navitron
 
Renewable Energy and Sustainability Forum
Welcome, Guest. Please login or register.

Login with username, password and session length
News: Anyone wishing to register as a new member on the forum is strongly recommended to use a "proper" email address - following recent spam/hack attempts on the forum, all security is set to "high", and "disposable" email addresses like Gmail, Yahoo and Hotmail tend to be viewed with suspicion, and the application rejected if there is any doubt whatsoever
 
Recent Articles: Navitron Partners With Solax to Help Create A More Sustainable Future | Navitron Calls for Increased Carbon Footprint Reduction In Light of Earth Overshoot Day | A plea from The David School - Issue 18
   Home   Help Search Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: So how does the SmartGauge work then..?  (Read 6154 times)
Outtasight
Hero Member
*****
Offline Offline

Posts: 886



WWW
« on: February 26, 2011, 01:24:27 AM »

Picking this up from the SG battery monitor thread so as to kick off another lively debate without polluting stephendv's thread...

You seem to have a very rigorous charging regime so maybe your cells never get out of synch. I'm still intrigued to know how it works by only (I think) measuring voltage. It's no good, I'm going to have to get one and see for my self  stir

The SmartGauge does indeed only measure the terminal Voltage of the battery.  The manual is very particular that the thing be connected directly to the battery posts and nothing else connected to the sense wires.  Any length of wire from the battery posts to the sense point would introduce reading errors of the terminal Voltage (by way of false Voltage drops/increases along the wire from loads or charge sources).

As far as I can tell by watching the behaviour of my battery and the SmartGauge, it is watching the Voltage over time and looking for tell-tale Voltage signatures.  When a battery has just been taken off load, the Voltage recovers over a short period (5 minutes) then levels out (over a few hours).  Based on the shape of the initial recovery, the levelling period, and the final level, you can take a stab at a SoC value.

When a battery is put under load, the terminal Voltage collapses quite a bit in a short time, but then tends to hold up.  The Voltage that it collapses by and how well it holds up, plus the slope of the decline thereafter also gives you a stab at the SoC.  

I can see this going on via my MSView logger from the charge controller.  The charge controller can record terminal Voltage at 24V with 2 decimal place resolution at 15 second time resolution.  This allows you to see quite fine detail in what the battery is doing over time. In the trace below, you can see when a large load causes a steep slope, then I turned some stuff off, then I turned something less heavy on for a few hours (I was watching a movie) and then finally turned most things off, pottered about with just the lights on before finally turning the inverter off for the night.  The recovery curve of the battery is quite distinctive when it went off load altogether (as opposed to just the load changing).  Then you can see a small short dip when someone went to the bathroom in the middle of the night and turned the lights on (the inverter has a wireless remote upstairs).  Note how the battery recovers almost immediately from such a short discharge event.  The charge controller uses zero current sense wires also connected directly to the battery posts, so is seeing exactly what the SmartGauge is seeing.

The gauge isn't fooled by short bursts of high load (that cause the terminal Voltage to collapse to as little as 12V) because it watches how quickly the Voltage collapsed, where it steadied, the slope of decay after it steadied, and how fast it recovered after the load was removed.  A full battery used to start a car may collapse to 11V during starting, but as soon as the load is removed, it will bounce back in a very short time to near full Voltage as it was only a "surface discharge".  Conversely a small initial Voltage drop means it was a small load but if the gentle slope of decay continues for a few hours, then it knows a large part of the capacity has been spent.

It's also not fooled by short duration high Amps charging.  The battery will "surface charge" and the Voltage will shoot up to the absorption limit.  The gauge will know from history that it may have only been at 50% charge so it will take a few hours held at the absorption level to actually get to anything like 100%.  If the charge source is removed early, it can spot the signature that the surface charge diffuses quickly and the Voltage falls back much more and much faster than is the case for a nearly full battery.  With each cycle, it learns and tunes the particular discharge signatures.
      
You have to tell the SmartGauge what type of battery you are using (proper wet, AGM, Gel, Calcium leisure, car...) and this tells the programme what type of internal resistance profile and Voltage signatures to expect.  You don't need to tell it how big the battery is, because it learns this from the Voltage-time responses to charge and load.  It doesn't need to know the size of the load in Amps.  The signatures will tell it how big the load was as a function of the battery size (it can guess from the initial Voltage drop and the slope whether it was a C100 or a C5 load).

As the battery ages, its effective capacity (in absolute Ah terms and also in terms of its internal resistance and so ability to deliver high current) declines.  The SmartGauge constantly learns what the capacity is and how loads affect it and so you are warned to keep within the "50% rule" more effectively.  As the battery ages, you notice that applying the same large loads as when the battery was new, now takes out a bigger percentage of the useful capacity.  You get to the 50% marker more quickly, so you learn to either go easy on the old battery or just hit it for shorter durations (or realise that it's time to buy a new one).  

If you used a Ah counter set for a 100Ah capacity and you kept drawing out 50Ah, thinking it was 50%, you'd soon be wrong and it would be 50Ah from a 80Ah effective battery (62%) and you'd hasten its demise.


* Discharge plot.gif (5.66 KB, 727x404 - viewed 1084 times.)
« Last Edit: February 26, 2011, 01:32:06 AM by Outtasight » Logged

http://solarbodge.blogspot.com/
3.58kWp & 800Ah LiFeYPO4 off-grid(ish). See 'Cobbled together PV in W.Sussex' (in "Show Us Yours")
Justme
Hero Member
*****
Offline Offline

Posts: 3531


« Reply #1 on: February 26, 2011, 10:49:31 AM »

I think the only info you missed out was that over time the SG gets more accurate whilst the amp hour counter gets less accurate. There was some talk that it also measures the resistance of the bat. The inventor posts on the canal boat forum a lot. He is a mine of information re bats, inverters & all things electrical.
Logged

Navitron solar thermal system
30 x 58mm panel 259L TS
1200watts solar 120vdc
FX80 Solar controller
Victron 12v 3000w 120a
6kva genny
6 x 2v cells 1550amp/h 5C
24 x 2v cells 700amp/h 5C
Total bank 4350 amp/h 5C
Eleanor
Hero Member
*****
Offline Offline

Posts: 2574



« Reply #2 on: February 26, 2011, 12:09:19 PM »

Thanks Outtasight. So, the SmartGauge is monitoring the behaviour of the bank as a whole. If the bank is made up of a series of 2V cells and one of the cells is lagging behind ie not fully charging what does the SmartGauge show? Will it refuse to show 100% SOC until that cell has been equalised because it is expecting to see a higher capacity? I can see it's a clever tool for measuring SOC and Amp hour counters don't seem to have the ability to determine capacity. They do record some information eg number of charge cycles, cumulative Amp hours drawn from the battery, depth of average discharge so perhaps some rough deductions on capacity could be made? The history will probably show up all my charging inadequacies  whistlie

My main question is can it be relied on by itself? We've already got the Victron battery monitor to tell us Ah used but is only any use if it's set correctly when the bank is fully charged and it doesn't tell us capacity. The SmartGauge would tell us % SOC so we would know capacity from the two instruments (unless the SmartGauge shows Ah capacity?). I can't see how either of them tell us the condition of the individual cells and how measuring SG can be avoided on eg a monthly basis or by using one of Stephen's loggers. I suppose a sophisticated charger or balancing circuits which adjust the charging for each cell would help overcome this. All the battery monitor information can be logged to PC so perhaps some measurements need to be done (or maybe I just need to get a life  Roll Eyes).
All just thinking aloud really  Huh
« Last Edit: February 26, 2011, 12:10:51 PM by Eleanor » Logged

I'm doing this for free, please be nice to me surrender
"Very few batteries die a natural death ... most are murdered" stir
Eleanor
Hero Member
*****
Offline Offline

Posts: 2574



« Reply #3 on: February 26, 2011, 12:23:52 PM »

No excuse for not getting one - Maplin are selling them :

http://www.maplin.co.uk/smartgauge-battery-monitor-399510

Does it need anything else to go with it?
Logged

I'm doing this for free, please be nice to me surrender
"Very few batteries die a natural death ... most are murdered" stir
EccentricAnomaly
Guest
« Reply #4 on: February 26, 2011, 12:42:28 PM »

I can't see how either of them tell us the condition of the individual cells and how measuring SG can be avoided on eg a monthly basis or by using one of Stephen's loggers.

I agree, monitoring individual cells is important. I'm not sure that monitoring SG as such is so important, though. Wouldn't monitoring the voltage across the individual cells be enough to alert you that something was amiss - if one or some weren't tracking the rest of the bank?
Logged
Outtasight
Hero Member
*****
Offline Offline

Posts: 886



WWW
« Reply #5 on: February 26, 2011, 07:06:55 PM »

I don't think it will cope well with a defective bank.  It probably assumes all the cells are equally matched.  It did say in the manual that you can monitor a 12V or a 24V bank at the terminal posts.  It can't do a 48V though.  You can however monitor a 48V bank by using a Smartgauge to just monitor a 12V or a 24V subsection of the total bank and it will read the state of charge of the subsection.  If the cells are matched then the reading will be good for the whole bank.

It doesn't display Ah, just Volts or % SoC.  You'd leave it showing % SoC as that's the only thing you're interested in.
Logged

http://solarbodge.blogspot.com/
3.58kWp & 800Ah LiFeYPO4 off-grid(ish). See 'Cobbled together PV in W.Sussex' (in "Show Us Yours")
Justme
Hero Member
*****
Offline Offline

Posts: 3531


« Reply #6 on: February 26, 2011, 07:41:47 PM »

It has other features as well.

It has a built in relay that can be set for high / low V or high / low SOC & then either unlatch due to the same thing or time or time after a set point.

We had it set to start the genny when soc got to 60% & stay on till 100% + X amount of time. Add in a timer so it cant start at night & your sorted.

The display can be always on, always off or to come on / go off at dif V & SOC levels.


Sneak preview, there is a big brother version coming soon with LOTS more features.
Logged

Navitron solar thermal system
30 x 58mm panel 259L TS
1200watts solar 120vdc
FX80 Solar controller
Victron 12v 3000w 120a
6kva genny
6 x 2v cells 1550amp/h 5C
24 x 2v cells 700amp/h 5C
Total bank 4350 amp/h 5C
Outtasight
Hero Member
*****
Offline Offline

Posts: 886



WWW
« Reply #7 on: February 26, 2011, 11:40:23 PM »

Yeah, I've not looked at using the relay for anything yet.  It could be easily used to drive a SoC based emergency disconnect for automatic load shedding (or as mentioned auto generator start up).

It would be really useful if the SoC reading could be made available via IP Ethernet.  Will increase the power consumption of the unit somewhat but would not need to be read more than once an hour, so a auto-sleep chipset could be used.  It might even work out about the same as running the old monitor in permanent display on mode (if you run it in display off mode but have periodic Ethernet access).

Also I'd hope the new version has an additional battery profile for the new Elecsol carbon fibre AGM type.
Logged

http://solarbodge.blogspot.com/
3.58kWp & 800Ah LiFeYPO4 off-grid(ish). See 'Cobbled together PV in W.Sussex' (in "Show Us Yours")
Outtasight
Hero Member
*****
Offline Offline

Posts: 886



WWW
« Reply #8 on: February 28, 2011, 12:09:58 AM »

Interesting thing Gibbo (the designer of the Smartgauge) mentioned here:

http://www.canalworld.net/forums/index.php?showtopic=34433&hl=gibbo&st=120

About the Smartgauge (SG):
Quote
Alternatively, if anyone already has one and uses those batteries it can be programmed, in the field, into the custom battery model hole in SG.
So in summary I don't need to touch the actual SG program, just the battery model. And a new custom model can be put in "in the field" via a hidden menu.

He was talking about a tweaked model for the Odyssey batteries that are similar to the Hawker ones that he was trying to test to destruction but were proving pretty indestructible (see his final results test quote below).

Quote
The (was) brand new battery has now done around 125 cycles to 10.5 volts at various (wildly different) currents. It is showing a capacity of 125ahrs (20hr rate - datasheet states 120ahrs) and a charge efficiency of 98.0%

That's astounding... still showing above nameplate capacity after 125 discharges to 100% DoD!

Logged

http://solarbodge.blogspot.com/
3.58kWp & 800Ah LiFeYPO4 off-grid(ish). See 'Cobbled together PV in W.Sussex' (in "Show Us Yours")
Justme
Hero Member
*****
Offline Offline

Posts: 3531


« Reply #9 on: February 28, 2011, 09:16:52 AM »

I know thats why I mentioned them. They do seem quite remarkable. I do actually have the method of doing the bat programming (not that I know what to change, I just got told what was needed) as my bank's indicated SOC climbs to fast to 100% on the meter so I then need to hold it at 100% (like an absorb phase) for a while. If I stop charging as soon as it says 100% I then dont get as long a running time as if I do the extra bit. But as I am just about to change the bat bank to a larger one I am holding off changing it until then. I have invited Gibbo to join in over here as I am sure he would fit right in. I hope he does.
Logged

Navitron solar thermal system
30 x 58mm panel 259L TS
1200watts solar 120vdc
FX80 Solar controller
Victron 12v 3000w 120a
6kva genny
6 x 2v cells 1550amp/h 5C
24 x 2v cells 700amp/h 5C
Total bank 4350 amp/h 5C
Outtasight
Hero Member
*****
Offline Offline

Posts: 886



WWW
« Reply #10 on: February 28, 2011, 01:21:27 PM »

My SG now seems to have the opposite "problem" or maybe not.  It takes a very long time to get to 100% (almost never in fact), despite the Ammeters showing next to no Amps after a couple of days of light use and long absorptions.  It will go up from about 50% to about 75% in a single sunny day and then the next day it will creep up to 85-90% if it gets a second full days charge in.  Over the last week of December, I didn't use the battery at all (no inverter use at all) and it eventually crept up to 100% after 5 days charging.

It could be right though, as after a solid charge for a couple of days the SoC reading goes down more slowly than if it had only had weak partial charges for a long time. I also notice that the split of discharge / charge changes between the AGM and Gel batts in the pack.  His posting on the canal forum does suggest that short term abuse of batteries can significantly reduce their effective capacity, but with some TLC, the capacity can come back (if the battery is a good one, like the Hawkers).  Others permanently get rotten and never recover.  At least it means I'm only using up to 50% of the capacity of the bank in it's current state of health, a very important factor that Amp.hour counters take no consideration of.

I'm not using the SG as strictly intended now though.  It's connected to the terminal posts of the gel pack (and using the gel battery model), but there's an AGM pack connected in parallel via a long and lossy "bridge" (2.5m of 6mm cable with a 32A breaker in the positive leg).  So I think the SG is reading the SoC of the gel pack with an "odd" remote load / charge source attached all the time (in addition to the inverter load and solar charge sources).  When the whole bank has been partial charged for quite some time, I've observed the internal resistance of the gel pack get quite high (the current for load / charge gets weighted more in favour of the AGM pack, even though it's at the end of all that "high resistance" wiring).  After proper charging, the internal resistance of the gel pack falls and it shares the charge and load much more equally with the AGM pack.

The charge regimes for the gel and AGM packs aren't so different.  The guide for the Marathon UPS AGM batts suggested the same absorption and float settings as the Deka/MK gel batts.  Their internal resistance behaviours are different though.  The AGMs keep low internal resistance, and pump out the Amps (probably until they self-destruct), but the Deka ones increase internal resistance dramatically, so that the Voltage collapses as they get to anywhere near discharged.  The acid starved design is part of their deep discharge protection (the acid runs out before the lead does).  Conversely, the gel batts take a while to start to charge up as their internal resistance is high and the AGMs greedily suck up all the charge (sometimes even tripping the 32A breaker).  But once the gels get going, they can be just as greedy.  I suppose that alone explains a big part of why the SG takes a time for its reading to creep up under charging, as the gel batteries really aren't charging up that much at first (the AGMs are stealing the lions share).  But it doesn't explain why it takes so long to creep up that last 5-10%, when both the gel and AGMs are sitting at float Voltage, drawing next to no Amps.

Not sure what impact the bank balancer is having on the SoC readings yet.  The AGM bank had two 100Ah packs that were unbalanced, but with the 1A bypass balancer I've got on it and the inter-pack mid-point common-rail to balance the three series strings in the AGM pack, it might change things again.  The balancer only activates above 12.8V, so has no impact during discharges.  Just stops the tendency for one half of a string Voltage to go above the other and start gassing (fatal for a AGM or gel battery).

Even with all this non-standard configuration, the Smartgauge seems to keep me on the right side of dead batteries and if it's a bit conservative on the remaining capacity and pessimistic on the fully charged indication, that's better than it being too optimistic... 
Logged

http://solarbodge.blogspot.com/
3.58kWp & 800Ah LiFeYPO4 off-grid(ish). See 'Cobbled together PV in W.Sussex' (in "Show Us Yours")
Justme
Hero Member
*****
Offline Offline

Posts: 3531


« Reply #11 on: February 28, 2011, 01:46:50 PM »

Wow thats some set up lol.

Did you use bat type 1 or 3 on the set up?

I think they are both for AGM's but for differing types.

I really do hope now that he pops in just to see what he makes of that.

Did you spot that once the bank is down to 50% of its original capacity that all bets are off? For small banks that would be the time to change them but I would guess that a very large forklift bank could still have a useful life left for light use off gridders.
« Last Edit: February 28, 2011, 01:51:29 PM by Justme » Logged

Navitron solar thermal system
30 x 58mm panel 259L TS
1200watts solar 120vdc
FX80 Solar controller
Victron 12v 3000w 120a
6kva genny
6 x 2v cells 1550amp/h 5C
24 x 2v cells 700amp/h 5C
Total bank 4350 amp/h 5C
billi
Global Moderator
Hero Member
*****
Offline Offline

Posts: 8975



WWW
« Reply #12 on: February 28, 2011, 11:52:45 PM »

Interesting , 

sound like a  toy i should have got years back

Logged

1.6 kw and 2.4 kw   PV array  , Outback MX 60 and FM80 charge controller  ,24 volt 1600 AH Battery ,6 Kw Victron inverter charger, 1.1 kw high head hydro turbine as a back up generator , 5 kw woodburner, 36 solar tubes with 360 l water tank, 1.6 kw  windturbine
Outtasight
Hero Member
*****
Offline Offline

Posts: 886



WWW
« Reply #13 on: March 01, 2011, 10:09:30 PM »

Wow thats some set up lol.

Did you use bat type 1 or 3 on the set up?

I think they are both for AGM's but for differing types.

I really do hope now that he pops in just to see what he makes of that.

Did you spot that once the bank is down to 50% of its original capacity that all bets are off? For small banks that would be the time to change them but I would guess that a very large forklift bank could still have a useful life left for light use off gridders.


Been out today at the EcoBuild exhibition in London.  Massive exhibition.

I used the Gel model (type 2).  I think the AGMs I have are the kind that Gibbo refers to as the type 3, as they have a requirement for quite low charge Voltages that are very similar to the gel ones.  If they'd been like some other AGMs that claim to like being charged at 14.4-14.6V absorption, then they'd have been totally incompatible with the gel ones that like no more than 13.8-14.1V.  So, apart from their different internal resistance foibles, they seem to be fairly good bed fellows.  I tried to compensate for the AGM low internal resistance by deliberately adding the extra wiring resistance.  I could maybe invent some kind of active balancing, using differential current measurement and use lengths of wire with short-circuiting relays to add or subtract resistance to the AGM bridge.  If the AGMs are taking too much charge / load, turn off a relay and it will add a loop of cable and a few milliOhms to compensate...

Or I could just for once buy a set of batteries that are all the same make / model / size / chemistry...  Roll Eyes  But where would be the fun in that?  Grin
Logged

http://solarbodge.blogspot.com/
3.58kWp & 800Ah LiFeYPO4 off-grid(ish). See 'Cobbled together PV in W.Sussex' (in "Show Us Yours")
Pages: [1]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.20 | SMF © 2013, Simple Machines
SMFAds for Free Forums
Simple Audio Video Embedder
Valid XHTML 1.0! Valid CSS!