SOC suddenly jumps to 100%

An area for discussing the OutBack FLEXnet DC product.

Moderators: OutBack Moderator Team, OutBack

Re: SOC suddenly jumps to 100%

Postby Kent Osterberg on Thu Apr 03, 2014 5:16 pm

David,

I agree that the FNDC shouldn't ever say 100% until the charged parameters are met. In fact, when charging, it would probably be good if it stopped increasing the displayed state of charge at 90%, or maybe 95%, until the charged parameters are met.
Kent Osterberg
Forum Emperor
 
Posts: 1503
Joined: Mon Jun 09, 2003 11:01 pm
Location: Cove, Oregon
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules

Re: SOC suddenly jumps to 100%

Postby HawaiianDrummer on Thu Apr 03, 2014 8:23 pm

Aloha, Yes Kent there is room for product development, I agree with you. And after reading tall girls criteria for battery management before totally putting the blame on the FLEXnet DC, today I re-entered a lot of the FLEXnet DC set points. Today I had full sun from dawn till dusk and got the 1,100 AH battery bank mostly full. I read the battery cells with the hydrometer and most were in the green at 1.265 to 1.275. I lowered the return amps to 11A instead of 22A, and time met to 1 HR. instead of 4 minutes, and raised my battery charge factor to 90% from 65% I also lowered the FM80 absorb times to 5 1/2 hrs. rather than 7 hrs. hoping that float will start kicking in more often instead of only Absorb time to finish off the daily charge. Now like an eagle I will watch the results over the next few weeks to see if any changes for the better are achieved in regards to the FLEXnet DC time required to reach 100%. I will update, David
HawaiianDrummer
Forum Guru
 
Posts: 323
Joined: Thu Apr 02, 2009 1:42 am
Location: Ninole, Hawaii
My RE system: 1-FM60's, 2-FM80's 2- FX3048's, Mate3, Hub 10, 12 US 64 w panels, 12 Kyocera
175 W panels, 21 Kyocera 225 W panels, 24 Trojan 2V 1,110 AH 48 V Battery
bank, Flex Net DC and Optics RE

Re: SOC suddenly jumps to 100%

Postby tallgirl on Sat Apr 05, 2014 7:27 am

David,

Absorb time should be set based on what it take to flatten the charging curve. That's the "Gold Standard" for end-of-charge. Were I to design a battery monitor (something I keep thinking about doing, but never have the time, and really don't have the time now), I'd say to hell with all the states-of-charge computations, parameters-met values, etc. and stick with one simple number -- the change in current with respect to a constant temperature corrected voltage.

The logic is very simple -- batteries are giant containers that do chemical reactions which obey certain laws. Those laws imply the shape of the charging curve, and the eventual bottoming-out, in an asymptotic manner, of the acceptance current.

There is still more math which can be used to determine if a battery bank is sulfated, and that's the #1 cause of poor battery monitor behavior. That's where I see the most "other" problems, followed by incorrect "charge factor" settings on products like the FNDC.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby vtmaps on Sat Apr 05, 2014 7:48 am

tallgirl wrote:Absorb time should be set based on what it take to flatten the charging curve. That's the "Gold Standard" for end-of-charge.
<snip>
The logic is very simple -- batteries are giant containers that do chemical reactions which obey certain laws. Those laws imply the shape of the charging curve, and the eventual bottoming-out, in an asymptotic manner, of the acceptance current.


I almost agree completely... but there can be a "false" end amps if Vabs is not high enough to break up stratification, or if (as you mentioned earlier in this thread) the current is so low that Vabs is reached too slowly.

--vtMaps
vtmaps
Forum Guru
 
Posts: 111
Joined: Mon Nov 07, 2011 6:09 pm

Re: SOC suddenly jumps to 100%

Postby tallgirl on Sat Apr 05, 2014 8:20 am

VT -

Yes, the FNDC is more than capable of "Garbage In, Garbage Out".

This is not, however, the fault of the FNDC. The other issue, which must periodically be mentioned, is that the SOC and "Parameters Met" calculations are mostly independent. The two should serve as checks on each other, which is why I never understand people who set the "charge efficiency" (or whatever -- I forget the field name) value to values like 70 or 80. I think mine is set to around 91 or 92 percent, I forget (I forget a lot about how my battery-backed system is set up -- I work on straight grid-tied systems mostly these days). If it ever gets much lower, I'll scrap the entire battery bank and start with fresh ones. That's the current, actual charge efficiency thingy. That's the number I use, and it works out (the "parameters met" values are set wrong at the moment, but that's a different store that involves a bus bar and a loose wire ...)
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby HawaiianDrummer on Sat Apr 05, 2014 11:06 pm

Aloha, Hey tall girl, I'm happy to see that you, "a technical genius" also has issues with the FLEXnet DC settings. In reference to why many of us were setting the BCF to lower numbers had to do with reaching 100% prematurely and trying to fool the system into charging longer until an actual high SOC was reached. But it doesn't hold true , So that's why I went back to the closer to default settings while at the same time now trying to manage the charging better with strict monitoring of the electrolyte hydrometer readings. and Yes, Thank You for reminding me about watching for the charge curve to flatten out to determine absorb end....I might raise the Absorb time back up as needed after I monitor my batteries day by day. (...although with my settings as they are now. these are the first days in months that I got my Float charging time back) So as with most mathematical permutations, ours here in rainy Hilo, Hawaii though the winter season, further hinder trying to lock in "magical numbers" I think it's coming down to using "seasonal" settings to try and stay on top of everything! Thanks, David
HawaiianDrummer
Forum Guru
 
Posts: 323
Joined: Thu Apr 02, 2009 1:42 am
Location: Ninole, Hawaii
My RE system: 1-FM60's, 2-FM80's 2- FX3048's, Mate3, Hub 10, 12 US 64 w panels, 12 Kyocera
175 W panels, 21 Kyocera 225 W panels, 24 Trojan 2V 1,110 AH 48 V Battery
bank, Flex Net DC and Optics RE

Re: SOC suddenly jumps to 100%

Postby tallgirl on Sun Apr 06, 2014 12:17 am

David,

Well ... my FNDC spends almost all the time working flawlessly. The settings are "wrong" right now because I had a problem with some of the unrelated wiring and the charge controller was off-line for several days waiting for me to get the free time. Rather than screw around with the batteries, getting them back to fully and completely charged, I just changed the "Parameters Met" settings so they got met real easy like.

Judging from the number of people who've reported that they set the "charge efficiency" thingy to some low value, it sounds like that is the more likely problem. The FNDC is really two battery monitors in one. The first "half" tries to track amp-hours. The second "half" tries to determine when the batteries are full. The first "half" requires that the "charge effiency" thingy be set very accurately =and= that you reach "Parameters Met" every few days. The second "half" requires that you set the values tightly enough that it's like you're threading a needle from low earth orbit.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby timmartin on Tue Apr 08, 2014 5:23 pm

Today, for completely unknown reasons, my FNDC did *not* reset to 100% when net Ah reached 0. It actually waited until the Charge Factor Corrected Ah reached 0! I did nothing, I changed nothing. Today it just decided to work.

Bet it won't tomorrow.

Outback Support has simply stopped responding to my emails. Very professional.
timmartin
Forum Guru
 
Posts: 273
Joined: Thu Dec 02, 2010 9:31 am
My RE system: ...
Sharp 120W Panels (12)
SunPerfect 230W Panels (9)
Whisper 100 Wind Turbine
Outback FM60 charge controller
Outback FM80 charge controller
Outback VFX3648 inverter/chargers (2)
Outback FlexNet DC
CALB CA400 LiFePO4 3.2V batteries
(32 cells paired together, 1 string = 51.2V 800Ah)

Re: SOC suddenly jumps to 100%

Postby HawaiianDrummer on Fri Apr 11, 2014 11:26 pm

Aloha, ...why should they respond, their Flexmax DC is performing exactly how they designed it to! ...sorry, David
HawaiianDrummer
Forum Guru
 
Posts: 323
Joined: Thu Apr 02, 2009 1:42 am
Location: Ninole, Hawaii
My RE system: 1-FM60's, 2-FM80's 2- FX3048's, Mate3, Hub 10, 12 US 64 w panels, 12 Kyocera
175 W panels, 21 Kyocera 225 W panels, 24 Trojan 2V 1,110 AH 48 V Battery
bank, Flex Net DC and Optics RE

Re: SOC suddenly jumps to 100%

Postby Kent Osterberg on Sat Apr 12, 2014 8:18 am

Aloha, ...why should they respond, their Flexmax DC is performing exactly how they designed it to! ...sorry, David


David,

I disagree. I can't believe that anyone would have intentionally designed it to work that way. However, it is working exactly the way it was programmed to do, and apparently Outback has no intention of fixing it. I wonder if they have being paying attention to what is happening with GM ignition switches these days. Sorry; that's a low blow - but this has gone on far too long.
Kent Osterberg
Forum Emperor
 
Posts: 1503
Joined: Mon Jun 09, 2003 11:01 pm
Location: Cove, Oregon
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules

Re: SOC suddenly jumps to 100%

Postby tallgirl on Sun Apr 13, 2014 9:57 am

Kent Osterberg wrote:
Aloha, ...why should they respond, their Flexmax DC is performing exactly how they designed it to! ...sorry, David


David,

I disagree. I can't believe that anyone would have intentionally designed it to work that way. However, it is working exactly the way it was programmed to do, and apparently Outback has no intention of fixing it. I wonder if they have being paying attention to what is happening with GM ignition switches these days. Sorry; that's a low blow - but this has gone on far too long.


I don't work for OutBack, and they don't pay me any money for anything I do. I don't get paid to post here.

From everything that has been written about the FLEXnet DC, including the "reset the SOC to 100%" feature that was explained, the FLEXnet DC is working correctly with one caveat about its data reporting. The FLEXnet DC includes a number of 16-bit values, and sometimes that values are reported incorrectly when a bit is carried or borrowed across the 8-bit boundary.

That's it. That's the sum total of all the problems I've ever found with the FLEXnet DC in 6 years or so of looking.

The biggest problem with the FLEXnet DC is that it is very challenging to properly configure and the documentation that OutBack has produced leaves a lot to be desired. The FLEXnet DC operates as two independent components. One counts amp-hours out and discounts amp-hours in by an "efficiency" value. It maintains a running value for the State of Charge based on that calculation. The other component looks at the net current to the batteries, the voltage at that time, and resets a flag (and potentially the SOC) when that happens.

The key configuration problems are fairly straight-forward. At one time I made a bit of pocket change helping people configure FLEXnet DCs, though that has waned over the years. And to be blunt, it has waned as more people have flat-out blamed the device rather than blaming the documentation or their own actions.

First and foremost, you CANNOT configure a FLEXnet DC until your batteries are fully and properly equalized, watered, and everything else that constitutes as healthy as they get. It can't be done.

Secondly, you cannot guess at any of the values. The "2% acceptance charge" rule of thumb for end-of-charge is an estimate. If it is higher than the acceptance charge for your batteries, you will chronically undercharge your batteries and they will gradually become sulfated. The "efficiency" value must likewise be determined, and if you don't want all these weird SOC jumps, it must be determined fairly accurately.

Thirdly, the "parameters met" values can take days or even WEEKS to get right. As an added bonus, they may have to be adjusted after you equalize your batteries, or as they age in general.

The sorts of work-arounds that I've seen reported here -- mostly lowering the "efficiency" value -- will cause the FNDC to behave badly. Not seeing the parameters met values correctly will do the same thing. Being impatient and thinking the current values are "close enough" will cause it to misbehave.

When I was consulting more actively, I never met an FNDC I couldn't get to work correctly -- other than the non-atomic value reporting issue. The only issue I've consistently seen with off-grid systems is people who don't get a full and complete charge every few (3 or 4) days. In those instances, the SOC value can become increasingly inaccurate as it doesn't include a self-discharge calculation in the amp-hour counter. There are even ways to account for that using the "efficiency" factor.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby timmartin on Mon Apr 14, 2014 10:14 am

tallgirl wrote:That's it. That's the sum total of all the problems I've ever found with the FLEXnet DC in 6 years or so of looking.

The biggest problem with the FLEXnet DC is that it is very challenging to properly configure and the documentation that OutBack has produced leaves a lot to be desired. The FLEXnet DC operates as two independent components. One counts amp-hours out and discounts amp-hours in by an "efficiency" value. It maintains a running value for the State of Charge based on that calculation. The other component looks at the net current to the batteries, the voltage at that time, and resets a flag (and potentially the SOC) when that happens.


Then how do you explain my many many examples of the SOC resetting to 100% when the net Ah reaches zero *instead* of waiting for the charge factor corrected net Ah reaching zero? My problem is with the first "component" you describe. The second "component" works perfectly, and you are right, it took a few days or even weeks to dial in those settings properly...

Yet another piece of evidence in my strange journey: two days ago I put the flexmax's into EQ. The sun went down before they were done, so the next morning they spun up out of snooze directly into EQ again. I watched the chart plotting the battery net Ah, and charge factor corrected net Ah. It was actually working properly while in EQ, but of course I hadn't intended for the EQ to start in the morning when the batteries weren't even at 100%. So I turned off EQ, back to bulk. As soon as I did that the FNDC reset to 100%.

Pasted_Image_4_14_14__10_12_AM.png


tallgirl wrote:When I was consulting more actively, I never met an FNDC I couldn't get to work correctly -- other than the non-atomic value reporting issue.


What's your hourly rate? I'm willing to pay for someone else to look at my data and tell me how to correct my situation. I'm open to the idea that it's something wrong with me, my batteries, etc. I just want to fix it... I'm not trying to blame outback, it's just that with the data and evidence I have it looks like something is wrong with my FNDC. I'm trying to work with them to figure it out but so far I just get canned standard responses, and it seems like nobody is looking at the data I have and actually explaining the issue.
timmartin
Forum Guru
 
Posts: 273
Joined: Thu Dec 02, 2010 9:31 am
My RE system: ...
Sharp 120W Panels (12)
SunPerfect 230W Panels (9)
Whisper 100 Wind Turbine
Outback FM60 charge controller
Outback FM80 charge controller
Outback VFX3648 inverter/chargers (2)
Outback FlexNet DC
CALB CA400 LiFePO4 3.2V batteries
(32 cells paired together, 1 string = 51.2V 800Ah)

Re: SOC suddenly jumps to 100%

Postby tallgirl on Mon Apr 14, 2014 8:22 pm

Tim,

The SOC jumps to 100% when the charge parameters are met. If that happens to coincide with when your reach 0 uncompensated amp-hours, so be it.

And my hourly rate is $150/hour.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby timmartin on Tue Apr 15, 2014 1:14 pm

tallgirl wrote:The SOC jumps to 100% when the charge parameters are met. If that happens to coincide with when your reach 0 uncompensated amp-hours, so be it.


Yes of course, I but *that* doesn't happen until later in the day in my case(s). Like I said, I fully understand and believe the charge parameters portion of the process to be configured and working correctly. It's the Ah counting and discounting portion of the product that is failing me...

SOC jumps to 100% at 12:55pm:
Pasted_Image_4_14_14__10_12_AM.png


Charge Params Met at 2:05pm:
Screen_Shot_2014-04-15_at_12_53_55_PM.png



So again, SOC jumps at ~12:55pm for unknown reason, charge params met over an hour later at ~2:05pm. I never even realized this issue until I started logging and graphing things. Once I saw the graphs I was very aware of how long the system "sat at 100%" until the charge params were met. My understanding is that my battery charge factor was too high, as this would cause the SOC to reach 100% "too early" so i started lowering it. As i lowered the value over a series of days, the problem become more and more obvious, as the system was *still* jumping to 100% an hour or two before Charge Params Met... for reasons that still elude me and do not appear to follow the "undocumented logic" that was expressed by outback support.

So I'm still looking for answers.

Julie, if you think you could figure it out, and believe with some confidence that the problem is *not* a bug or problem with FNDC, then I'd be willing to pay you for your time.
timmartin
Forum Guru
 
Posts: 273
Joined: Thu Dec 02, 2010 9:31 am
My RE system: ...
Sharp 120W Panels (12)
SunPerfect 230W Panels (9)
Whisper 100 Wind Turbine
Outback FM60 charge controller
Outback FM80 charge controller
Outback VFX3648 inverter/chargers (2)
Outback FlexNet DC
CALB CA400 LiFePO4 3.2V batteries
(32 cells paired together, 1 string = 51.2V 800Ah)

Re: SOC suddenly jumps to 100%

Postby tallgirl on Tue Apr 15, 2014 6:58 pm

Tim,

I did look at your charts. My conclusion was that you didn't have good enough resolution to know what all was going on. What I've seen with systems, especially when they really are close to fully charged, is that the sweep of the array by the charge controller can create a false "parameters met" sort of scenario. In particular, as I understand the "undocumented feature", it can meet that.

What happens at very high states of charge is the density of the acid near the plates has increased to a point that the batteries are "more charged" than they actually are -- ion migration slows down as the specific gravity of the battery as a whole increases.

I'm not sure what your specific issue is, since you're jumping from 98 to 100 percent. The ones that concern me a lot more are the people where the jump is from the lower 90's or even the 80's to 100. In 100% of the instances I've examined, that is caused by incorrect settings.

But as I've said repeatedly -- there are two different "computations" in the FLEXnet DC. The SOC is one, which I mostly ignore. The other is parameters being met. That's the only "value" that is based on physics, and I'm real fond of physics.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby Kent Osterberg on Tue Apr 15, 2014 10:46 pm

As long as the FNDC Charge Parameters Met Time is not set to zero, a scan by the charge controller should not fool the FNDC into believing the charge parameters are met when in fact they aren't. Considering that Outback has acknowledged in private conversations that there is undocumented logic involving the uncorrected amphour total reaching zero and we can see the SOC charge resetting when the uncorrected amphour total reaches zero, we are probably seeing the effects of the undocumented logic.
Kent Osterberg
Forum Emperor
 
Posts: 1503
Joined: Mon Jun 09, 2003 11:01 pm
Location: Cove, Oregon
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules

Re: SOC suddenly jumps to 100%

Postby timmartin on Wed Apr 16, 2014 10:42 am

tallgirl wrote:I did look at your charts. My conclusion was that you didn't have good enough resolution to know what all was going on. What I've seen with systems, especially when they really are close to fully charged, is that the sweep of the array by the charge controller can create a false "parameters met" sort of scenario. In particular, as I understand the "undocumented feature", it can meet that.


I can promise you with 100% certainty this isn't what's happening. Even if something is happening in between my 5 min interval for data logging, the "days since params met" would go to zero if the params were met. And it doesn't... it continues to increase until hours later when params are actually met. I also have the data streaming in every minute (just not recording to the database every data point), plus I can watch it live on the Mate3's web UI and/or the LCD itself.

Kent Osterberg wrote:As long as the FNDC Charge Parameters Met Time is not set to zero, a scan by the charge controller should not fool the FNDC into believing the charge parameters are met when in fact they aren't. Considering that Outback has acknowledged in private conversations that there is undocumented logic involving the uncorrected amphour total reaching zero and we can see the SOC charge resetting when the uncorrected amphour total reaches zero, we are probably seeing the effects of the undocumented logic.


I agree with this as well. My Charge Params Met Time is 10 minutes. I intentionally set my logging interval to half of this to make sure that I'm logging at least two data points that are within the parameters.

That logic *could* be triggered by a resweep. To test that I could change the mode of my charge controllers, right? I have auto restart in mode 2, perhaps 1 would avoid trigger the issue, or certainly 0. That is only if it's actually caused by resweeps.
timmartin
Forum Guru
 
Posts: 273
Joined: Thu Dec 02, 2010 9:31 am
My RE system: ...
Sharp 120W Panels (12)
SunPerfect 230W Panels (9)
Whisper 100 Wind Turbine
Outback FM60 charge controller
Outback FM80 charge controller
Outback VFX3648 inverter/chargers (2)
Outback FlexNet DC
CALB CA400 LiFePO4 3.2V batteries
(32 cells paired together, 1 string = 51.2V 800Ah)

Re: SOC suddenly jumps to 100%

Postby tallgirl on Wed Apr 16, 2014 8:46 pm

Kent Osterberg wrote:As long as the FNDC Charge Parameters Met Time is not set to zero, a scan by the charge controller should not fool the FNDC into believing the charge parameters are met when in fact they aren't. Considering that Outback has acknowledged in private conversations that there is undocumented logic involving the uncorrected amphour total reaching zero and we can see the SOC charge resetting when the uncorrected amphour total reaches zero, we are probably seeing the effects of the undocumented logic.


Kent,

You'd be mistaken. Setting the time to 1 minute is a great way to have issues.

As I understand that "undocumented logic", the "undocumented logic" is correct. My understanding may well be wrong, and OutBack is very unlikely to let me look over the FNDC firmware to figure out if it's right or wrong, but the description of what it does, and how it supposedly does it, make sense.

It's nigh on impossible for a battery to stay near Absorb when it isn't at or near 100% charged =and= not being charged. It takes more than 5% of my amp-hour capacity to keep my batteries at Absorb when the batteries are at 90% SOC, but only 1.5% when they are 100% charge. That's why this "undocumented logic" works.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby timmartin on Wed Apr 16, 2014 10:11 pm

tallgirl wrote:That's why this "undocumented logic" works.


Yeah, except outback told me (and I quote):

Net amp hours > 0 AND Battery volts < ( Absorb setpoint - 0.5 ) AND Average net amps (to the battery) > 0


The voltage has to be BELOW the absorb set point. It sure is easy to get the net amp hours above zero and have the voltage BELOW the set point *AND* it's currently charging... that's an every day occurrence!
timmartin
Forum Guru
 
Posts: 273
Joined: Thu Dec 02, 2010 9:31 am
My RE system: ...
Sharp 120W Panels (12)
SunPerfect 230W Panels (9)
Whisper 100 Wind Turbine
Outback FM60 charge controller
Outback FM80 charge controller
Outback VFX3648 inverter/chargers (2)
Outback FlexNet DC
CALB CA400 LiFePO4 3.2V batteries
(32 cells paired together, 1 string = 51.2V 800Ah)

Re: SOC suddenly jumps to 100%

Postby vtmaps on Thu Apr 17, 2014 12:33 am

I think I've thought of an inelegant way to work around the reset problem. It involves fooling the FNDC into counting the net amphours wrong. You do this by putting up a small string of panels and a small controller on the wrong side of the shunt. Have the small controller shut off at a voltage below Vabs, so that it doesn't interfere with FNDC end amps.

More amphours will be going into the battery than the FNDC knows about. If you size the extra string properly, it should delay the reset until the parameters are met.

--vtMaps
vtmaps
Forum Guru
 
Posts: 111
Joined: Mon Nov 07, 2011 6:09 pm

Re: SOC suddenly jumps to 100%

Postby tallgirl on Thu Apr 17, 2014 6:05 am

timmartin wrote:
tallgirl wrote:That's why this "undocumented logic" works.


Yeah, except outback told me (and I quote):

Net amp hours > 0 AND Battery volts < ( Absorb setpoint - 0.5 ) AND Average net amps (to the battery) > 0


The voltage has to be BELOW the absorb set point. It sure is easy to get the net amp hours above zero and have the voltage BELOW the set point *AND* it's currently charging... that's an every day occurrence!


I'd like to see someone from OutBack post the precise math involved in this "undocumented" feature. I know how to very easily make what you've described happen with a system (multiple days without being fully charged confuse the FNDC because it doesn't track self-discharge). A lot of people don't make it to "Full" every day and every single FNDC ever installed doesn't "jump".

At this point, I'm going to bow out. The solution to FNDC problems isn't messing with the battery charge factor or by adding extra charging sources or otherwise lying to them.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby tallgirl on Thu Apr 17, 2014 7:49 am

I did just verify that, inspite of there only being one FNDC firmware version in production, the behavior that's being described didn't happen yet again. I don't normally at the corrected and uncorrected amp-hours, but did spend about 2 hours today collecting second-by-second data to do that.

And I'll repeat what I've said countless times -- the best way to insure an FNDC works correctly is to program it correctly.
Julie in Texas

I ride bicycles. A lot.
User avatar
tallgirl
Forum Emperor
 
Posts: 4334
Joined: Wed Sep 26, 2007 9:59 am
Location: Austin, TX
My RE system: 16 Kyocera KC175GT (2,800 watts DC), MX-60, 2 OutBack GVFX3648, 8 GC2 batteries (6v @ 215AH),

Re: SOC suddenly jumps to 100%

Postby HawaiianDrummer on Wed Jun 11, 2014 1:41 am

Aloha, Here's my update after resetting FNDC back to around start up settings. Success, Now my hydrometer readings on all cells are daily at 1.275 by early afternoon and float times work again for first time in I can't remember how long! By setting Absorb times down to a reasonable time of around 5 hrs I also get Float times of 1 1/2 to 2 1/2 hrs most sunny days. SO now that the batteries are fully charged "everyday" now in our sunny Hawaiian Summer days, my new concern is OVER charging! Mainly because now for the 1st time I am getting Black residue on the inside of most all battery caps with Dark greyish color of the electrolyte! So I am thinking that now that I have great charging being completed everyday that I might have to set my FM80 Absorb voltage back down to the recommended V around 59.8...for my Trojan batteries ( I have to look that up again) instead of the 60 .4 I have been using in the past.... maybe that is why it is eating my battery plates a little bit now...any other ideas out there? Anyways it is good to see my hydrometer readings back to 100% ish. I also conclude that the FNDC works for most features but that SOC function is more or less to be disregarded. I'm happy again, Thanks OutBack Power! David
HawaiianDrummer
Forum Guru
 
Posts: 323
Joined: Thu Apr 02, 2009 1:42 am
Location: Ninole, Hawaii
My RE system: 1-FM60's, 2-FM80's 2- FX3048's, Mate3, Hub 10, 12 US 64 w panels, 12 Kyocera
175 W panels, 21 Kyocera 225 W panels, 24 Trojan 2V 1,110 AH 48 V Battery
bank, Flex Net DC and Optics RE

Re: SOC suddenly jumps to 100%

Postby BlackCherry04 on Wed Jun 11, 2014 7:53 pm

David, you could be over charging them unless they are filled with 1.280 electrolyte ( 2v Trojans industrial batteries usually are ) . If they have 1.280, you just about right. Are you using a RTS sensor to adjust the charge voltage to temperature ?

1) lower the Absorb voltage by .5 and get a new base line.
2) Adjust the Return amps higher so the FNDC will end the Absorb charge sooner.

Your batteries will last longer if you under charge them by a small amount on a daily basis and then do a short EQ once a month and bring them to 100%.
BlackCherry04
Forum Virtuoso
 
Posts: 87
Joined: Sat Feb 09, 2013 8:50 am
My RE system: Various Different On and Off Grid

Re: SOC suddenly jumps to 100%

Postby HawaiianDrummer on Sat Jun 14, 2014 1:52 am

Aloha, Thanks Blackcherry for your reply, 1st yes I use a RTS. 2nd I will think about lowering the Absorb V setting, BUT here in Ninole, Hawaii I got 3 days of cloudy weather (which is typical) and one of those days was heavy clouds with rain all day, so my hydrometer readings were in the red again, so I am afraid to lower the Absorb voltage setting until it stays sunny....which is unpredictable here.... 3rd I could raise the end amps setting to 2% instead of 1% of my 1,110 AH battery bank to quit charging sooner, Thanks for all your insightful recommendations. When the weather is sunny for weeks, I never need to use my generator, but when it is cloudy I need to turn he generator back on to Auto start and it runs every morning and some evenings if it's been heavily clouded all day, so that my batteries are satisfied (We are heavy load users here off grid....I can bring in 18 to 24 KWH in each good day and use it all up each day with our house and guest house and aquaponics green house and the chicken coop heat lamp, etc. etc.
What I would like to see from Out Back Power is a device that "only turns gen on when it's CLOUDY and needed, BUT NOT when it is Sunny and not needed....so far the FNDC just turns it ON everyday no matter what settings are used......so overkill. Bottom line...OFF Grid High use Sun needed.... Cloudy Gen. needed. The saga continues, stay tuned! David
HawaiianDrummer
Forum Guru
 
Posts: 323
Joined: Thu Apr 02, 2009 1:42 am
Location: Ninole, Hawaii
My RE system: 1-FM60's, 2-FM80's 2- FX3048's, Mate3, Hub 10, 12 US 64 w panels, 12 Kyocera
175 W panels, 21 Kyocera 225 W panels, 24 Trojan 2V 1,110 AH 48 V Battery
bank, Flex Net DC and Optics RE

PreviousNext

Return to FLEXnet DC

Who is online

Users browsing this forum: No registered users and 1 guest