Charge controller and battery meter questions

Plowman · October 2014

I've been having on-going issues with my two Trojan T-105 batteries not getting a full charge. The panels seem to be doing their job, I'm pretty sure the weak link is my Morningstar TS-45 PWM controller.

I've got 632W of panels for 225 Ah @ 12V of batteries, it seems like the batteries should be easily charged on a sunny day. But lately the charge controller hasn't been going into float, it spends most of the day at absorption trickling in the amps at what seems to me to be a very low level (often <2 amps). It's set at 14.8V absorption, as Trojan recommends. Weather has been somewhat variable lately, but even when there's good sun all day I haven't been getting a full charge (usually maxs out around 95% based on SG of 1.265-1.270).

My daily demand averages out to 48 Ah/day @ 12V, including inverter draw. This is only 21% of battery capacity.

Been thinking about replacing the Morningstar TS-45 with a Midnite Kid. Unfortunately their sizing calculator says I'd need 1.5 Kids to match my array (four 158W panels, 18.9V VMP):

Rated PV Array Current: 33.88 Amps
Battery Charging Current @ 14.4 V: 43.9 Amps

Array Power (Wattage) : EXCESSIVE
KIDs Required : 1.5

My Morningstar appears to go through the bulk stage early (I think by 11 am usually), then spends the rest of the day at absorption trickling in a miserly current. I'm thinking that bulk isn't the problem, it's absorption. Am I right? Would it be OK to use a Kid knowing that much of the power coming from the panels at maximum output would be wasted during the bulk stage? I've never seen all four panels at maximum rated output.

I also need a shunt-type battery meter, preferably one that can store a week or more of data. I need more than snapshots based on hydrometer and clamp meter readings.

The options I've been considering within my limited finances (<$400 budget) are:
(1) Stick with the Morningstar TS-45 and get a Trimetric 2030 meter
(2) Sell the Morningstar and get a Midnite Kid with Whiz Bang. But if the Kid is too small.....

Advice?

mike95490 · October 2014

Re: Charge controller and battery meter questions

Bulk and Absorb should be set to the same voltage. Only difference is, while in absorb, the amps should lessen eventually, as the battery fills up, and the timer kicks out and drops you to float.

If you force an EQ (when was your last EQ) do the amps go up ? What's the water usage been like, little usage ? Lots of gassing ?

I question running 632W of PV (45A) into a Morningstar TS-45 PWM ??? Limits ? Running warm ? Any loads running while charging ?

And, then there is the heat, hot days reduces the voltage of the panels. You may be limited there. You could try the "Ice Bucket Challenge" and see what cold water on hot glass does

and if your charge increases.

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

Yes the Kid is too small as that much panel can put out 40 Amps on 12 Volts.

On the TriStar PWM they'd put out about 33 Amps.

The problem is the TriStar is staying Absorb too long. This can be a problem with any controller that has a fixed Absorb setting. I can't remember exactly how they function, but it sounds like the Absorb extend time has been set?

Adding a TriMetric won't do any good for changing that problem.

You could get the Kid, and remove one panel:
3 * 158 = 474 Watts * 0.77 / 12 = 30 Amps peak current. That should in fact be enough for 225 Amp hours of battery.

Plowman · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

The problem is the TriStar is staying Absorb too long. This can be a problem with any controller that has a fixed Absorb setting. I can't remember exactly how they function, but it sounds like the Absorb extend time has been set?

That's the conclusion I've come to. But I can't find a way to adjust the absorption aside from the voltage set point.

Adding a TriMetric won't do any good for changing that problem.

I know a battery meter won't fix the problem, just saying I need a battery meter so I can see exactly what's going on with the incoming and outgoing amps. The Kid + Whiz Bang combo looked perfect, but then I realized the Kid won't handle my array.

You could get the Kid, and remove one panel:
3 * 158 = 474 Watts * 0.77 / 12 = 30 Amps peak current. That should in fact be enough for 225 Amp hours of battery.

I have a cheap 10 amp PWM charge controller that I'm not using, would it work to run one panel on the 10 amp CC and the other three panels on the Kid?

Plowman · October 2014

Re: Charge controller and battery meter questions

mike95490 wrote: »

Bulk and Absorb should be set to the same voltage. Only difference is, while in absorb, the amps should lessen eventually, as the battery fills up, and the timer kicks out and drops you to float.

The TS-45 doesn't have separate set points for bulk and absorption. I have it set to 14.8V, but bulk is lower.

If you force an EQ (when was your last EQ) do the amps go up ? What's the water usage been like, little usage ? Lots of gassing ?

I equalized a couple days ago for a couple hours. Not sure if amps go up. There's some water loss, don't have a good sense of how much.

I question running 632W of PV (45A) into a Morningstar TS-45 PWM ??? Limits ? Running warm ? Any loads running while charging ?

8.47 amps * 4 = 33.88 amps * 1.25 = 42.35 amps. I don't see why a TS-45 couldn't handle this. No it's not running hot, and yes I do sometimes have some loads running during the day. CC output increases proportionally when I turn on a load. As I said, the panels appear to be doing their job, it's the extended time in absorption at a trickle charge that appears to be the problem.

vtmaps · October 2014

Re: Charge controller and battery meter questions

Plowman wrote: »

I've been having on-going issues with my two Trojan T-105 batteries not getting a full charge.
<snip>
but even when there's good sun all day I haven't been getting a full charge (usually maxs out around 95% based on SG of 1.265-1.270).

You have more than enough panel to charge those batteries. I question the premise of your complaint... I suspect that your batteries are getting fully charged.

The low current you see in the afternoon is probably the 'end amps' of the battery... that's the amount of current that flows into a battery when it is fully charged. Since the battery is fully charged, the energy that this current carries cannot go to charging the battery... it goes to heat, electrolysis of water, and corrosion. To avoid excessive heat, electrolysis of water, and corrosion, it is good to switch from absorb to float as soon as the battery is fully charged.

I think Cariboocoot's assessment is correct:

Cariboocoot wrote:

The problem is the TriStar is staying Absorb too long. This can be a problem with any controller that has a fixed Absorb setting. I can't remember exactly how they function, but it sounds like the Absorb extend time has been set?

I'm not familiar with your charge controller, but I think there may be some software that allows you to reduce the absorb time. I don't think you need a more efficient controller... you don't need to wring every last kwh out of your panels... you have plenty of solar panel for those batteries.

If you get a Trimetric monitor, it can be configured to reset to 100% using 'end amps' criteria. For example, you tell the trimetric that your batteries are fully charged when the voltage has been (for a full minute) at 14.8 volts with the battery current less than 3 amps. When 'end amps' is reached, the Trimetric knows your batteries are charged, but it can't tell that to, or control, your charge controller.

More advanced charge controllers (such as the Kid) can make the switch from absorb to float based on 'end amps' criteria. By the way, Bogart (maker of the Trimetric) has a new charge controller that works with their trimetric monitor. Like the midnite Kid, it is a 30 amp controller.

--vtMaps

Plowman · October 2014

Re: Charge controller and battery meter questions

vtmaps wrote: »

You have more than enough panel to charge those batteries. I question the premise of your complaint... I suspect that your batteries are getting fully charged.

I'm basing this on SG readings. Just took a reading, it's currently 4:30pm here, been sunny all day. 1.255 in test cell of battery #1, 1.260 in test cell of battery #2. That's only ~90% SOC, lower than normal.

Lately I've usually been seeing 1.265 by late afternoon, which is ~95%. Loads were a little higher than usual this morning since it's a weekend, but I was gone most of the afternoon and the only load was the inverter.

Charge controller was putting out 0.88 amps @ 14.82V with the inverter turned off, according to my clamp meter. Obviously something not right.

vtmaps · October 2014

Re: Charge controller and battery meter questions

Plowman wrote: »

Lately I've usually been seeing 1.265 by late afternoon, which is ~95%. <snip> Charge controller was putting out 0.88 amps @ 14.82V with the inverter turned off, according to my clamp meter. Obviously something not right.

Something may be wrong, but it's not obvious to me that something is wrong. Maybe 1.265 is as high as SG gets in your batteries, measured with your hydrometer. Or maybe not. If you adjust your Iota voltage and do a real EQ for an hour, you can find out high your SG can be.

One thing you can do that might shed light: discharge the battery to at least 80% SOC. Then recharge it, and while it is in absorb stage, make a graph of time vs charging current. The current should start out high and taper down to 'end amps'.

I don't have your batteries, so I'm not sure how low your end amps should be... 0.88 amps sounds a bit low, but not so low that I would automatically think something is wrong.

On the other hand, you might be right that the SG you are measuring is too low. If you are correct, it means that there is some sulfation. If there is some sulfation, that would raise the internal resistance of the battery and would result in end amps being lower than normal.

Bottom line: equalize, and then know for sure where you stand. If it turns out that your batteries are getting fully charged, then adjust your controller to spend less time in absorb. If they are not getting fully charged, we can troubleshoot... maybe it is the controller... but first you need to know for sure the SOC of your batteries.

SG is, of course, the gold standard for SOC. But you cannot read your hydrometer to a resolution high enough to differentiate between 99% SOC and 100% SOC. Consider this: If you are repeatedly charging up to 99.9% SOC, that 99.9% becomes your new 100%. After awhile, your battery is reduced to 99.9% of 99.9% of 99.9% etc. This is slow deficit charging and the SG slowly declines as more and more electrolyte is lost to sulfation.

If you do have sulfation and it has been longstanding, it is probably not reversible. 95% may be the new 100% for your batteries. Or maybe not. An EQ will bring your batteries to their fullest potential.. what ever that may be. If you reverse some sulfation (assuming there is sulfation and that it is reversible), you can expect to see a bit of an increase in your end amps.

--vtMaps

Plowman · October 2014

Re: Charge controller and battery meter questions

vtmaps wrote: »

Bottom line: equalize, and then know for sure where you stand. If it turns out that your batteries are getting fully charged, then adjust your controller to spend less time in absorb. If they are not getting fully charged, we can troubleshoot... maybe it is the controller... but first you need to know for sure the SOC of your batteries.

SG is, of course, the gold standard for SOC. But you cannot read your hydrometer to a resolution high enough to differentiate between 99% SOC and 100% SOC. Consider this: If you are repeatedly charging up to 99.9% SOC, that 99.9% becomes your new 100%. After awhile, your battery is reduced to 99.9% of 99.9% of 99.9% etc. This is slow deficit charging and the SG slowly declines as more and more electrolyte is lost to sulfation.
--vtMaps

Thanks VT, I did as you said. Equalized with my generator and Iota for 2 hours after work this evening. Voltage at battery was 15.42, amperage was 1.7 right before I shut the generator off. This is measured with an Entech clamp meter.

Here's what I got:

SG, before and after equalization (measured with Hydro-Volt with auto temp compensation):

Cell 1 Battery 1: 1.260 / 1.275
Cell 2 Battery 1: 1.265 / 1.275
Cell 3 Battery 1: 1.275 / didn't measure

Cell 1 Battery 2: 1.260 / 1.275
Cell 2 Battery 2: 1.265 / 1.275
Cell 3 Battery 2: 1.270 / didn't measure

As you can see, I did need to equalize, and the equalization appears to have worked. 1.275 is what I consider "fully charged," Trojan says 1.277 is 100%. It hasn't reached this level since 9/17/14 based on my spotty records.

I usually only see 1.275 after manual equalization with the charge controller. In automatic mode it usually only achieves 1.265, 1.270 if I'm lucky. The latter fact, and my most recent test, suggests to me the charge controller is not fully charging my batteries on a regular basis. Thoughts?

vtmaps · October 2014

Re: Charge controller and battery meter questions

Plowman wrote: »

suggests to me the charge controller is not fully charging my batteries on a regular basis. Thoughts?

That may be a good thing. On not. Very controversial.

RE batteries are a niche market. Most deep cycle batteries are used in situations where they are charged from the grid (forklift, golfcart,etc).

Most battery manufacturers (including Trojan) recommend a charging protocol that has some regulation by current, and usually has daily charging voltages that briefly get as high as your EQ voltage (15.42 volts). These types of charging protocols are NOT the usual bulk, absorb, float that we use in the RE industry (and they may use the terms bulk, absorb, and float differently). The type of charging we do in the RE industry involves regulation by voltage.

One of the nice things about regulation by voltage is that all chargers and controllers can be on the same page... they all see the same battery voltage, so they know what's going on. That's why we can mix and match multiple inverter/chargers and controllers on the same battery bank.

When you do regulation by current, the charger or controller has to know how much current is going into the battery. In most industrial chargers, when it's time to charge the battery, the battery is connected to the charger. There are no loads or other charging sources, so the current produced by the charger is the current going into the battery. One other thing about regulation by current... the power source (grid or generator) must be reliable... the regulation by current protocols don't work too well when clouds pass by.

So now, back to your question... The controversy about battery charging involves how often to get all the way to 100% SOC. And I mean 100.000%, not 99.999%. (you can't read the difference on your hydrometer).

Realize that sustained high voltages are tough on a battery.... with industrial chargers the time spent at 15+ volts on a daily basis is minimal.

If you were to set your RE absorb voltage to 15+ volts, your battery would get to 100% daily, but it would spend a great deal more time at the higher voltages. Too much time for good battery life.

So what to do?

Trojan and many other battery manufacturers recommend "routine" equalization (as opposed to "corrective" equalization). In effect, they are recommending that you deficit charge on a daily basis (easier on the battery), but catch up to 100% every week on a routine basis.

Why every week? Every time you charge or discharge your battery, stratification builds up. That means more of the sulfuric acid is at the bottom of the cells and it takes higher than normal voltage to drive sulfate off the bottom of the plates (more work to push sulfate off plate into higher concentration of sulfuric acid). As you cycle your bank during the week, never quite getting to 100%, this stratification builds up and the sulfate at the bottom of the plates will harden (over time). A "routine" equalization will restore order.

Some manufacturers specify how often to do a routine EQ by the amount of cycling... that is one 50% cycle causes as much stratification as five 10% cycles or two 25% cycles. For most RE situations about 1 week is about right.

One more thing about battery charging... try to look at it from the battery manufacturer's perspective:
Here is something I wrote recently:

vtmaps wrote: »

Suppose a battery manufacturer produces battery cells that have the following characteristics:
99.5% of the cells will be fully charged with a voltage of 2.37 volts (14.2 volts in a 12 volt system).
99.95% of the cells will be fully charged with a voltage of 2.43 volts (14.6 in a 12 volt system).

If they recommend a charging voltage of 14.2, a half percent of their cells will fail during warranty, but that means almost 3% of their 12 volt batteries will fail (6 cells per battery). On the other hand, if they recommend charging at 14.6, only 0.3% of the 12 volt batteries will fail. If you buy that 12 volt battery, what charging voltage will you use? There is a 97% chance that if you use the lower voltage your battery will last longer.

You are lucky that you have flooded cells... you can use a hydrometer to tune your charging voltages and time to what works for your battery in your system in your environment. In all likelihood you can (if you have adjustable chargers and controllers) use less time and voltage than the manufacturer recommends and extend the life of your bank.

By the way, your Iota charger has some non-RE charging protocols... have you noticed that Iota specifies separate bulk and absorb voltages?

--vtMaps

Photowhit · October 2014

Re: Charge controller and battery meter questions

vtmaps wrote: »

Trojan and many other battery manufacturers recommend "routine" equalization (as opposed to "corrective" equalization). In effect, they are recommending that you deficit charge on a daily basis (easier on the battery), but catch up to 100% every week on a routine basis.

Perhaps I'm doing this just to make 'Coot proud, I'm a strong advocate of monthly equalizing, but Trojan and I think Deka? do NOT recommend routine equalizing. They want you to only equalize for correcting out of balance SG.

I thing this is why they also have a fairly high absorb voltage, just opinion on this...

vtmaps · October 2014

Re: Charge controller and battery meter questions

Photowhit wrote: »

but Trojan and I think Deka? do NOT recommend routine equalizing. They want you to only equalize for correcting out of balance SG.

I thing this is why they also have a fairly high absorb voltage, just opinion on this...

OK, thanks... my mistake on Trojan --vtMaps

Plowman · October 2014

Re: Charge controller and battery meter questions

Photowhit wrote: »

Perhaps I'm doing this just to make 'Coot proud, I'm a strong advocate of monthly equalizing, but Trojan and I think Deka? do NOT recommend routine equalizing. They want you to only equalize for correcting out of balance SG.

I thing this is why they also have a fairly high absorb voltage, just opinion on this...

I tried 14.6V absorption for a little while this summer to avoid tripping the overvoltage "feature" on my little Xantrex inverter---had the same problem I'm having now, my batteries weren't getting a full charge and CC wasn't going into float. Switched back to 14.8V and it seemed OK for a while, until the last few weeks.

So how to proceed? Options as I see them:

(1) Get a new charge controller---would like a Midnite Classic with Whiz Bang, but don't think I can justify the expense right now

(2) Keep things as they are, but equalize once a week

(3) Reduce absorption voltage to 14.6V and equalize once a week

As far as I can tell my options are limited when it comes to programming my Morningstar TS-45, but I still have yet to fully explore their software (MSView), which doesn't work all that well on my laptops.

I can probably get it to switch to float earlier, but that won't help. It was in float mode yesterday when the batteries were well below 100%.

mike95490 · October 2014

Re: Charge controller and battery meter questions

I'm assuming you are using the TS-45 as your solar charger ? And you are programming different absorb voltages via RS-232 port ? You should be able to increase the ABSORB TIME, say, in 20 minute steps, and see if longer absorb times pumps enough amps in the battery to charge it. That's the definition of Absorb, after Bulk get the voltage to X, Absorb holds it there for some amount of time, till the battery finishes filling up.

Plowman · October 2014

Re: Charge controller and battery meter questions

mike95490 wrote: »

I'm assuming you are using the TS-45 as your solar charger ? And you are programming different absorb voltages via RS-232 port ? You should be able to increase the ABSORB TIME, say, in 20 minute steps, and see if longer absorb times pumps enough amps in the battery to charge it. That's the definition of Absorb, after Bulk get the voltage to X, Absorb holds it there for some amount of time, till the battery finishes filling up.

I haven't programmed it, like I said the software isn't playing nice with my laptops. Just using the DIP switches to adjust voltage.

Right now sunset is usually the timer on my absorption stage. Trickling in <1 amps for an hour or two longer probably won't help, even if I could.

My Iota was still pushing in >1.5 amps after 2 hours of equalizing at 15.42V, which got the batteries to what I consider 100%. Seems to me that the charge controller is throttling the amps down too much, too soon, resulting in less than a full charge on a day-to-day basis.

My new thinking on the subject, after this thread and thinking more on what folks have said in other threads, is that I don't need to achieve a full charge every day. Once a week is fine as long as SOC stays at a reasonable level (>75%). I can achieve a full charge once a week with manual equalization. Valid strategy?

I don't care if these batteries last 4 years vs. 5 years. I'll have gotten my money's worth either way. I consider them kind of a starter set. But one of these days I'd like to invest in a larger off-grid system and would like to know how to properly manage a larger (more expensive) bank.

vtmaps · October 2014

Re: Charge controller and battery meter questions

Plowman wrote: »

My new thinking on the subject, after this thread and thinking more on what folks have said in other threads, is that I don't need to achieve a full charge every day. Once a week is fine as long as SOC stays at a reasonable level (>75%). I can achieve a full charge once a week with manual equalization. Valid strategy?

Yes, a very valid strategy. Even if you could program your controller to get you to 100% every day, you might be better off with this "once a week" strategy.

--vtMaps

Plowman · October 2014

Re: Charge controller and battery meter questions

vtmaps wrote: »

Yes, a very valid strategy. Even if you could program your controller to get you to 100% every day, you might be better off with this "once a week" strategy.

--vtMaps

Thanks, I think I'll give it a go. If nothing else it'll reduce the number of times I have to break out the hydrometer.

Since the Midnite Kid is too small for my system, and I can't afford a Classic + Whiz Bang, I think I'll break down and get a Trimetric. I need something to see amps coming in and going out. Hopefully it'll also help me fine tune my absorb voltage set point. I'll stick with 14.8V for now.

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

Plowman wrote: »

Since the Midnite Kid is too small for my system, and I can't afford a Classic + Whiz Bang, I think I'll break down and get a Trimetric. I need something to see amps coming in and going out. Hopefully it'll also help me fine tune my absorb voltage set point. I'll stick with 14.8V for now.

I don't see why you think the Kid is too small, as it will handle 30 Amps and that is more than enough for 225 Amps hours which apparently is what you've got.

inetdog · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

I don't see why you think the Kid is too small, as it will handle 30 Amps and that is more than enough for 225 Amps hours which apparently is what you've got.

Not if the PV will be powering daytime loads at the same time that it is charging the battery. (Which is also a good reason for using a shunt to control the actual battery amps during Bulk, and AFAIK that either is or will soon be a possible option with the Kid.)

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

inetdog wrote: »

Not if the PV will be powering daytime loads at the same time that it is charging the battery. (Which is also a good reason for using a shunt to control the actual battery amps during Bulk, and AFAIK that either is or will soon be a possible option with the Kid.)

No.
225 Amp hours only needs 22 Amps peak current. That number includes room for averaged loads while charging, never mind an additional 8 Amps.

You should not try to run super-large PV on battery systems, even with a current-limiting controller.

As it is he has the potential for >20% peak charge rate, and that's quite steep. Charge acceptance would be low and heating would be high and all it takes is for those loads to not be constant (which they rarely are).

But then again everyone should just ignore anything I say because I've only fixed more systems than most people have ever seen so what would I know about anything, right?

inetdog · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

No.
225 Amp hours only needs 22 Amps peak current. That number includes room for averaged loads while charging, never mind an additional 8 Amps.

You should not try to run super-large PV on battery systems, even with a current-limiting controller.

As it is he has the potential for >20% peak charge rate, and that's quite steep. Charge acceptance would be low and heating would be high and all it takes is for those loads to not be constant (which they rarely are).

Maybe I am missing something here? When you say 20% charge rate, you mean C/5 right? Definitely too high for FLA, although reasonable for AGM.
But how is 30A greater than C/5 for a 225AH battery? My version of arithmetic tells me that 225AH/5H is 45A, not 30A.
Does the Kid have the capability of delivering a short term output greater than the 30A continuous rating?

And finally, you mention that there is room for an additional 8A of load while delivering a perfectly satisfactory C/10 rate to the battery.
With a 12V battery system, that is less than 100W of AC load. Not what I would call a major daytime load.

I am not doubting your advice, which I do respect, just your math behind it.

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

The potential charge current he has available now from all his panels vs. dropping one panel and using a Kid.

Four 158 Watt panels = 632 Watts. Right now that's on a PWM TriStar and can put out over 35 Amps (if I recall the Vmp/Imp on those panels correctly). On an MPPT it would be 40 Amps +/-. That's pretty near 20% of the 225 Amp hours. 17.7% if you want to be irritatingly picky. At that rate charge acceptance is noticeably lower (a couple of us tested this problem this Summer).

Hence my original suggestion of dropping one panel and using a Kid. 474 Watt array and about 30 Amps output.

If expansion is an issue you can double up Kid controllers giving you 60 Amps capacity. The combined cost of two of them is less than one Classic 150, with a small sacrifice in current capacity (36 Amps on a 12 Volt system). If you are looking at going beyond that point for stored capacity it's worthwhile considering upping system Voltage (if viable) or replacing/supplementing the Kid with a Classic.

Photowhit · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

You should not try to run super-large PV on battery systems, even with a current-limiting controller.

Well curious about this, as I'm about to add to my already over paneled array.

I hope to add @2200 watts to my 4000 watt array. I intend to run an electric water heater via relay, once batteries are charged. Had planned on making it a stand alone system, but decided with the Classic's features, I would be better off doing it this way. Currently I do not regulate the battery charging as I can't normally expect more than 15% charge rate.

I know this is out of norm, the panels cost me $1000 for 2600 watts, so pretty darn cheap. (New panels warehouse accident)

The idea is to have some water heating on marginal days.

Cariboocoot wrote: »

But then again everyone should just ignore anything I say because I've only fixed more systems than most people have ever seen so what would I know about anything, right?

Always respect your opinion, and almost always agree...

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

Photowhit wrote: »

Well curious about this, as I'm about to add to my already over paneled array.

I hope to add @2200 watts to my 4000 watt array. I intend to run an electric water heater via relay, once batteries are charged. Had planned on making it a stand alone system, but decided with the Classic's features, I would be better off doing it this way. Currently I do not regulate the battery charging as I can't normally expect more than 15% charge rate.

I know this is out of norm, the panels cost me $1000 for 2600 watts, so pretty darn cheap. (New panels warehouse accident)

Always respect your opinion, and almost always agree...

This is exactly the kind of situation that can get you in trouble.
Batteries fully charged before opportunity load is active? No problem.
Adding a large amount of PV to supply a big opportunity load? Possible problem.

You don't want to limit the controller output or else you will also limit the power available to the opportunity load (and thus wouldn't gain anything). The Outback FNDC has the potential to work around this (never programmed one so I don't know if it actually can) because it has three shunts and theoretically at least can tell the difference between current from the controller and current to the battery, thus it should be possible to limit the latter without affecting the former (power going to loads only).

If you just pile on the PV then expect the full current potential to be available to the batteries at some point. Thermostatically controlled loads will turn on/off as they see fit, and there may be some lag in the reaction of the controller to the reduced load.

That is providing you don't get into a situation where (too) large amounts of current go to the battery during Bulk and/or early Absorb even without the load cycling. Too short a Bulk stage is not good just as too short an Absorb stage isn't.

There are a lot of variables to look at, including what the maximum current will be as a charge rate (too high for FLA's? Consider using AGM's) and how the charge controller will handle the current (limiting how?) and fluctuating load demand (will it react fast enough?). I've not tried this with a Classic, but suspect it would be the best save for the current 'dividing' between charge and loads. I suspect Bob & Robin looked at the shortcomings of their old MX and made good on them. The MX, btw, has quite noticeable lag in reaction times as well as needing to sweep periodically (programmable) to obtain the best power point for conditions. They are also not good at limiting current (will rise above programmed max).

Now in some cases this momentary whack with too high current will be brief and rare enough that the only result will be a small amount of time shaved off the batteries' lifespan. In the worst case it will be a lot of time shaved off. All a matter of how much too high the current is, how long it is applied for, and how often it happens.

jcheil · October 2014

Re: Charge controller and battery meter questions

So what is the proper "ending amps" to consider ending absorb?

Photowhit · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

You don't want to limit the controller output or else you will also limit the power available to the opportunity load (and thus wouldn't gain anything).

I would impose a 120 amp input to my 800Ah battery, 15% is limit manufacturer suggests for max battery life. It's not a huge opportunity load, 900 watts (may try to go to a 4400watt 240 element run on 120 for about 1100 watt load) I do hope to gain warm showers on over cast days...

I believe a single shunt at the battery connection is all that is needed to allow for extra generation from the array to cover loads, since the CC should see the amperage there and be able to make the adjustments to maintain the charge current. Do to the electro chemical nature of charging, I don't think brief over current of a couple seconds will create problems, though I've been wrong before, I think from my reading at Midnite this is their intent.

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

jcheil wrote: »

So what is the proper "ending amps" to consider ending absorb?

Normally around 2% of capacity at the 20 hour rate.
But depending on the system that may need to go up/down, including the need for load allowance.

Someday MidNite may announce the WhizBang Sr. which will operate with multiple shunts (like the FNDC) using a small external processing board to report not only SOC to the Classic but also to adjust its current limit based on current to the batteries (including from multiple sources) with perhaps a Bulk stage algorithm to improve charge acceptance (starting at a lower current percentage at low SOC and moving up).

But they haven't yet.

vtmaps · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

The Outback FNDC has the potential to work around this (never programmed one so I don't know if it actually can) because it has three shunts and theoretically at least can tell the difference between current from the controller and current to the battery, thus it should be possible to limit the latter without affecting the former (power going to loads only).

Yes, Outback can do that, but also needs a Mate3.

This feature has been discussed a few times on the Midnite forum, and will be implemented in future firmware. The Midnite controllers will only need one shunt (with whizBang). They will regulate their output to cover loads above and beyond the maximum charging limit. Apparently the Classic knows how much power it is producing without an external shunt. The external shunt and whizbang allow the Classic to know the battery's current as well as its own current.

Also discussed has been the possibility of using the whizbang to allow some constant current protocols (for example, equalization with current limiting).

--vtMaps

jcheil · October 2014

Re: Charge controller and battery meter questions

Cariboocoot wrote: »

Normally around 2% of capacity at the 20 hour rate.
But depending on the system that may need to go up/down, including the need for load allowance.

Just so I understand, on a 740ah bank, that would be 14amps?
Meaning when the charge controllers get to the point of only putting 14amps into the battery, it is considered time to leave absorb.
But I question how it knows the difference between 14amps that the batteries are able to take vs the array for example on a cloudy day suddenly putting out only 10. I would think that would confuse the "end amps" calculation.

I have an FNDC and the SOC capability is awesome (so far) but I just wanted to understand how the end amps work and wondering if I should be using it also.

Cariboocoot · October 2014

Re: Charge controller and battery meter questions

jcheil wrote: »

Just so I understand, on a 740ah bank, that would be 14amps?
Meaning when the charge controllers get to the point of only putting 14amps into the battery, it is considered time to leave absorb.

That is correct. It gets to the set End Amps and changes to Float, regardless of how much time is left on the Absorb timer.

But I question how it knows the difference between 14amps that the batteries are able to take vs the array for example on a cloudy day suddenly putting out only 10. I would think that would confuse the "end amps" calculation.

This is the problem with load allowance. The controller can not know how much is going to battery and how much is going to load unless there is a shunt to measure each. From what vtMaps says the Classic is reading its internal current and comparing it to the external measurement on the WhizBang Jr. to determine this. That makes sense.

Cloudy days are unlikely to have an effect of End Amps. Voltage level must be maintained, and if there is a sudden drop in current this is not likely to happen; V drops below Absorb level and the controller knows it hasn't finished Absorb. The MX will slip to Bulk and run more time up on the clock until Absorb level is reached again. Also, the current must remain below End Amps for a given amount of time before the stage changes. This is to avoid transient effects near the end of the cycle.

I have an FNDC and the SOC capability is awesome (so far) but I just wanted to understand how the end amps work and wondering if I should be using it also.

Technically End Amps is redundant when SOC Absorb End is used. Hopefully the SOC change will be more accurate.

Some of us still struggle with eight-year-old controllers that cost $900 new and don't do half the stuff a Classic can. But it still works. Flawlessly I might add.

Blackcherry04 · October 2014

Re: Charge controller and battery meter questions

FNdc Charge Parameters Met Time

During this period, the charge parameters must meet or exceed the user-set charge parameters before the battery is considered full.

Programmable from 0 to 240 minutes

Setting this time to 0 effectively eliminates any filtering provided by this timer. The first time the charge parameters are met, the unit will consider the parameters met and perform accordingly.

The default setting is one minute. Longer or shorter periods may apply to certain sources such as wind generators or hydros.

I think I have mine set to 10 Minutes.

Charge controller and battery meter questions

Comments

Categories