SOC suddenly jumps to 100%

An area for discussing the OutBack FLEXnet DC product.

Moderators: OutBack, OutBack Moderator Team

Lance
Forum Expert
Posts: 53
Joined: Thu Aug 21, 2014 4:36 am
My RE system: 30xCS6P-260M,2xFM80,Radian 8048,GSLC-175PV,FNDC,MATE3,24xL16RE-B

Re: SOC suddenly jumps to 100%

Post by Lance » Tue Jan 06, 2015 7:11 am

Hmmm.. This is really helpful. I'll be all over this when I get back to my system later this week..
The issue I seem to be stuck on is setting the right Absorb target. I have 24 L16RE-Bs that are less than a couple months old.
Trojan says that charged SG should be 1.277.
If I set my return amps to zero and the target to 56.4-ish, the SG never seems to get above 1.260. I've even tried setting the timer to 4 hours.
But, if I raise the target to 59.1, things seems to level off at about 11 amps. That's spot on at 1% (of 1110). I'm told that newer batteries will get there, older ones wont. And the bonus is that SGs get to about 1.275ish.
So given that:
Am I ok doing this providing I keep a close eye on SGs?
Should the other targets be raised proportionally?
Where I live, my days are prone to either no sun or a full day of it. Consequently, I'm either charging off the Radian and supplementing with PV, or tons of PV and trying to sell back by about 1pm. If I could only get the Mate3 to connect all on it's own to start selling - I'd be tickled!

Thanks!
Lance
30xCS6P-260M,2xFM80,Radian 8048,GSLC-175PV,FNDC,MATE3,24xL16RE-B

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

Re: SOC suddenly jumps to 100%

Post by tallgirl » Tue Jan 06, 2015 5:43 pm

Lance,

Use the target Trojan gave you. 56.4 is too low, unless they are AGMs, which they aren't since you can stick them.

Yes, chronically low SG after being "fully charged" is a sign of being under-charged. But charging too high can cause its own problems, so ... use the target that Trojan gave you.
Julie in Texas

I ride bicycles. A lot.

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Wed Jan 21, 2015 12:57 am

My questions were escalated to the software engineering team, the next few posts will document the conversation I've been having with them...
Software Engineering wrote: This could be ÔÇ£Float ModeÔÇØ

It was designed as a way for the FNDC to self-correct drift from the actual state of the battery.
If it is hitting this, the customer may want to review their charge parameters. How old is the battery bank?


But this happens even if the charge parameters is turned off! The battery bank is 4 years old and is still operating well if I disregard the FNDCÔÇÖs behaviors.
Software Engineering wrote: All of the following have to be true for float mode to be triggered.

Code: Select all

Net amp hours > 0 (positive) AND
Battery voltage < (Absorb setting - 0.4v) AND
Average net amps > 0 (positive)
IÔÇÖve gotten this info before (both from outback and second-hand, although each time itÔÇÖs a tiny bit different). This seems illogical to me because that would mean that if the batteries were on their way to being charged (but not all the way there yet) and a cloud passes overhead, this would get triggered. Look, it happened today before the absorb point was ever reached.

At about 12:38 the battery net Ah went over zero, while the Charge Factor Corrected Ah will had 45 Ah to go.
PastedGraphic-4.png
Battery voltage was more than 0.4 volts below the absorb setting in fact it hadnt even gotten up to the absorb setting yet:
PastedGraphic-6.png
But the charge controllers were still pumping in 48 amps of course trying to reach the absorb point!
PastedGraphic-5.png
And this situation caused the SOC and Ah counters to all reset. This essentially happens almost every single day! This cant be right

oddly it would maybe make sense if one of those evaluations was swapped:

Code: Select all

Net amp hours > 0 (positive) AND
Battery voltage > (Absorb setting - 0.4v) AND
Average net amps > 0 (positive)
If the voltage was remaining high (close to absorb) after net Ah was zero and the charger was positive at all.. that kinda would make sense although I still think youÔÇÖd get false positives.

It should be noted that the lower you set your FNDC Charger Factor the more noticeable this behavior is. I have mine set quite low right now because 1) i was trying to get the system to reach 100% SOC at or close to when the parameters would be met and 2) IÔÇÖm really trying to figure out what is wrong with the system and this setting allows me to see things more obviously.

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Wed Jan 21, 2015 1:06 am

Software Engineering wrote:Perhaps this could cause a false positive. I would only expect that to happen if the battery were already near full.
With cloud cover, the current would change, and with it the net amps could easily swing to negative depending on loads.
It could, but you canÔÇÖt depend on the charging amps to go negative. The logic as described is so loose that it would trigger float if the batteries were at 46 volts with 1 amp of charging current. How could that ever be misinterpreted as being a charged battery?

HereÔÇÖs an example from 3 days ago. (this pre-mature float mode has happened 4 days in a row as of today)
PastedGraphic-7.png
Notice that the float triggered at about 3:24, and despite having a positive charging amperage itÔÇÖs pretty clear the batteries were nowhere near being charged on that day. It hadnÔÇÖt yet even gotten close to the setpoint of 58.8 volts.
Software Engineering wrote: The intent of Float Mode was to catch when the battery appears full but the settings are expecting a longer charge.
That is, the FNDC aligning its calculation of the SOC to what it measures coming in and leaving the battery.
This seems in opposition with having users try and set the settings to the proper values. If the system was auto configuring thatÔÇÖd be different, but the users are expected to input the proper configuration information. In my case (and others on the forums) it is causing the FNDC to artificially lie to us regularly about the SOC and net Ah. IÔÇÖd say 8 out of 10 days this happens with my system. And the more consecutive days of not reaching ÔÇ£charge params metÔÇØ the worse the situation gets. You say itÔÇÖs meant to address drift, but I believe in many many cases itÔÇÖs causing drift in the opposite direction.
Software Engineering wrote: With the Charge Factor set low, the FNDC will expect more charging before reaching 100%.
If that is the case, and Float Mode is triggered, we will see a jump of the calculated percentage.

The 48 amps from the chargers is a reasonable return amp value for a battery size of 1110 ahrs (approx. 4%).
Ah, but the problem is that it wasnt 48 amps because thats all the batteries were accepting to reach the setpoint. It was 48 amps because thats the maximum that could be generated from the PV at the time the voltage could be _anything_!
Software Engineering wrote:I recommend raising the charge factor.
That will push the points of Float Mode and Charge Parameters Met closer together, and closer to the actual state of charge of the battery.
Similarly, setting the charge factor high would cause Charge Parameters Met to trigger first.
I dont see how that does anything other than hide the problem. When i arbitrarily set the Charge Factor higher, then the FNDC reports 100% earlier (and possibly avoids the jumping only because were narrowing the statistical window in which it will happen) but then it will stay at 100% for a long time before Charge Params Met. Hours and I felt the intention of the charge factor settings existence was to dial it in so that you reach 100% at roughly the same time that your charge params will be met. Do I have those intentions wrong?

I guess I dont understand why this drift protection needs to be in there at all. You ask the user to input a charge factor and then it is systematically ignored, why ask for it? The system will be reset to 100% when charge params are met isnt _that_ the protection against drift? Id love to understand better the scenarios where this logic was decided to be necessary.

For what itÔÇÖs worth, it wasnÔÇÖt until I started logging and looking at the data that I noticed this. For the first couple years, and when my battery bank was new, I went to work and when i came home it was 100% and params had been met. I had no idea that it was jumping to 100% too early in the day. But weÔÇÖre pushing the system harder now, the batteries are aging (although i wouldnÔÇÖt call them old!) and as IÔÇÖm logging all the data I quickly saw this strange behavior (feel free to browse my data: http://finleyridge.com/power/historical.html). And iÔÇÖm not the only one, there are active threads on the forum regarding this and IÔÇÖve had many other forum members send me private messages asking if iÔÇÖve learned more or had progress in discussing this issue with Outback Support.

Thank you again for your time,
.tim
Last edited by timmartin on Wed Jan 21, 2015 1:16 am, edited 1 time in total.

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Wed Jan 21, 2015 1:13 am

Today it jumped from 72% to 100% while it was still in the midst of charging (at about 2:48pm.) The net Ah hit zero, the chargers were still trying to charge the batteries, the voltage never got even close to the absorb setpoint. How on earth is this proper behavior? I cant use this product to monitor my batteries if its constantly lying about the SOC the generator SOC triggers dont ever work because the SOC is arbitrarily reset and Ive had several low voltage system cut-outs because the other numbers are no longer reliable.

Interactive data graphs: http://finleyridge.com/power/historical ... 2015-01-12

Screenshots of the data graphs:
PastedGraphic-1.png
PastedGraphic-2.png

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Wed Jan 21, 2015 1:28 am

I previously wrote:It could, but you canÔÇÖt depend on the charging amps to go negative. The logic as described is so loose that it would trigger float if the batteries were at 46 volts with 1 amp of charging current. How could that ever be misinterpreted as being a charged battery?
Software Engineering wrote:If the battery were at 46 volts, the net ah would _not_ be near 0 or positive unless the system settings were set wildly different than what the system really needs.
I must be missing something. How are the net Ah and current voltage correlated? To my understanding, net Ah is a counter and since weÔÇÖre talking about the value which is NOT charge factor corrected, it can reach zero or a positive value at practically point in a charge cycle. For instance, IÔÇÖve had a full week of partial sun each day: the net Ah and corrected-net-Ah drift further and further apart the more charging/discharging cycles there are without reaching 100%. So as of today they are about 100Ah different. So net Ah will reach zero 100Ah before the batteries are even close to being charged.

Is my logic correct?
I previously wrote:Ah, but the problem is that it wasnt 48 amps because thats all the batteries were accepting to reach the setpoint. It was 48 amps because thats the maximum that could be generated from the PV at the time the voltage could be _anything_!
Software Engineering wrote:Was there any load on the system?
Yes, this system is always under some load. ItÔÇÖs my primary residence where I live with my family of four.
Software Engineering wrote:Is there a battery temp sensor in the system?
Yup.
I previously wrote:I dont see how that does anything other than hide the problem. When I arbitrarily set the Charge Factor higher, then the FNDC reports 100% earlier (and possibly avoids the jumping only because were narrowing the statistical window in which it will happen) but then it will stay at 100% for a long time before Charge Params Met. Hours and I felt the intention of the charge factor settings existence was to dial it in so that you reach 100% at roughly the same time that your charge params will be met. Do I have those intentions wrong?
Software Engineering wrote:It is not hiding a problem.
Unless you count not knowing in advance which settings will match your system
I have verified my settings over and over. IÔÇÖve had others check them, the installers, other users, forum members, and even Katee and other tech support people said that my settings all appear to be correct. Furthermore, the settings donÔÇÖt change the behavior of this ÔÇ£floatÔÇØ mode.
I previously wrote:I guess I dont understand why this drift protection needs to be in there at all. You ask the user to input a charge factor and then it is systematically ignored, why ask for it? The system will be reset to 100% when charge params are met isnt _that_ the protection against drift? Id love to understand better the scenarios where this logic was decided to be necessary.
Software Engineering wrote:If the FNDC is wrong about the state of the battery, it will trigger Float Mode when the battery appears full.
That is, in this direction, Charge Parameters Met will correct the other way.
Can you explain this more? I donÔÇÖt know what you mean by ÔÇ£the other wayÔÇØ. Charge Params Met triggers when the conditions are such that the system thinks the batteries are charged. What youÔÇÖve told me is that float mode is doing the same or similar thing, but not based on settings IÔÇÖve entered but instead other information.
Software Engineering wrote:Do you have a record of when the Charge parameters were met over the same days in the log?
Yes. The charge params met value is recorded, but it is only ÔÇ£yesÔÇØ for a very short time (it changes to ÔÇ£noÔÇØ as soon as the batteries start discharging). But it records the ÔÇ£days since full chargeÔÇØ so I can query the database to see on which days the parameters are met. IÔÇÖve been meaning to add that to the graphs/charts so that itÔÇÖs easier to see which days they were met.
Software Engineering wrote:I see that sometimes the net Ah and corrected Ah are positive and suddenly drops to 0 net Ah. Such as January 3rd.
That appears to be the opposite of your example from (at the time) 3 days ago. My guess would be Charge Parameters Met.

That has happened a few times that IÔÇÖve seen. The log has ÔÇ£noÔÇØ for every datapoint for Charge Params Met, but sure enough shows the ÔÇ£days since chargeÔÇØ goes to zero at 1:42pm.
Software Engineering wrote: Do you have the Charge Factor over the last week or two recorded?

That value isnÔÇÖt logged, so no. Would it be useful for me to look at that value on the MATE3 periodically and write it down?
Software Engineering wrote: An ideal charge cycle will result in a net Ah of zero.
Batteries are not ideal, and we can adjust for that with the Charge Factor.
If set too far one way, the net ah after a charge cycle could be positive, but we really can't store more than the battery can hold
Yes, I recognize this which is why I was very confused when the Ah counters both went positive and didnÔÇÖt snap back to zero automatically. The following couple of days then showed massively incorrect numbers since it said the batteries were at 100% in the morning after 8+ hours of normal power loads.
Software Engineering wrote:
If we don't pick the right Charge Factor, our meter does not know any better.
Float Mode is there to try and account for that.
I hate to sound like a broken record, but can you explain how resetting to 100% under these conditions is the proper thing to do? IÔÇÖve read every document I could regarding my batteries, proper charging, and IÔÇÖve observed the behavior of my own system for a few years now. The puzzle pieces just donÔÇÖt fit together for me! IÔÇÖd be happy to call and talk to someone on the phone if thatÔÇÖs an easier to way explain it.

IÔÇÖm still coming from he perspective that my settings are all correct, but thereÔÇÖs actually a bug or something wrong with the FNDC. If you were to assume I was right for a moment, what would think was the issue? What could be wrong or incorrect in such a way that FNDC was behaving incorrectly?

.tim

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Wed Jan 21, 2015 1:48 am

tallgirl wrote: Yes, the Net AH was greater than zero when this started. Mine usually is. That's how you can tell you were in Absorb or Float recently.
This practically never happens for me. I get little saw-tooth style graphs as it sometimes starts to go positive but then immediately gets zeroed out. The rare occasion where that hasn't happened and my net Ah's went way into the positive values made the next 24 hours of data horrible ÔÇö i went to bed and woke up and the system was still at 100% SOC since the net Ah were still positive.
sawtooth.png
big positive.png
tallgirl wrote:"Parameters Met" happened at 23:19. The SOC reset fixer thingy doesn't happen until AFTER the current drops. Not before. Chicken, egg, generator didn't stop because the SOC reset.
I want to point out that I turned off "Charge Parameters Met" entirely and my SOC still jumps to 100%. I'm pretty sure I turned off float coordination too. I might try that again since i'm not positive I did.

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Re: SOC suddenly jumps to 100%

Post by Kent Osterberg » Thu Jan 22, 2015 11:29 pm

At least a software engineer at Outback has confirmed the existence of a trigger that resets the FNDC to 100% other than the charged parameters met condition. However, resetting the displayed SOC to 100% at any time other than when the charged parameters are met is a mistake! It is nearly as bad a believing that the FNDC is correct when it is first powered up.

User avatar
Kurt Lundquist
Alpha Energy Engineering
Posts: 1447
Joined: Tue May 31, 2011 7:23 am
My RE system: VFX3524, Mate3, 200Ah 24V AGM Battery Bank
Location: Arlington, WA

Re: SOC suddenly jumps to 100%

Post by Kurt Lundquist » Fri Jan 23, 2015 3:32 pm

What a monster post! I have hesitated to respond on this one because up to this point all I can offer is my opinion. For example I would love it if the FNDC had logic to perform Peukert calculations for more accurate battery capacity readings.

From my experience few people actually use the FNDC as a battery monitor. Most people, (even pre-wired OutBack Systems) use the FNDC as a device monitor. One or two shunts are typically dedicated to RE sources, with a final shunt monitoring the inverter. Power that flows from the RE source to the inverter loads are counted as amps in/out of the battery bank. Realistically the amps from the RE source go path of least resistance and flow directly out on the inverter shunt. The battery bank doesn't really see much if any of that current.

To use the FNDC as a battery monitor you should have one shunt at the battery bank, with all other DC devices on the other side of the shunt. Amp flow on the shunt will be true amps in/out of the battery. I like how the 3 shelf OutBack IBR battery rack has a setup to put one shunt per battery string. You can look at the shunts to see if the strings are balanced or not. To me, that is truly using the FNDC as a battery monitor.

Regardless of the setup I have always considered the FNDC to be like a vehicle gas gauge. The accuracy depends on a wide range of variables, many of which are beyond our control. Over all you get a fair idea of how much juice is left in the tank.

-Kurt
Kurt Lundquist
Renewable Energy Project Engineer
http://alphaenergy.us/

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Fri Jan 23, 2015 5:14 pm

klundquist wrote:From my experience few people actually use the FNDC as a battery monitor. Most people, (even pre-wired OutBack Systems) use the FNDC as a device monitor. One or two shunts are typically dedicated to RE sources, with a final shunt monitoring the inverter. Power that flows from the RE source to the inverter loads are counted as amps in/out of the battery bank. Realistically the amps from the RE source go path of least resistance and flow directly out on the inverter shunt. The battery bank doesn't really see much if any of that current.
Don't use it as a battery monitor? I don't know why else you'd buy it... if you just want to monitor your charge controllers or inverters, they provide that data directly and can be displayed via MATE and logged/plotted/etc. I really don't know why you would install a FNDC and not use it for monitoring your batteries. I suppose you could use it to monitor non-outback components on the DC side of your system. I use it for that as well, with my wind turbine. As you describe, a shunt for one or more RE sources and one on your inverter(s). Typically the solar/wind sources are all input-only. The inverter(s) are mostly output but sometimes are input when charging with a generator. As long as you have a shunt in place for all DC connected items then it works as a battery monitor just fine. It's main display statistic is State of Charge as a percentage. Some people just use ONE shunt, which works fine as long as all DC components are flowing through the shunt to get to the battery. It doesn't give you the ability to see what the different parts of your system are actually doing though. If you have your shunts set up properly the RE shunt should see the incoming RE (solar/wind) current and then also count it on the way back out of the inverter shunt. If you have them connected on the wrong side of the shunts then sure, the current will just flow from the RE to the inverter without getting counted. That doesn't make sense, why would you hook them up that way?
klundquist wrote:To use the FNDC as a battery monitor you should have one shunt at the battery bank, with all other DC devices on the other side of the shunt. Amp flow on the shunt will be true amps in/out of the battery. I like how the 3 shelf OutBack IBR battery rack has a setup to put one shunt per battery string. You can look at the shunts to see if the strings are balanced or not. To me, that is truly using the FNDC as a battery monitor.
The per-string method is interesting, but personally I like to know more about the different RE sources etc.
klundquist wrote:Regardless of the setup I have always considered the FNDC to be like a vehicle gas gauge. The accuracy depends on a wide range of variables, many of which are beyond our control. Over all you get a fair idea of how much juice is left in the tank.
Yes, that's all it is. Just counting the electrons going in and out. But properly configured it is pretty useful since voltage while under load isn't a great indicator and SG is a pain to constantly measure.

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Re: SOC suddenly jumps to 100%

Post by Kent Osterberg » Fri Jan 23, 2015 5:43 pm

up to this point all I can offer is my opinion.
Does this mean that Outback won't officially be disseminating any information on this topic? Have the gags been put on?

Considering that this topic has been active on the forum for three years now with virtually no reply from Outback, dribbling a little information to folks who call is inappropriate. It is time for Outback to step up and explain precisely and clearly what is going on.

The responses from the software engineer are not clear and concise.
Software Engineering wrote:All of the following have to be true for float mode to be triggered.

Code: Select all

    Net amp hours > 0 (positive) AND
    Battery voltage < (Absorb setting - 0.4v) AND
    Average net amps > 0 (positive)
Is this truly representative of the code causing this? It absolutely makes no sense to reset the indicated SOC to 100% under these conditions. It bypasses the conditions the user is trying to establish with the charged parameters.
Where does the FNDC get the absorb setting? Does it query it from some other device in the system? Which one?
Is the 0.4V set in the code?
If so why does the user input the charged voltage parameter?
Software Engineering wrote:If the FNDC is wrong about the state of the battery, it will trigger Float Mode when the battery appears full.
That is, in this direction, Charge Parameters Met will correct the other way.
Someone is not keeping track of signs very well. Charged parameters met removes an accumulation of negative adjusted amphours and so does the reset to 100% that we see in the data.

User avatar
Kurt Lundquist
Alpha Energy Engineering
Posts: 1447
Joined: Tue May 31, 2011 7:23 am
My RE system: VFX3524, Mate3, 200Ah 24V AGM Battery Bank
Location: Arlington, WA

Re: SOC suddenly jumps to 100%

Post by Kurt Lundquist » Mon Jan 26, 2015 7:52 am

timmartin: I'm sorry I was not more clear with my post. Everyone uses the FNDC as a battery monitor. What I meant to say is there are two different ways to install the FNDC. In your system the FNDC is set up to monitor the flow of DC current from RE sources to the batteries AND to the inverter. The battery side of your shunt bus allows amps to flow almost directly from the RE source to the inverter. The amps are not necessarily flowing into the battery. Consider a grid tied system that is selling all the RE. Amps flow in on the RE shunt(s) and then they either go to the battery or the inverter. Charged batteries have high resistance so most of the amps are actually going out on the inverter shunt without going into the battery. The problem is the FNDC does not know the amps are not really going into the battery. In this setup you have greater visibility of RE/Inverter activity and less visibility of amps into your battery. That is why I said this is more like using the FNDC as a device monitor than a battery monitor.

If you only have one shunt, or one shunt per battery string, the amps that flow across the shunt are always going into the battery. To use the grid tied example again, amps can flow from RE source to inverter without going through a shunt. This is a subtle difference in setup but I think it does have an impact on accuracy.

Kent: I do not work for OutBack. Your use of the word gags made me smile. I agree with you in that it does not seem like the FNDC is a high priority item for the engineering team. I'd rather see some NEC 2014 compliant gear like Arc Fault Combiners and a Rapid Shut Down system. Regardless I hope that the points in this thread are considered for any FNDC updates or future products.
Kurt Lundquist
Renewable Energy Project Engineer
http://alphaenergy.us/

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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%

Post by timmartin » Tue Jan 27, 2015 9:51 am

klundquist wrote:If you only have one shunt, or one shunt per battery string, the amps that flow across the shunt are always going into the battery. To use the grid tied example again, amps can flow from RE source to inverter without going through a shunt. This is a subtle difference in setup but I think it does have an impact on accuracy.
That really depends on how the FNDC does it's math. If you have your charge factor set to 100%, for the sake of argument, then it wouldn't matter which way you have your shunts set up ÔÇö as long as all charging and loads are properly being measured. This is because if your RE is charging with 50 amps, and you have 25 amps of load, then you'll see 50 - 25 = 25 net. So the system will properly count only 25 amps going into the battery.

However, with the charge factor set to anything realistic or reasonable, the FNDC is "discounting" the incoming charging amperage. If it discounts positive shunt values PER shunt then you are absolutely right. Let's show 75% charge factor to make the math easy:

Code: Select all

(50 charging amps * 75% charge factor) - 25 load amps = 12.5 net amps to battery
But what it should be doing (and i hope it is!) would be to add the shunt values together FIRST and then apply the charge factor if the net value is positive.

Code: Select all

(50 charging amps - 25 load amps) * 75% = 18.75 net amps to battery
If you had a single shunt/battery-only configuration your shunt would read 25 amps and it would be discounted to 18.75 just the same. But in that configuration your data logs, charts, graphs, etc wouldn't show any information about your loads or what was happening etc.

I personally find the detailed information very useful, I can look at the daily chart of my system and I can literally see when the pressure pumps turns on/off and when it's meal time and the microwave or toaster oven gets used. I can sure as hell see when a major appliance like the dishwasher or washer/dryer is getting used and It's very helpful to troubleshoot or make decisions about when to do certain chores.

timmartin
Forum Guru
Posts: 278
Joined: Thu Dec 02, 2010 8:31 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)

Most egregious jump yet.

Post by timmartin » Sun Jun 14, 2015 8:13 am

Two days ago I had a problem that caused my system to decline into shutdown. The SOC and voltage dropped to the point where the system shut off to preserve the batteries. This was my own fault... but what happened the next day was not.

I charged on solar for the first part of the day, while I hooked up my new generator which has been sitting there unused. When I finished, at about 3:08 PM, I ran the generator and everything was working perfectly.

After charging for a few hours, at 6:26 PM, I turned off the generator. At this point SOC read 76%, remaining Ah required for charge was ~260 Ah (corrected value). And the chargers were holding the absorb voltage (58 V) with about 34 amps. For reference, my end return amps is 12.

So, within moments of shutting off the generator, the voltage dropped as expected and everything seemed good. Well, except that FNDC decided to jump the SOC to 100% all of a sudden.

*scream* I hate hate hate this bug, and Outback doesn't even acknowledge it exists!

So yes, the net amp hours was positive. Yes of COURSE the voltage was below the absorb setting. And during this transition, YES the net amps was positive because the sun was still out and I was still getting some amount of charging.

This is wrong wrong wrong. I just need 30 min on the phone with one of the engineers.
6-13 jump.png

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Re: SOC suddenly jumps to 100%

Post by Kent Osterberg » Sun Jun 14, 2015 9:44 am

*scream* I hate hate hate this bug
Me too! And the other bugs in it.
It is high time that Outback recalls the FNDC and fixes every one of them.

vtmaps
Forum Guru
Posts: 111
Joined: Mon Nov 07, 2011 5:09 pm

Re: SOC suddenly jumps to 100%

Post by vtmaps » Mon Aug 17, 2015 8:41 am

Bill Gordon wrote: My opinion of a fully charged battery is when the "Battery Charge Factor compensated Net Amp Hours" have been replaced.
That's not how I would want to define 100% SOC. I want charge parameters met to define a fully charged battery. I don't care what the coulomb counter says. The battery charge factor is a moving target and I don't trust it to estimate net amphours by counting coulombs.

This thread is about a bug that Kent and others have reported, where the SOC on the Flexnet can jump to 100% without parameters being met.

--vtMaps

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Re: SOC suddenly jumps to 100%

Post by Kent Osterberg » Tue Sep 29, 2015 9:13 am

Bill,

It's been three weeks since your post regarding Outback's tests. Please let us know if Outback has figured out anything yet. I'm particularly interested in hearing if they are actually going do do anything.

jcheil
Forum Expert
Posts: 54
Joined: Fri Mar 21, 2014 6:55 am
My RE system: Off-Grid in Central Florida since 2005, Full-Time since June 2014 | 12 X Sovello 205w panels, 9 X ToPoint 220w panels, 36x ToPoint 225w panels (12,525 watts total) | Custom built single-axis ground mounts | Complete FP2 Outback System: 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1110ah @ 48v | Honda EU7000is Generator and a pile of "other" Generators | Home-Made PVC solar hot water collector | Custom data logging software http://www.somewhatcrookedcamp.com/monitormate.html
Location: Holopaw, FL

Re: SOC suddenly jumps to 100%

Post by jcheil » Mon Oct 26, 2015 1:32 pm

I also agree the bug is horrible. it actually defeats the purpose of the FNDC. I can't even trust it half the time.
pretty much EVERY day mine will jump from somewhere in the low 90%'s to 100%.

However, even if OB does "fix" the code/bug, I am not sure how they "distribute" the fix to everyone easily.
Because from what I am told, the FNDC cannot have the firmware remotely updated. So that means sending our units back to OB for the new firmware like the inverters and charge controllers.

And I did get confirmation of this "undocumented feature" in an email from OB about a month ago:

Diane Lawrence <dlawrence@outbackpower.com>
Sep 25

to me
Hello,

The FLEXnet DC has undocumented logic which can cause a jump to 100% SOC.

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

Adjusting charge factor is one way to ensure the smoothest possible transition to 100%.


Best Regards,

Diane Lawrence
Technical Support, OutBack Power Technologies
Mailing address: 17825 59th Ave NE, Suite B, Arlington, WA 98223
Shipping Address: 17827 59th Ave NE, Arlington, WA 98223
360.435.6030 Main | 360.618.4363 Tech Support
www.outbackpower.com
OutBack Training Videos
11,175watts of Solar | 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1,100ah @ 48v | Honda EU7000is Generator | Live data [url]http://www.somewhatcrookedcamp.com/monitormate.html[/url]

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

Re: SOC suddenly jumps to 100%

Post by tallgirl » Sat Feb 27, 2016 6:24 pm

I'll repeat what has been said countless times before - check your charge factor.

Happy fun "bug" happens when your net amps into the batteries (and the FNDC knows how to calculate that and it also reports it in the data stream -- it's the Charge Factor Corrected Net Amp Hours) is positive. If it isn't really positive (because you're not really at 100% SOC ...), lower your charge factor.

I do believe there's a bug, but I believe it's in how CFC net amp-hours is computed, but this thing y'all keep saying a bug seems much more related to an improperly set charge factor than anything else.
Julie in Texas

I ride bicycles. A lot.

jcheil
Forum Expert
Posts: 54
Joined: Fri Mar 21, 2014 6:55 am
My RE system: Off-Grid in Central Florida since 2005, Full-Time since June 2014 | 12 X Sovello 205w panels, 9 X ToPoint 220w panels, 36x ToPoint 225w panels (12,525 watts total) | Custom built single-axis ground mounts | Complete FP2 Outback System: 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1110ah @ 48v | Honda EU7000is Generator and a pile of "other" Generators | Home-Made PVC solar hot water collector | Custom data logging software http://www.somewhatcrookedcamp.com/monitormate.html
Location: Holopaw, FL

Re: SOC suddenly jumps to 100%

Post by jcheil » Sat Feb 27, 2016 6:43 pm

I disagree that the charge factor cause it. I have shown that it happened EXACTLY as the outback support documentation many times in my system. You can see those exact things happening and poof it goes to 100% regardless of charge factor and I have tried numerous charge factors from 70% to 95%. It doesn't change anything other that how quickly or slowly the bug occurea each day.
I log data every minute online (link below) and you can see for yourself. And Tim has show this on his system. Data also.
11,175watts of Solar | 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1,100ah @ 48v | Honda EU7000is Generator | Live data [url]http://www.somewhatcrookedcamp.com/monitormate.html[/url]

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

Re: SOC suddenly jumps to 100%

Post by tallgirl » Sat Feb 27, 2016 10:40 pm

jcheil wrote:I disagree that the charge factor cause it. I have shown that it happened EXACTLY as the outback support documentation many times in my system. You can see those exact things happening and poof it goes to 100% regardless of charge factor and I have tried numerous charge factors from 70% to 95%. It doesn't change anything other that how quickly or slowly the bug occurea each day.
I log data every minute online (link below) and you can see for yourself. And Tim has show this on his system. Data also.
Then it's not the bug described above because your SOC can't be anything but 100% if net amp-hours is greater than 0. The state of charge should be (battery capacity minute charge factor correct amp-hours) / battery capacity * 100.

There's supposedly another bug with "Parameters Met", and that's the other thing that resets your SOC to 100%.

I'm not saying the FLEXnet DC isn't as buggy as a picnic lunch in the middle of a swamp, but I have seen WAY more of them programmed incorrectly than correctly. The biggest problem I see is the "Parameters Met" values being just plain wrong.
Julie in Texas

I ride bicycles. A lot.

jcheil
Forum Expert
Posts: 54
Joined: Fri Mar 21, 2014 6:55 am
My RE system: Off-Grid in Central Florida since 2005, Full-Time since June 2014 | 12 X Sovello 205w panels, 9 X ToPoint 220w panels, 36x ToPoint 225w panels (12,525 watts total) | Custom built single-axis ground mounts | Complete FP2 Outback System: 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1110ah @ 48v | Honda EU7000is Generator and a pile of "other" Generators | Home-Made PVC solar hot water collector | Custom data logging software http://www.somewhatcrookedcamp.com/monitormate.html
Location: Holopaw, FL

Re: SOC suddenly jumps to 100%

Post by jcheil » Sun Feb 28, 2016 7:07 am

Well in the testing I have done, and in my case, I rarely if ever get to "parameters met". So it's not because of that.
But again, when those exact sequences of events shown above takes place, regardless of any other setting in the FNDC it immediately jumps to 100% SOC.

I'm getting to the point where I can almost predict it based on the time of day and the amount of sunshine out.
11,175watts of Solar | 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1,100ah @ 48v | Honda EU7000is Generator | Live data [url]http://www.somewhatcrookedcamp.com/monitormate.html[/url]

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

Re: SOC suddenly jumps to 100%

Post by tallgirl » Sun Feb 28, 2016 8:13 am

RIght, but if your "charge factor correct net amp-hours" are greater than 0 when that happens, the corrective action is lowering your charge factor so your charging amps are discounted more, and you reach 0 CFC net amp-hours after still more charging.

What is your "Parameters Met" current as a percentage of capacity? For a healthy battery bank it should be on the order of 1-2 percent. Generally speaking, the lower you can set it, the better. The "Parameters Met" bug that gets discussed sometimes (which is very different from THIS "bug") is caused by high-speed change in irradiance and the voltage not being high enough that the "parameters" are met before the battery voltage can drop below the value. Others have said that the voltage merely being ABOVE the setting, followed by the current being BELOW its setting, causes that bug to be trigger.

And again, not saying the FNDC isn't buggier than a picnic lunch in the middle of a swamp.
Julie in Texas

I ride bicycles. A lot.

jcheil
Forum Expert
Posts: 54
Joined: Fri Mar 21, 2014 6:55 am
My RE system: Off-Grid in Central Florida since 2005, Full-Time since June 2014 | 12 X Sovello 205w panels, 9 X ToPoint 220w panels, 36x ToPoint 225w panels (12,525 watts total) | Custom built single-axis ground mounts | Complete FP2 Outback System: 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1110ah @ 48v | Honda EU7000is Generator and a pile of "other" Generators | Home-Made PVC solar hot water collector | Custom data logging software http://www.somewhatcrookedcamp.com/monitormate.html
Location: Holopaw, FL

Re: SOC suddenly jumps to 100%

Post by jcheil » Sun Feb 28, 2016 8:25 am

My charge factor is at 78%. Which after years of trial and error seems to be the best number. With that setting my SG is pretty much dead in line with the SOC indicator on the mate throughout the day. If I set it any higher, it gets to 100% SOC per the display, but the SG is still below 1.280.
The only negative I have found haing it that low is that it takes forever to get from 90 to 100% SOC. Mostly because the batteries are only pulling in 5 amps at that point.
11,175watts of Solar | 3 x FM80, 2 x VFX3648, X240 Transformer, FLEXnet-DC, Mate-3, Hub-10, FW500 AC/DC | 24 x Trojan L16RE-B Batteries 1,100ah @ 48v | Honda EU7000is Generator | Live data [url]http://www.somewhatcrookedcamp.com/monitormate.html[/url]

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

Re: SOC suddenly jumps to 100%

Post by tallgirl » Sun Feb 28, 2016 8:35 am

What is your "Parameters Met" voltage and current? Five amps into an 1,100 amp-hour bank is fully-charged. Ten amps into that size bank is fully charged ...
Julie in Texas

I ride bicycles. A lot.

Post Reply