a nut shell!

[goboatingnow]

Technically voltage monitors like Smartgauge dont need resets to remain accurate BTW ( they dont monitor current anyway).

Current monitors need resets because of the fundamental issues with current monitoring, ie the inability to differentiate what the sources of current flows are from.

In reality a good battery monitor should separately measure charge currents and load currents and therefore could better tell what the battery is actually receiving, this could then auto reset on full far more accurately.

dave

[Paul L]

Originally Posted by goboatingnow

The terminal voltage of a large battery bank is almost totally determined by the battery voltage rather then any charge sources.

If you look at the research papers http://facta.junis.ni.ac.rs/eae/fu2k72/5mischie.pdf

Youll see that for a given load ( ie discharge) or charge source , there is a voltage relationship.

The trick is devising a software algorithm that models the curves with sufficient accuracy and also that the model copes with different battery capacities, etc

Hence the model accommodates a variable discharge load and variable re charge sources.

dave

If I read this paper correctly, this applies to a fixed load or charge currents. Have you tried the Smartgauge? Does it work on a cruising boat type situation?

[goboatingnow]

Originally Posted by Paul L

If I read this paper correctly, this applies to a fixed load or charge currents. Have you tried the Smartgauge? Does it work on a cruising boat type situation?

Yes it does, but it "seems" accurate but Ive no proper data.

Why I referenced the research paper, is to show you that using advanced battery modelling you can interpret SOC ( which is what you want to know) from terminal voltages , even in a dynamic situation

heres a simpler description of using voltage in a charge only situation http://www.scubaengineer.com/documen...ing_graphs.pdf

Heres a IEEE paper with a bit more depth for those so inclined

http://www.nuigalway.ie/power_electr...onics_2007.pdf

dave

[sailorboy1]

I asked a similar question a month ago about BM and SOC restting when using a solar panel. My Victron seems to be working well with a setting of 14.5 volts and a 2% current level. While it is possible to get the high voltage and low current if a cloud when over at just the right time, from what I'm been seeing it pretty much appears to have gotten pretty close to charged once the voltage even got to that level.

[Paul Elliott]

(Re: How can "Smartgauge" accurately determine battery State of Charge with no current monitoring?)

Originally Posted by goboatingnow

The terminal voltage of a large battery bank is almost totally determined by the battery voltage rather then any charge sources.

If you look at the research papers http://facta.junis.ni.ac.rs/eae/fu2k72/5mischie.pdf

Youll see that for a given load ( ie discharge) or charge source , there is a voltage relationship.

The trick is devising a software algorithm that models the curves with sufficient accuracy and also that the model copes with different battery capacities, etc

Hence the model accommodates a variable discharge load and variable re charge sources.

dave

But this paper (and the ones you reference later) show that both battery voltage and current are needed to determine State of Charge. Terminal voltage alone is not sufficient, unless you somehow know, without measuring, the charge or discharge current.

Elsewhere you say:

Originally Posted by goboatingnow

Why I referenced the research paper, is to show you that using advanced battery modelling you can interpret SOC ( which is what you want to know) from terminal voltages , even in a dynamic situation

Again, the papers tell me that the terminal voltage will depend on both SOC and current. And not just instantaneous current, but also the current averaged over some duration. The models in these and other papers use multiple layers of capacitors and series/shunt resistors to mimic this dynamic battery behavior.

So, I remain puzzled as to how accurate the Smartgauge can be without measuring current.

[goboatingnow]

There is a documented, understandable relationship between battery voltage and SOC, even in dynamic situations, ie , where the discharge rate changes and there is a switching between discharge and recharge cycles.

The terminal voltage is an indication of discharge current, its just that deriving SOC from that isnt straightforward.

SmartGauge from what I can see, uses an iterative approach, building discharge maps to refine the underlying algorithm

dave

[secrabtree]

From my reading of the SmartGuage info it appears that the definition of state of charge is the ratio between the amount of charge (energy?) in the battery compared with the amount of charge (energy?) that the battery could hold. This is an interesting number to me, but only out of curiosity. What I would like to know is the amount of energy that the battery holds and that is a completely different number than the ratio between what exists and what is possible. I haven't yet found any instrument that can instantaneously tell that. Battery voltage while at rest gives a fair indication, but that isn't perfect. The "percent of charge" reading of an Amp hour counter may not be an accurate reading of the percentage of charge, but it seems just as accurate of an indication of the amount of energy still available as a true "state of charge" meter like the SmartGuage. A new 100Ahr battery has a certain amount of available energy. At an SOC of 75% it has about 50%, realistically, of the original available energy. But as it gets older and, let's say, realistically has the capacity of a 50Ahr battery at a 100% SOC it will only have the same available energy as a new 100Ahr battery at 75% SOC. So SOC is a meaningless number unless one knows what the available amount of energy is when the SOC = 100%. And if one knows that then an Amp-hour counter will be perfectly fine.

In my case I would mainly use the Amps function (not on the SmartGuage) to help determin when to run the watermaker and/or refrigerator, the largest consumers of electricity on my boat. Negative charge and I must turn off one or both. The "Amp Hours Remaining" function I use only as an indication of when I must be careful and miserly about cabin lights and other discretionary uses of power. I have 210Ahr of battery capacity so I never want to use more than 105Ahr, and, to play safe, I try to keep it under 75Ahr. And, realistically, with my 6 solar panels I rarely get that low.
Now you know my reasoning behind my selection of an Ahr counter rather than a SmartGuage.
I have selected the TriMetric45 mainly on the recommendation of the Morningstar technition and not hearing anything bad about it. (Although Morningstar makde my TriStar 45 controller, and the similarity of the TriStar and TriMetric names, Morningstar does not make the TriMetric, Bogart Engineering does.
I guess my question has been answered about as well as it ever will be.