Charge Conrollers not staying in Absorb cycle

Anawa

As I get my new system up, running, and de-tweaked, I'm concerned that the FM Controllers are "short cycling" through the absorb stage. I've been online with the PV system now for about 8 weeks. I scrolled back through the Mate3 data log and noticed that most days the absorb cycle has only been 20-30 minutes even though the bulk stage was 1-3 hours. The system seems to go from bulk-to-float with only a courtesy-wink at the absorb stage. To date, my average range for DoD is 87-92%. Obviously very shallow and where I wanted the usage to be.
I've taken SG readings for the last couple of days and found readings around 1.223, thus my batteries are not 100% SOC even though the Mate3 reports as such. My last (and only) equalization was on 4/21/2015. Also, the SG readings have not been taken after a 3 hour rest on the batteries. The reason for not resting the batteries is that I have some ongoing construction and refrigeration that would be disrupted by doing so. I know that the SG readings need to be accurate and I will follow proper protocol, but I was thinking that first I would need to resolve the Absorb stage issues with the CC's.

I would like some advice in resolving these issues.

Thanks
Paul

benthere

It sounds to me like "Absorb End Amps" is stopping your absorb. I tried that setting for a couple days on my FM80 then disabled it because my SG also dropped, IIRC.

Anawa

Benthere, the End Amps setting was at zero (disabled?) until I set it at 10 amps yesterday. Waiting to see what happens with absorb stage today. Right now, it's in bulk.

Blackcherry04

Anawa wrote: »

Benthere, the End Amps setting was at zero (disabled?) until I set it at 10 amps yesterday. Waiting to see what happens with absorb stage today. Right now, it's in bulk.

Go advanced and set the FX's absorb time to say 3 hrs. I think there may be a glitch even though you have set the Mate / FNdc to control the charge cycle and ending amps. I had the same problem and thats how I solved it, then the Mate settings worked.

Vic

Hi Anawa,

OK, from your Signature, it appears that you do not use the FNDC -- which has some reported problems in the way it functions in its ending Absorb.

What is your setting for Absorb voltage, and Absorb time?

And, yes, with the FM CC's EA set to 0, the EA function is disabled.

You note that you are using an "industrial" battery, is this battery the Trojan Ind battery ... this might have come up in another Thread ... forget.

Thanks, Vic

Anawa

Vic, pretty sure I do use the FNDC, at least, I am not aware of any settings that have disabled it. The Mate3 reflects that it's being used and the system was factory wired (Outback FP-1).

For both FM's: absorb = 57.8V, absorb time = 3.0 hrs, absorb end amps = 10 amps.

Batteries are 804 ah G/B Industrial FLA's.

BTW, batteries already in float at 11:20 AM, absorb time today = 13 minutes. DoD was 87%.

Blackcherry, I'll check-out the absorb setting in the FX

Paul

Anawa

The FX absorb was 1:00 hr. Reset to 3.00 hr to match FM's.
Paul

Vic

Hi Paul,

Believe that the GB battery is a Forklift battery. These usually run 1.280, or even higher SG electrolyte. These batteries often like fairly high Vabs, in the range of about 60 volts, or so.

On the Outback Forum, there have been extended discussions of some of the issues with the way that the FNDC ends Absorb. You might want to review some of these comments:

http://www.outbackpower.com/forum/viewforum.php?f=28&sid=bdc510d6162e30935c0ab11e27044257

FWIW, Vic

Anawa

Vic, you are correct about the G/B's, they are technically forklift batteries. The 57.8 absorb voltage is G/B's recommended setting.

Thanks for the Outback links, I'll check them out.
Paul

Anawa

Vic, you're also correct about G/B's SG being 1.280.

Vic

Hi Paul,

AM sure that GB knows their batteries. BUT, the customary charge cycle with Grid-powered chargers differs considerably from what we off-gridders with PV power must do. Usually this means that, even in the Summer, that a Vabs of about 60 or so Volts is needed, especially with an SG of 1.280.

So would do just what you are doing; increase the Absorb time, and leave the Vabs as it is, and see what happens. The FNDC appears to have some bugs, and this might be an additional variable in how well your battery gets charged.

Have only run the OB MX-60s, so cannot comment on the FNDC performance from direct experience. Good Luck, Vic

zoneblue

Your DOD is small, so your absorbs wont be long however, the key thing is that the absorb stage completes properly. You dont really want to be guessing at these settings. Take a clamp meter and taking current readings every 20 minutes through absorb. From this curve you can see what your end amps figure should be. And whether its time or amps that the controller is triggering on.

Anawa

ZB, one issue I have is that taking a reading on absorb is almost impossible. For instance, today the cycle lasted only 13 minutes and there are many days that the absorb time is 0 minutes!

I'm reviewing the Outback forum comments on this issue and right now believe that a place to start is to get my batteries to 1.285 SG and then reset the "Charge Parameters" in the FNDC.

QUESTION: I know that the accuracy of the SG reading is impaired when the battery is under load, but is it also impaired similarly during charging? In other words, during an equalization cycle, are intermittent SG readings accurate? I ask this because I think I will going through equalization(s) to get the SG level to where it needs to be and I do not want to be shutting down the system (removing all loads) during the process.

Thanks
Paul

zoneblue

You need to temporarily set the abosrb time to 10 hours for your test cycle. If need be disable the fndc, or set EA to a very low amount. You need to find out what it triggering absorb termination. Here our bank is lightly cycled like yours at least 3 out of 4 cycles. Absorb is around 2.5 hours. For deeper discharges is might run trhough 4 or 5 hours. Its very important that you complete absorb, as each cycle that doenst complete permanently locks up hard sulphate and the batterys capacity reduces by that amount. However if you run longer float times in the afternoon, that can make up for it to a fair extent.

Are you sure the end amps is actually running off the battery current? It might be that your loads are confusing it.

Vic

Anawa wrote: »

ZB, one issue I have is that taking a reading on absorb is almost impossible. For instance, today the cycle lasted only 13 minutes and there are many days that the absorb time is 0 minutes!

I'm reviewing the Outback forum comments on this issue and right now believe that a place to start is to get my batteries to 1.285 SG and then reset the "Charge Parameters" in the FNDC.

QUESTION: I know that the accuracy of the SG reading is impaired when the battery is under load, but is it also impaired similarly during charging? In other words, during an equalization cycle, are intermittent SG readings accurate? I ask this because I think I will going through equalization(s) to get the SG level to where it needs to be and I do not want to be shutting down the system (removing all loads) during the process.
Thanks
Paul

Paul,

First, your target SG is probably 1.275 or so, as Hydrometer readings are probably +/-- about 5 points.

It sure seems like the FNDC may well be screwing with your Absorb end. And its report of SOC may be inaccurate, because its reset to 100% may not be accurate. of it does not allow a complete Absorb to occur.

SG readings during Absorb should be accurate, as are those when you have just entered Float, or during EQ. NO resting should be required for accurate G readings, and this is good, as few off-grid systems would allow resting the battery for hours in order to be able to take accurate SG readings.

Believe that the FNDC will report accurate battery charge currents, via a Shunt. So, you should have an accurate reading on the charge current going into the battery.

Have never had any problem feeling that I was getting accurate SG readings during Absorb, EQ, or in Float. In Float, if the Vfloat is not accurately set, to just, barely keep the battery charged, then as the time in Float increases, it is possible that the battery might not be quite as well charged as when the transition to Float just occurred.

Glad that you are reading comments on the OB Forum, regarding the operation of, and possible bugs with the FNDC. Have not used one, so cannot comment from personal experience.

Your battery is young, and it is Flooded, so it is good that you are paying attention to the battery. Good Luck, Vic

Photowhit

Anawa wrote: »

. My last (and only) equalization was on 4/21/2015.

With Fork lift type batteries you MUST equalize every 30 days and sometimes more often. So once you get the battery fully charged...

Dave Angelini

Force a bulk charge until you get your SG going the right way. Just curious, with a cooling load where you live, are you sure you are not pulling down the voltage during absorb? Are the SG readings similar on all the cells? If this has been mentioned by others I did not have time to read.

Anawa

Dave, I don't think my cooling load is a factor in what's going on with the Absorb dysfunction. Our current weather conditions have been very mild and the mini-split only runs in the afternoon, after the Absorb stage has either terminated or was skipped over. Yesterday, 0.00 hrs for the Absorb stage. Now that I've got some other chores out of the way and weather permitting, I will be going through a battery maintenance routine ..... equalization, bulk, or what ever necessary to get the SG up to 1.285 (G/B battery target). Vic, I read through a bunch of the Outback forum comments on the FNDC and discerned that yes, there are probably issues with how it regulates the battery status, but with so much techno-jargon and engineering agenda issues, I got lost in how all of this pertains to my situation. Nonetheless, I believe the time I spent reading was worth the effort as my off-grid education was further expanded. Not sure it was enhanced, but maybe down the road I will have one of those "Ah-Ha" moments and remember some of the stuff I read. Thanks again for the link. My current plan is to commence with getting the SG up to 100% SOC and then re-setting the FNDC charge parameters using Outback's documented recommendations. I thought I would start today with a 3 hr Equalization charge and each hour record the SG's in all the cells. Anyone that would like to weigh-in on this process .... I'm listening. Thanks Paul

zoneblue

DO you posess a dc clamp meter? Takes a lot of the guessing out, so after it kicks to float, check the charge acceptance rate. Keep checking to see if it reduces significantly. Sure sign that absorb was not done.

Anawa

ZB, tell me more about using the clamp meter to monitor the absorb cycle. Being a newbie at all this and at the risk of being just plain rude, a bit more specific instructions are greatfully appreciated.

I'm assuming the objective here is to accurately measure current and not voltage. If this assumption is correct, then what will a significant decrease in current tell me why the absorb cycle is not done? Most days, the absorb cycle is 0.00 hrs which tells me the absorb cycle never started, thus it never completed either. Not being argumentative, just what's happening to my system.

Right now I'm trying to get my batteries to 100% SOC (measured by SG) through forced equalization and/or forced bulk. Do you have any recommendations on this process, another EQ or bulk? Yesterday the starting SG was 1.220 (most) and after a 3 hr EQ, they increased to 1.240. The SG variance at the start across all cells was 1.217 to 1.230, with most cells at 1.220. My SG target is 1.285 per manufacturers recommendations.

I'm also assuming that the process you're suggesting occurs after a "normal" daily charging cycle? Should I continue working on the 100% SOC and then measure the current?

Thanks
Paul

Vic

Hi Paul,

I am NOT zb,

BUT, I believe that he is saying that if there is a significant amount of current INTO the battery at Vfloat, and this current diminishes over a course of hours, then there is battery charge going on, when in Float. And that this means that the Absorb stage was insufficient compared to the needs of the battery.

And, of course, if you can see that there was NO Absorb for a day, this just has to be true, even just from self-discharge of the battery.

I have not studied all of the Posts on the OB Forum for a number of months, but, it seems to me from past readings, that the FNDC has significant BUGs, and this appears to be detrimental to battery charging.

Personally, I would take the FNDC out of the picture, at least until you can reliably get the battery fully charged almost every day that you have good sun.

The Clamp DC Ammeter will allow you to measure the actual battery charge current. BUT, the main draw-back with the Clamp DC ammeter, is that the Zero is unstable, and these inexpensive versions of these meters drift significantly over time, and need to be re-zeroed a number of times per minute, in my experience.

The FP1 must have a shunt in the negative lead of the battery, for the FNDC to use. This shunt will have a voltage that is proportional to the DC current flowing into the battery. The only problem in trying to measure this battery charge current with a voltmeter, is that at the finishing currents that you are interested in represent hundreds of microvolts, which most DMMs cannot measure accurately.

Perhaps, even with the FNDC disabled for charge termination, there is a way to see what the battery charge current is during Absorb and Float ... dunno, have never used any OB hardware, except the MX-60

FWIW, Keep at trying to get the SGs up. Try forcing a Bulk, and then, you need to get an Absorb stage done. You can do this Absorb by manually adjusting Float voltage to the desired Absorb voltage, if the system will not allow a normal three or four, or five hour Absorb. It is better to manually do the Absorb, than to just go to EQ, as the battery is not getting fully charged, when there is a very short or, NO Absorb.

EDIT:

AND, if the loads on the inverter/s are light, then, the output current shown on the FM CCs, when totaled should mostly be current into the battery. Near the end of Absorb, you could probably shut off one FM, to make this CC output current easier to read, as there could be enough PV power from one PV array and one CC to continue the Absorption (or simulated Absorb0 from setting the Vfloat up to the desired Vabs ... unless this is so late in the day, that there is too little PV power from just one PV array/CC ... FWIW ...

More later, Good Luck, Vic

Anawa

Thanks "Not ZB" (Vic) for the response. Well, after a "normal" cycle and the CC's went straight to Float (0.00 Absorb) around 11:30 A.M., the manual forced bulk didn't work. Right now, I've reset the float to 58.0 volts and will let it run for a couple more hours and I'll measure the SG again.

Something I noticed with the FM's is that when port 3 was in Bulk, port 4 was in Absorb. Not sure how long this was out-of-sync, but is this a concern?

Paul

Vic

Hi Paul,

Thanks for the added info.

The OB MX-60s had what was a Bug, in my opinion, and the FMs probably still have it ...

IF you are using CC EA to end Absorb, and you TRY to Force Bulk, the CC will go to Bulk, but, if the CC is delivering less than the EA setting in the MX (and probably the FMs too), then Bulk will end instantly, IIRC. YES, this logic seems incorrect, but only when Forcing Bulk from Float. EDIT SO, it seems that Force Bulk will not work if the CC is in Float (MX-60s, anyway) when the CC output current is below the CC EA setting. So here, in the past, have turned on an electric heater, which demands about 20-ish additional Amps from the CC, which is above our customary shunt EA of 12 A. Then, Force Bulk works fine from the Float Stage ... at least this is what I believe was the sequence, and just my feeling of what was happening with the OB MX-60s. ... end EDIT.

The MidNite Classics seemed to use this same logic, IIRC. One fix for this for the MX (and Classics) was to turn on an AC load for the inverter, that was larger than the CC EA setting, Then, Force Bulk, and then, the CC would actually run Bulk then Absorb and ultimately Float, if the day was long enough.

You might try this, or set CC EA very low, but not zero, and see if this does the job.

This is from somewhat distant "memory".

Setting Float to the desired Vabs is simple, too.

Now, regarding your Vabs; This is really too low for off-grid charging. It is possible, and have not looked at the GB charging recommendations in several years, BUT, most of the Industrial/forklift battery manufacturers specify a fairly high Bulk voltage for some time period, followed by a longer, lower V for Absorb. But, IMO with a 1.280 or 1.285SG electrolyte, using PV charge sources, the day may not be long enough to get a full charge done, when the battery is discharged a considerable amount.

Personally, if this battery were mine, would raise the Vabs to about 60 or so volts. This is not EQ territory, and this will result in more charging per hour, and seems from your SG readings, you do need to catch up. These Forklift batteries can accept a lot of current during the early/mid portions of Absorb, and you are probably PV limited, anyway.

EDIT: And there are several members still here who use Forklift batteries, some of them do use GBs, at that. So please listen to Photowhit, and those with experience with FL batts.

Just opinions. Good Luck, Vic

zoneblue

While there is some urgency to get the SG up, just remember to factor that some L16 form factor batteries are known to take a few dozen cycles to easily reach the full target SG. Rolls is one brand known to do this in quite a stuborn way. But you do need to perserve even if it means increasing the absorb setpoint and time a bit during this break in period.

What i was trying to explain was about how a battery charges. During bulk, the battery will take all you can throw at it, and the current is limited in practice by the Imp of the PV array that day. Once the absorb setpoint is reached things change, and the controller prevents the voltage rising, As it does this the current will begin to decrease in a concave manner. This is the classic charge acceptance taper. The only blackbox image that i can readily lay my hands on is at the bottom of this page: https://code.google.com/p/theblackboxproject/

The bottom graph blue curve is the actual battery current, and while it wasnt a clear blue day, eventually you see the taper take effect around 11.30. Because the weather was a bit frumpy the controller kicked in and out of absorb for a while which you can see on the top graph red curve. Its that battery current "charge acceptance" curve that i suggest is important to get to know, for your bank. It will follow this same curve each day, albeit sometimes longer sometimes shorter depending on the initial SOC. YOur battery documentatiion shoudl state a tpyical end amps figure, if not you will have to resort to average values for the battery type. For AGM, the typical figure is 0.5% of the bank capacity. So for example i have 400 Ah bank, 0.5% of 400 is 2Aamps. In practive i add 0.5 amp to this to give the controller some margin for waveform noise and set my end amps to 2.5 amps.

Now if the battery does not absorb, but goes straight to float, the charge acceptance taper still keeps tapering, just at a slower rate owing to the lower voltage setpoint. The critical thing is that (regardless of charge stage) the taper tapers until its flat. Thats the point when the battery accepts no more current, and is the point when the battery is fully charged. The actual process of absorb always has a higher voltage for several reasons. (such as electrolyte stirring, and speeding up the charge process with limited daylight etc) But the important thing is that the bank reaches that point of accepting no more current. If not daily then a couple of times a week.

With a clamp meter or as Vic said any other semi reliable method of measuring current (if you have a mate3, you oght to be able to get similar graphs to the blackbox ones) then by clamp metering the battery positive cable, you can monitor the charge acceptance amps in real time.

Remember that If you try to EQ before the taper has near completed you run the risk of creating high currents, and temps, as mentioned above.

zoneblue

Anawa wrote: »

Right now I'm trying to get my batteries to 100% SOC (measured by SG) through forced equalization and/or forced bulk. Do you have any recommendations on this process, another EQ or bulk? Yesterday the starting SG was 1.220 (most) and after a 3 hr EQ, they increased to 1.240. The SG variance at the start across all cells was 1.217 to 1.230, with most cells at 1.220. My SG target is 1.285 per manufacturers recommendations.

No your first priority is to get the batteries charging correctly. Trying to EQ a half charged bank will not acheive what you want. "Force bulk" is a hack to start a new absorb cycle. THats not going to work either because as teh bank is over 80% SOC, it will just flick straight to absorb then straight to float as before and you are back to where you started. There are only about 3 things that terminate absorb. Max time, or EA. On the midnite controller there is also skip bulk, but pretty sure that FM doesnt have that. So, for now set the max time to say 6 hours, and the EA to, well, for you 2% of 800Ah, ie 16A. To be on the safe side turn all your (non essential) loads off. If that doesnt make it absorb, then mate, something is busted.

It is important to understand that end amps can work in two ways, via actual battery current, or by controller output current. (only systems with a battery shunt and a controller capable of using it). Clearly the latter will be confused by any loads on the system. On the midnite there is a setting to force it to use the battery shunt. However the confusion will look like delayed transition to float not early transition.

Big industrial banks like that need long absorbs. Yet theres only so many hours in the day, can be a issue. As Vic said higher Vabs helps.

Anawa

Thanks guys, I'll keep working on getting the 100% SOC. I appreciate y'all staying with me as a new week starts and I move through the process. Right now I'm "in-town" and going out to the off-grid site in the AM.

Paul

Anawa

What about using my generator and the FX's to get the 100% SOC? Although my signature does not reflect, I have a 8.5kW RES propane Kohler that I'm sure wouldn't mind a little exercise. I've been so focused on the FM's that I forget it.

Paul

Vic

Hi Paul,

BTW, Radio Shack has had a $50-ish DC Clamp On ammeter, if one near you is still in business... you might call a close on to check stock, if one is close. Their Clamp meter is pretty good, but most of these need to be re-zeroed several times per minute.

AND, why not try this;
Set the Float voltage on each FM CC to your desired Vabs? Would suggest somewhat more than your 57.XX, but even that for hours and hours would make use of free PV power, unless it is cloudy. You would need to monitor progress, as the CC should stay in Float for the entire day, after the usually short or nonexistent Absorb.

Believe that this has not been mentioned, but, also make certain that you use the Remote Temp Sensor/s on each charge source. Your system should be able to share on RTS, using the Hub, or something like that device.

No harm in trying the genset, but if there is good sun, why not try Absorbing, using Float stage with the Vfloat cranked up to your desired Vabs?

You have been bombarded with many thoughts, and am sure that it can be a bit to process. FWIW, Good Luck, Vic

zoneblue

Bombard stopped for now

Anawa

Thank you both for your concerns, but I feel neither bombarded nor overwhelmed. Just more appreciative that there are knowledgeable folks "out there" that want to help others that are "technologicically" challenged in this environment. Before I stated my solar journey, I read enough to know that the knowledge ramp-up would be steep and, most likely, a forever "work-in-progress". So far this seems to be true.

I'll keep you guys posted as I move along this phase of the learning process.

Paul

zoneblue

This is the one i use, its never given any grief.
http://www.dx.com/p/ut203-1-4-lcd-digital-clamp-multimeter-red-black-1-x-9v-6f22-battery-113298