Battery Charging Graphs

[nicknorman]

This topic for the battery-obsessed only!

 

I have a few Mastervolt devices including a Mastershunt (AH counting gauge) all connected via Masterbus. This gives the possibility to connect a PC and using their Masteradjust software, to log pretty much any of the parameters recorded. So I set it up to record battery current, voltage, SoC and AH consumed, every 5 seconds. I then started the engine this morning from a mid-60s SoC. Unfortunately we stopped for lunch before reaching fully charged so the point of this first post is just to observe the first part of the charge. Later I will complete the charge and post the graph approaching 100% SoC.

 

Batteries are 450AH of Trojan T105s, 2 years old.

 

 

 

The behaviour shortly after starting is interesting, a large current peak (we have a 175A alternator) falls off quickly but then the current starts to increase again (with concomitant slight reduction in voltage as the load on the alternator increases. Here is a zoom of this area:

 

Any thoughts or comments to explain the shape of the graph in this initial charging region?

 

For those who don't realise, the graphs are images and clicking them will enlarge them.

 

 

Just wondering what happens if I post the actual excel spreadsheets:

 

Data.xls

 

Yes that works but I guess you need to have Excel or equivalent on your computer.

Edited February 14, 2016 by nicknorman

[Guest]

Just opened your XLS sheet no problem with LibreOffice.

[nicknorman]

Just opened your XLS sheet no problem with LibreOffice.

Yes it even opens on my iPad, but a lot of the formatting is lost. Libre should be pretty good at working with xls though.

[Keeping Up]

I can't explain that inital effect, except t say that from visual observation on my Nasa BM2 I get the same effect, and I have wondered about it too. In particular is it the alternator or the battery which is causing the effect? (ETA or an effect within the Mastershunt?)

 

Is yours also the Iskra alternator?

Edited February 14, 2016 by Keeping Up

[nicknorman]

It is an Iskra. It definitely isn't the alternator because it is the battery that determines the current, the alternator's voltage is them a reflection of the load. As well as the current increasing after a minute or two, the voltage is falling. If the voltage were rising one could perhaps put it down to the alternator, but it isn't, it's falling!

[Keeping Up]

Point taken, I agree it does sound as if it must be the battery which is causing the effect. How about it being the result of the inital high current knocking the thin sulphurous coating off the plates?

 

However nothing is ever 100% certain so I'm keeping all options open. Including the option that the measurement system is giving an odd effect. I say that because mine has an odd effect: for a few seconds after starting the engine it shows a large DISCHARGE current of up to 150 amps which I don't believe, before then showing the expected charge current. I have some theories as to why - we can possibly discuss it after we've all given more thought to your figures here.

[nicknorman]

We are heading back to the marina tonight (Jeff has to go back to work) but I'm staying down til Tuesday evening so it will give the opportunity to repeat the excercise. We have a 100A charger (and shore power) which is obviously not as big as the alternator so it will be interesting to see if the same phenomena manifests itself.

[dmr]

Trying to replicate this with your big charger is the obvious next step, but as you have a clever charger it might just have some soft start built into it?.

 

I have also observed both the short term and slightly longer term effects that you see.

A few of thoughts (which I have not thought too deeply about).

A knackered battery will sometimes take no charge at first, but sometimes it increases later if left on charge. Could the increase in current be a much smaller version of this "waking up" effect?

I still think temperature is a factor in this, the winding go from cold to maybe 100degC pretty quickly, this must play a part.

I half agree the voltage maybe goes against this but I am not sure.

Can we assume an alternator is a simple voltage generator (V = L di/dt sort of thing) or does it have some constant current characteristic too?

What about the diodes warming up? Don't semiconductors have a negative temp coefficient!

 

Thank you so much for doing these tests and sharing your results, measurements are so much better then pontifications!

 

..................Dave

[nicknorman]

Trying to replicate this with your big charger is the obvious next step, but as you have a clever charger it might just have some soft start built into it?.

 

I have also observed both the short term and slightly longer term effects that you see.

A few of thoughts (which I have not thought too deeply about).

A knackered battery will sometimes take no charge at first, but sometimes it increases later if left on charge. Could the increase in current be a much smaller version of this "waking up" effect?

I still think temperature is a factor in this, the winding go from cold to maybe 100degC pretty quickly, this must play a part.

I half agree the voltage maybe goes against this but I am not sure.

Can we assume an alternator is a simple voltage generator (V = L di/dt sort of thing) or does it have some constant current characteristic too?

What about the diodes warming up? Don't semiconductors have a negative temp coefficient!

 

Thank you so much for doing these tests and sharing your results, measurements are so much better then pontifications!

 

..................Dave

The alternator is capable of putting out 175A for prolonged periods, and I can prove it by including in a log the current going into the inverter as well as into/out of the battery. Then with the engine running, I put the 2kw kettle on which will pull about 180A into the inverter. I'll do that later. As you say there will be some effects due to heating of the alternators bits but I am confident that in no way does it explain the phenomena, primarily because as the battery current increases the voltage is reducing.

 

It is perhaps worth mentioning that there is a fairly long time constant on the MS' current reading as it take several seconds (maybe 5) for the reading to reach 170A or whatever, and an equal amount of time to fall back to a low reading if I stop the engine. I am pretty sure this is the only reason the current doesn't seem to go straight to an initial high value but rises over a few seconds.

[dmr]

The alternator is capable of putting out 175A for prolonged periods, and I can prove it by including in a log the current going into the inverter as well as into/out of the battery. Then with the engine running, I put the 2kw kettle on which will pull about 180A into the inverter. I'll do that later. As you say there will be some effects due to heating of the alternators bits but I am confident that in no way does it explain the phenomena, primarily because as the battery current increases the voltage is reducing.

 

It is perhaps worth mentioning that there is a fairly long time constant on the MS' current reading as it take several seconds (maybe 5) for the reading to reach 170A or whatever, and an equal amount of time to fall back to a low reading if I stop the engine. I am pretty sure this is the only reason the current doesn't seem to go straight to an initial high value but rises over a few seconds.

 

As I think I have said before, I am surprised you can get your 175amps long term.

Ive got the 100 amp Iskra and it starts at about 85 amps then drops to about 50 where it stays for the bulk charge.

I know this is not battery limited because it comes up to 80 isk if I short in the second alternator.

Alternator cable is a bit dodgy though, one of the few thick wires that I have not replaced yet, plan to buy some cable on eBay this evening.

 

I think fitting that larger frame 175 amp alternator to my engine would be financially traumatic.

 

It makes total sense for the MS to do some averaging, its not easy getting good quality data out of a shunt.

 

...............Dave

[WotEver]

It could be nothing more than an overshoot on the MS logging software. Do you have any other ammeter to compare with? Maybe your clamp meter?

 

Tony

[nicknorman]

It could be nothing more than an overshoot on the MS logging software. Do you have any other ammeter to compare with? Maybe your clamp meter?

Tony

Well it wouldn't be specifically the logging software because it's also what I see on the display. However yes I do have a Clampmeter and so I can check it tomorrow morning (I'll stay off shore power tonight). I'm 99.9% sure it won't be, but easy to check and thus eliminate.

Ive got the 100 amp Iskra and it starts at about 85 amps then drops to about 50 where it stays for the bulk charge.

I know this is not battery limited because it comes up to 80 isk if I short in the second alternator.

 

...............Dave

I think what you are seeing is just the consequence of the soft knee of the regulation curve. I think if you added some load to pull the voltage down a bit, your alternator would go up to 100A, presuming it's spinning fast enough. Not into the battery of course, but total output ie battery + additional load.

[sjc]

Consider the following ideas :-

1) An immediate start of the alternator with a heavy load already attached could force the regulator to overshoot ?

2) A step input to the batteries ( do they present a capacitive load ? ) will force a short sharp current pulse ?

[Tony Brooks]

Consider the following ideas :-

1) An immediate start of the alternator with a heavy load already attached could force the regulator to overshoot ?

2) A step input to the batteries ( do they present a capacitive load ? ) will force a short sharp current pulse ?

 

 

1. but the self current limiting design of most alternators should have caught that. The regulator is probably fully conductive at that time.

 

2. Interesting, the insides look very much like the open variable capacitors of old. I do not know how the electrolyte would modify any capacitive behaviour or even if there is any.

 

 

Well done Nick, very interesting work & topic.

[nicknorman]

Consider the following ideas :-

1) An immediate start of the alternator with a heavy load already attached could force the regulator to overshoot ?

2) A step input to the batteries ( do they present a capacitive load ? ) will force a short sharp current pulse ?

1) the regulator regulates voltage and the data shows the voltage at the peak current, of 14v. The final regulated voltage is at least 14.4v so it has not overshot on voltage.

 

2) I suppose the battery will have slight capacitance, not much though due to the plates being fairly far apart. Whereas the peak of current lasts nearly a minute and is perhaps 40A over the settled value. That would take a massive capacitor.

 

Sorry to be negative, keep coming up with the ideas!

[WotEver]

The battery has a HUGE capacitance - many thousands of coulombs. The typical battery equivalent circuit is this capacitor in series with an RC network in series with a resistor.

 

Tony

[nicknorman]

The battery has a HUGE capacitance - many thousands of coulombs. The typical battery equivalent circuit is this capacitor in series with an RC network in series with a resistor.

Tony

Hmmm, no I prefer my Indian restaurant equivalent circuit!

[by'eck]

Interesting graphs and agree that the initial charge voltage/current phenomena is purely down to battery characteristics.

 

Not surprising the voltage drops during the relatively brief charge current increase although there seems to be lag in the voltage increase after current starts to fall again. My only thought is its an initial surface charge effect caused by the relatively high (C/3?) charge ratio.

 

I suspect differing battery constructions would provide substantially different characteristics under the same conditions during this early part of the charge cycle. Would be interesting to see a similar graph for leisure batteries for example.

[nicknorman]

Interesting graphs and agree that the initial charge voltage/current phenomena is purely down to battery characteristics.

 

Not surprising the voltage drops during the relatively brief charge current increase although there seems to be lag in the voltage increase after current starts to fall again. My only thought is its an initial surface charge effect caused by the relatively high (C/3?) charge ratio.

 

I suspect differing battery constructions would provide substantially different characteristics under the same conditions during this early part of the charge cycle. Would be interesting to see a similar graph for leisure batteries for example.

 

<off topic, just got back to the marina, your boat is looking fine if a little lonely!>

[by'eck]

<off topic, just got back to the marina, your boat is looking fine if a little lonely!>

 

Thanks Nick.

 

Back on topic I suspect starter batteries of same capacity and given the same charge ratio and source would exhibit this effect to a much lesser extent, given the greater plate surface area in contact with electrolyte.

 

I would also suspect very mild sulphation as suggested by Allan (KU) would contribute to the effect.

[smileypete]

Could try it with and without a HEAVY discharge right before recharging, see if there's any difference to the voltage peak.

[nicknorman]

Thanks Nick.

 

Back on topic I suspect starter batteries of same capacity and given the same charge ratio and source would exhibit this effect to a much lesser extent, given the greater plate surface area in contact with electrolyte.

 

I would also suspect very mild sulphation as suggested by Allan (KU) would contribute to the effect.

Just picking up on the sulphation comments, I've no idea if this is the explanation but just to mention the batteries have been on float charge for weeks, discharged a bit on Friday (85%SoC) recharged by chugging to Fradley for 5 hours, discharged overnight and then the charge curves you have seen. So hopefully no exposure to less than 100% SoC except overnight.

[nicknorman]

So here is tonight's data. We stopped for lunch with the SoC at 93% and the current at 14A. I was very good and boiled the kettle whilst the engine was still running! When we started up again unfortunately I forgot to restart the log until the engine had been running for a few minutes, so no possibility for initial transients. But anyway by the time I started logging the SoC was 91% (just 2% less than when I stopped for lunch) but interestingly the current was much higher at about 40A. So perhaps giving the batteries a long rest most of the way through the charge, is a way to reduce overall genny running time (if you are the genny sort of boater)? It is the same concept as surface charge etc, the long rest allowed the charge to permeate more evenly through the plates so the surface became more receptive to the re-started charge.

 

On the graph you will notice a couple of things - we got back to the marina and I put it on shore charge at the same voltage, but there is a slight glitch where the Combi took over. Then later on I decided to up the voltage to 14.9 which is a good Trojan finishing voltage. The current of course went up, but not by that much, and soon fell back to fairly close to what it had been.

 

But I think the most notable point is that having started at 91% SoC, the batteries were still not fully charged 5.5 hours later, as can be seen by the fact that the current is still falling slowly. So (bearing in mind this a favourite point of mine) at what point should one consider the batteries to be fully charged?

 

 

Needless to say the MS' SoC was slightly out of kilter as evidenced by its jump from 97% to 100%.

 

The actual xls attached if anyone is interested:

 

Data3.xls

Edited February 14, 2016 by nicknorman

[by'eck]

Just picking up on the sulphation comments, I've no idea if this is the explanation but just to mention the batteries have been on float charge for weeks, discharged a bit on Friday (85%SoC) recharged by chugging to Fradley for 5 hours, discharged overnight and then the charge curves you have seen. So hopefully no exposure to less than 100% SoC except overnight.

 

Maybe the very fact they have recently been left on float for an extended period may have caused minor sulphation in excess of that they would have received from regular daily charge/discharge use.

 

In any case the current/voltage blip may better be explained by plate electro-chemical dynamics (just made that up ) in relation to rate of absorption of charge into plate depth.

[nicknorman]

Could try it with and without a HEAVY discharge right before recharging, see if there's any difference to the voltage peak.

Yes that is the problem, there are loads of different combinations of conditions to test, and they will probably all give different pictures! I started this morning's charge after using the electric kettle and coffee machine for breakfast, but that was over an hour before I started the engine