Hello all,
About a year ago I purchased my first Pv setup. It consisted of a Renogy 100W Monocrystalline panel and their "30A" PWM charge controller (which I found out later only handles ~300W @ 12V). I also bought one Power Patrol (Interstate Batteries) SLA1165 12V 55AH AGM battery. I had a few small inverters already so I used one of them to run a laptop or two off the system. I mounted the panel on my back patio roof and bought some 8AWG Pv cable (50ft) to connect it up to the charge controller inside. I was trying to spread out the cost of the system I wanted to build by buying this setup first. I eventually bought more of the SLA1165 batteries for a total of 4. I found out after noticing some problems and doing some research that this was not a good idea. At first the system worked fine, but the 4 batteries in parallel eventually got unbalanced.
Today I have a MidNite Classic 150 with 3 of the Renogy 100W panels in series. I've purchased a Samlex 2KW pure sine wave inverter (SSW200012A) for large temporary loads, and a Go Power 175W modified sine wave inverter (GP-175) for small continuous loads. The 2KW inverter doesn't get a lot of use, just short duration loads just to get some time out of it. It is intended for power outages only. The 175W is always on and usually has laptops or other small electronics connected to it. I actually just bought a 2nd one since they are very efficient and have a surprisingly low idle power utilization. I've made several other improvements and started putting in more safety equipment as I've learned what is needed.
Before I get to my questions, this is what the system looks like now, starting from the roof:
Renogy 100W panel, 22.5Voc, 5.75Isc - 3 in series
50ft, 8AWG Cu, Pv cable
1000V 20A in-line fuses (one on each Pv conductor)
MidNite Classic 150
2ft, 6AWG Cu, welding cable with PP75 connector
2ft, 6AWG Cu, welding cable with PP75 connector and 32V 80A in-line AGU fuses (one on each conductor)
Wire combining blocks (7 screw-down connections)
1ft, 4AWG Cu, welding cable to each battery with crimped-on rings
-------------------(Also connected to wire combining blocks)--------------
2ft, 1/0AWG Cu, welding cable with SB350 connector and 32V 250A in-line ANL fuses (one on each conductor)
1ft, 1/0AWG Cu, welding cable with SB350 connector, connected with rings to 2KW Inverter
-----
3ft, 6AWG Cu, welding cable with PP75 connector and 32V 100A in-line AGU fuses (one on each conductor)
PP75(1) to PP45(4) adapter, PP45 and PP30 can connect.
(several devices connected using PP30 connectors, including 175W inverter, with their own fuses)
-----
I've already learned that my 4 parallel-connected 12V 55AH batteries are going to have problems based on balance. When they finally die, I plan to replace them with 6V or 4V batteries in series, one string. However, I'd like to get at least a few more years out of these batteries. I've read using the equalize function on the charge controller is a bad idea with AGM batteries. How would I equalize the batteries? I know they are off because they used to charge and discharge showing the same amperage flow to/from each battery using a clamp meter. Now they are off by about 15% after a year. This is my biggest issue right now and the main reason I wanted to post on this forum. The batteries used to settle out at night around 12.9V after a full charge with light load, but now they settle around 12.7 and continue down to 12.4 with only light load overnight. This change in voltage is when I first realized something was wrong.
For my 2nd question, which may be irrelevant depending on the answer to the first one above, what would be the recommended settings on the Classic for my current battery bank? The spec sheet only has the following (per battery):
Charge Voltage (constant)
(float) 13.5V ~ 13.8V
(cycle) 14.4V ~ 14.7V
Max Charge Current 16.5A
So I setup the classic with Bulk and Float at 13.8V max (I think), and Absorb at 14.7V (1hr max). Equalize is disabled. Re-bulk is set to 12.5V.
Any suggestions?
For my 3rd question, any comments on the wiring? My maximum expected continuous amps is 200A from the batteries and the wiring to the 2KW inverter matches the manufacturer's suggestion. Obviously if I change the battery bank I will change out to 2/0 or better wiring for the battery connections.
Edit: 4th question. I eventually want to have 1200W of Pv on the roof, but I want to keep the 12V battery system. It will be all the same panels that I have now. I planned two strings of 6 panels, but I've read on here that can cause stress on the Classic. Suggestions?
This is all off-grid, obviously, and I have no intention to ever connect it to the grid. The system final design is to take the place of a 2KW portable emergency generator. I intend it to operate my refrigerator and small electronics during an extended power outage. I realize right now I will likely have an 80% DOD and kill the batteries if I put the refrigerator on it, but that is an acceptable risk right now. Let's hope the power doesn't go out for very long anytime soon! Day-to-day operation will be just small electronics to keep the DOD at 30% or better.
As a side note, I am in West Central Florida, so plenty of sun. However, my panels are facing North so I don't get the most optimum power generation that I could. I also have some shading issues at the beginning and end of the day due to trees.
Thought I would attach some of the graphs for my setup from MyMidNite. The lowest voltage on the graph is 12.4V.
I am sure you know what that graph is telling you... you are deficit charging, as the Voltage line is ever so slightly decreasing as time advances, if the hard numbers don't tell the same story, it will just be a matter of time.
300W (17.5A) of poorly located is marginal to recharge 220Ah @ 12 v , normally you would need 22A from ~ 4 well placed panels to give you a 10% charge rate... see what I am getting at?
you would be much better off with 4 6V GC2 batteries
hth
Thanks for the reply Westbranch!
I think the graph may look like deficit charging since I have been progressively putting more load on the system over the last few weeks since it seemed like I was hardly utilizing the power generated by the panels. You may notice the first day on the graph I put a little too much load on it during the day and it never reached float charge. On the 7th day I put load on the system late in the day after it was already float charging, so the float charge amps increased as seen by the increased wattage from the Pv array.
You are right, I don't have enough solar panels yet. I plan to add 3 more this year as money allows. I'm just not sure I am deficit charging since the load on the system is very small. My worse issue currently (I think) is the batteries are unbalanced causing the voltage to drop off quickly after the sun goes down, even with only 200mA of load. I know a newly designed battery system is in my future, but I can't do that this year.
What do you think?
You have a Classic, did you get the WhizBang jr? If not it is a recommended powerful little piece of electronics as it allows you to log the actual power going into the battery, measured at the battery (well as close as you can get almost).
It will also let you know if you have to raise/or lower the Absorb voltage based on those crude specs you posted... as well as the amount of Absorb time you set in the classic..
hth
I keep forgetting to mail in the postcard for that! I'm also not sure how best to hook it up given the way my batteries are connected. I also don't have a shunt, does it come with one?
I tried to wire everything to minimize loss due to cable/connections. It will be interesting to figure out how to rewire to include the shunt.
Quote from: kf4hzu on January 28, 2015, 01:02:01 AM
I keep forgetting to mail in the postcard for that! I'm also not sure how best to hook it up given the way my batteries are connected. I also don't have a shunt, does it come with one?
I tried to wire everything to minimize loss due to cable/connections. It will be interesting to figure out how to rewire to include the shunt.
You will have to purchase a shunt, the WBjr does not come with a shunt. It is pretty simple to connect it up. Just remember this rule. One end of the shunt goes to the negative of your battery bank and that is the only connection on that side of the shunt. All other negative loads including the Classic Battery negative, inventor negative connect to the other side of the shunt.
here is a good pic of the WBJR mounted,
http://www.midnitesolar.com/productPhoto.php?product_ID=519&productCatName=Charge%20Controllers%20-%20Classics&productCat_ID=21&sortOrder=9&act=pc
not knowing how your wiring goes, here is a pic of my Epanel with shunt in middle, taken in progress...
http://www.altestore.com/store/Meters-Communications-Site-Analysis/Shunts/500A-50mV-DC-Shunt-for-Current-Monitoring-Meters/p4638/
Will this shunt work fine, or is there a better resource for one? I usually shop on Amazon but can't find the MidNite Solar shunt on there. I'd prefer to use known good components.
Also I have attached a picture showing how my batteries are connected to everything else (also reference my first post on this thread). This is just the negative connection in the picture, but the positive one is identical.
The orange circle indicates where the 4 batteries are connected. The blue circle is the charge controller negative, green is 2KW inverter, red is 75A "general" feed where the 175W inverter is, among other things.
So inserting a shunt is going to require some new hardware and rewiring. Suggestions welcome!
Also I had a few other questions in the original post, including how/when/if to equalize the batteries. Any ideas on those?
Inserting a shunt into your existing wiring would not be trivial.
You could tidy up all this wiring and improve control and safety with a Midnite Epanel or an MNDC box. These can simplify your wiring and they come with a detailed wiring diagram, spaces for circuit breakers and mountings for the shunt. Epanels include a shunt but you need to get one for the MNDC
Dgd
dgd,
I think the MNDC is the way to go for me, I don't have a lot of space. It may be a little while before I can drop the money on it, along with the money for all of the new wiring, breakers, etc I would need to connect everything. If I am going to up my game to a box like that, I would want to move to 2/0 or better cable for the batteries and main inverter connections.
So I read in the manual that it is not acceptable to use welding cable for residential wiring. Does that mean my current system violates code? That may expedite changing out the cabling and moving to a wiring box like the MNDC.
Hi Jesse(?),
Know that you want to run a 12 V system; a 2 KW Inverter is fairly large on a 12 V system, this could hammer your smallish AGM batteries, especially if each battery is not equally sharing the current demands.
In my opinion, strings of five of those PVs would better than six, especially on a 12 volt system. Believe that strings of six, with 1200 watts STC may well push things too much, but have not run that through the MidNite Classic String Sizer. Even five 12 V nominal PVs might push the Classic a bit on a 12 V system.
The Linked Shunt looks fine to me. Deltec is the standard Shunt for our types of systems, and that Shunt looks just like the Deltec. This IS the most common shunt used around here.
You should only EQ your AGM batteries if it is recommended by the manufacturer. Often EQ on AGMs is really a L-O-N-G Absorb, and the customary Vabs.
Regarding using Welding Cable in your system. It would probably be OK, but might not pass an Inspection ... but your system may not even need to be inspected. Often welding cable does not meet some specs, especially for use in residential applications. AND, you would generally need to use lugs that are rated for use on finely stranded cable, and you might want to use breakers or fuses that have terminal connections that either have studs for lugs, or are rated for clamping fine stranded cable, etc.
FWIW, Vic
Quote from: Vic on January 29, 2015, 11:43:35 PM
Hi Jesse(?),
Know that you want to run a 12 V system; a 2 KW Inverter is fairly large on a 12 V system, this could hammer your smallish AGM batteries, especially if each battery is not equally sharing the current demands.
Yep I'm Jesse!
I want to stick with a 12V system because I have so many 12V devices, cables, fuses, etc with Anderson PowerPole connectors, that I hardly need to use inverters with the system for most of the things I would need in an emergency situation. I've been in to ham radio for many years, and I've built out much smaller portable emergency power configurations so I could work in a disaster area if needed (Helped out after Charlie hit). I used "Field Day" events to test everything, using small generators to charge small AGM batteries. I also want the power system (minus batteries and CC) to be portable so I can move it to my car (running the alternator) if I'm somehow in dire need of power and the solar power is non-functional.
The 2KW inverter is over-built for my needs. I only planned 1.5KW max (microwave), and only about 600W semi-continuous (fridge/freezer), so I bought an inverter that is slightly over what I need so I don't run it at 100% duty cycle for any length of time. (Yes, that is 2.1KW, I would unplug the fridge before running microwave) I think it may hurt efficiency, but I don't run the 2KW inverter 24/7 because it is very noisy, especially under load. It only runs when I would have normally turned on a generator in an extended power outage (or brief periods just to test it and get some time out of it).
Quote from: Vic on January 29, 2015, 11:43:35 PM
In my opinion, strings of five of those PVs would better than six, especially on a 12 volt system. Believe that strings of six, with 1200 watts STC may well push things too much, but have not run that through the MidNite Classic String Sizer. Even five 12 V nominal PVs might push the Classic a bit on a 12 V system.
I did use the string sizer setting up 12 of my panels in 2 strings and it said it was ok on the Classic 150 down to (IIRC) -8F which is unlikely in this part of Florida. But "ok" and "optimum" I know are different things. 5+5 may be fine for me, I'll need to re-do some math to figure it out.
Quote from: Vic on January 29, 2015, 11:43:35 PM
The Linked Shunt looks fine to me. Deltec is the standard Shunt for our types of systems, and that Shunt looks just like the Deltec. This IS the most common shunt used around here.
Cool, I'll order it after I get the WBJr, thanks!
Quote from: Vic on January 29, 2015, 11:43:35 PM
You should only EQ your AGM batteries if it is recommended by the manufacturer. Often EQ on AGMs is really a L-O-N-G Absorb, and the customary Vabs.
Long absorb sounds safe, the guys at Interstate Battery didn't really have an answer for me. I increased my normal Absorb to 2 hours to help. How about I setup the Equalize for 14.7V for ~5 hours every 30 days (if I can even do that with the Classic)? Also, they are unbalanced right now, badly. I have a "smart" charger that runs off the wall. Any suggestions on bringing them back in to balance soon-ish, or just let the CC do it over time?
Quote from: Vic on January 29, 2015, 11:43:35 PM
Regarding using Welding Cable in your system. It would probably be OK, but might not pass an Inspection ... but your system may not even need to be inspected. Often welding cable does not meet some specs, especially for use in residential applications. AND, you would generally need to use lugs that are rated for use on finely stranded cable, and you might want to use breakers or fuses that have terminal connections that either have studs for lugs, or are rated for clamping fine stranded cable, etc.
FWIW, Vic
I purchased all the major artery cables (6GA to 1/0GA currently), excluding the orange 4GA cables, from powerwerx.com because I didn't have the right tools to properly crimp connectors on cable that large. I believe they would have given me the correct connectors for the assembled cables. I didn't want to skimp and risk issues when messing around with this much DC potential. I think before I rebuild the wiring as discussed earlier in this thread I'll purchase the proper tools to make my own connections. I didn't even know there were different ratings for clamping different strand types. I've only been aware of differences between solid and stranded connectors.
I am less concerned about random inspection, and more concerned about doing it wrong, causing a fire, and THEN an inspection finds out I did something wrong (like selecting the wrong cables/connectors) which caused the fire and therefore I don't get insurance coverage for my house. Worse would be if someone gets hurt by a bad decision on design or components. This is in a closet that I close when I have visitors, but there isn't a lock on the door. I've had thoughts about building a new small closet dedicated to the Pv system, but that is not going to happen anytime soon.
I appreciate your input and advice. Definitely learning a lot of things in the few days I have been on this forum. I've also been browsing around old posts, but I have some unique issues that I haven't seen on here.
Looking forward to more discussion!
Please, do not try to use any Equalize settings on your AGM batteries. Go to the Mfg's website and look up the maintenance charge voltages.
When charging AGM, once you exceed the internal gas recombiner, it Vents gas, and therefor you loose electrolyte that can not be put back.
Quote from: kf4hzu on January 27, 2015, 11:13:50 PM
How would I equalize the batteries? I know they are off because they used to charge and discharge showing the same amperage flow to/from each battery using a clamp meter. Now they are off by about 15% after a year. This is my biggest issue right now and the main reason I wanted to post on this forum. The batteries used to settle out at night around 12.9V after a full charge with light load, but now they settle around 12.7 and continue down to 12.4 with only light load overnight. This change in voltage is when I first realized something was wrong.
For most folks, by the time they realize something is wrong with their parallel batteries, serious damage has been done. You seem to be more on top of it than most folks.... you have a clamp meter and use it, and you are sensitive to tenth-of-a-volt differences. You might be able to rescue your bank.
What you MUST do is separate those batteries and charge them and evaluate them one at a time.
Some AGM manufacturers do recommend an "equalization" when needed... but it's usually just an extra long absorb, or its very few tenths of a volt higher than a routine absorb. I think you mentioned that you have interstate batteries... they're OK batteries (in my opinion) but you must understand that Interstate's tech support folks are clueless, and have no clue that they're clueless. If you can find out who made those particular batteries for interstate, you can look up that manufacturer's recommendations for the battery.
For safety reasons it is a good idea to put a fuse or circuit breaker in each string of a parallel battery bank. If you had done that it would be very easy to switch in or out any particular battery in your bank. Of course, all the extra connections involved in fusing each string add connections (which are points of failure) to the battery wiring. You seem to understand some of the issues with parallel batteries. Do you know that the problems are exacerbated with AGM batteries? This is because their lower internal resistance makes the resistance in the wiring relatively greater.
--vtMaps
Quote from: kf4hzu on January 30, 2015, 12:47:36 AM
I didn't even know there were different ratings for clamping different strand types. I've only been aware of differences between solid and stranded connectors.
A recent thread over at NAWS may be of interest to you:
http://forum.solar-electric.com/showthread.php?26114
Also, this Home Power article:
http://www.homepower.com/articles/solar-electricity/design-installation/inverter-battery-cables
--vtMaps
Quote from: kf4hzu on January 30, 2015, 12:47:36 AM
I did use the string sizer setting up 12 of my panels in 2 strings and it said it was ok on the Classic 150 down to (IIRC) -8F which is unlikely in this part of Florida. But "ok" and "optimum" I know are different things. 5+5 may be fine for me, I'll need to re-do some math to figure it out.
Actually, optimum (for a 12 volt system) would probably be three of those panels in series. How long a run is it from the panels to the controller?
--vtMaps
A lot of information since my last visit, thanks! This may be a bit long of a post...
Quote from: mike90045 on January 30, 2015, 01:55:39 AM
Please, do not try to use any Equalize settings on your AGM batteries. Go to the Mfg's website and look up the maintenance charge voltages.
When charging AGM, once you exceed the internal gas recombiner, it Vents gas, and therefor you loose electrolyte that can not be put back.
Yep that seems to be what everyone says. I did read the spec sheet; that information is in my original post.
Quote from: vtmaps on January 30, 2015, 05:52:38 AM
For most folks, by the time they realize something is wrong with their parallel batteries, serious damage has been done. You seem to be more on top of it than most folks.... you have a clamp meter and use it, and you are sensitive to tenth-of-a-volt differences. You might be able to rescue your bank.
Thanks, I hope so. After all the reading I've done over the last few weeks I plan to use the clamp meter (which I've found to be very accurate) to check the voltage and charge/discharge amps from each battery every week going forward.
Quote from: vtmaps on January 30, 2015, 05:52:38 AM
What you MUST do is separate those batteries and charge them and evaluate them one at a time.
Some AGM manufacturers do recommend an "equalization" when needed... but it's usually just an extra long absorb, or its very few tenths of a volt higher than a routine absorb. I think you mentioned that you have interstate batteries... they're OK batteries (in my opinion) but you must understand that Interstate's tech support folks are clueless, and have no clue that they're clueless. If you can find out who made those particular batteries for interstate, you can look up that manufacturer's recommendations for the battery.
For safety reasons it is a good idea to put a fuse or circuit breaker in each string of a parallel battery bank. If you had done that it would be very easy to switch in or out any particular battery in your bank. Of course, all the extra connections involved in fusing each string add connections (which are points of failure) to the battery wiring. You seem to understand some of the issues with parallel batteries. Do you know that the problems are exacerbated with AGM batteries? This is because their lower internal resistance makes the resistance in the wiring relatively greater.
Since I kept it a 12V system, I can actually remove the entire bank from the system and replace it with a 12V 35AH AGM battery I have allowing me to do any charge/eval cycle needed on them. What do you suggest I do, and what results should I look for?
I've looked for who may make the PowerPatrol SLA1165 batteries but only find spec sheets that don't even have the word "Equalize" in them. I do believe the tech support people don't know what they are talking about, which is why I asked here :)
I actually bought fuses and fuse holders to go between the batteries as you suggest. I decided not to put them in because I was concerned about adding resistance and increasing my voltage sag during heavy load. Do you feel it is best to put them in anyway? If so, is it fine to only fuse the positive leads?
I recently learned my AGM batteries are the worst for parallel banks. Wish I would have known before I spent the money on them. I just need to get some more time out of them before I can buy another bank that is designed properly. Maybe LiFePO2? :)
Quote from: vtmaps on January 30, 2015, 06:23:06 AM
A recent thread over at NAWS may be of interest to you:
http://forum.solar-electric.com/showthread.php?26114
Also, this Home Power article:
http://www.homepower.com/articles/solar-electricity/design-installation/inverter-battery-cables
Ugh, none of that is good news for me. The only thing I have that meets compliance are the rings that powerwerx.com used, which ARE UL listed. The cable doesn't have the required markings on it to pass. The future cost of this system is going to exceed the budget I've allocated to it. :(
Quote from: vtmaps on January 30, 2015, 06:26:48 AM
Actually, optimum (for a 12 volt system) would probably be three of those panels in series. How long a run is it from the panels to the controller?
It is a 50' run from the roof, not counting the short interconnect cables on the panels. (I put that all in my first post)
Using Vmp (18.9V) and Imp (5.29A) for the math, building it out the way you suggest, http://www.calculator.net/voltage-drop-calculator.html says I would have about a 2.35% loss. My original design of 6x2 would be 0.58% loss. At 1200W, that would be an increase of 21W of loss if my math is right. I've read a lot of articles suggesting to keep cable loss at or below 1% if possible. Would having a 3x4 configuration with that cable loss be better due to the Classic's losses in converting down from ~113.4V on a 6x2 configuration? I know I'd need to change out the fuses for the Pv side since I only have 20A fuses now.
I purchased the Classic because I thought it could easily and efficiently handle the 6x2 configuration I had planned when I reach 1200W of Pv. The string calculator on the site said it was ok. There was no mention of estimated losses in the controller itself due to the configuration. Maybe this should be added to the calculator if it really does make a significant difference in losses which also increases heating of the controller, reducing its lifespan. Perhaps someone from MidNite can chime in on this one?
On a good note, another 100W panel is on the way! My wife decided it would make a good birthday present. So at least I will be at the 10% charge amps previously mentioned in this thread on a good day.
Thank you all for the info so far! Still a lot to learn...
Thought I'd attach another graph. I put too much load on it so never even reached Absorb. Shed some load last night and full 2hr absorb with a little over an hour of float today! With the 4th panel coming I should be good at the current load levels. Just need to repair the battery bank issues.
I have noticed something that may be an issue. The relay in the charge controller clicks every few minutes in the morning and evening when the light levels are low. In the morning it switches between BULK MPPT and RESTING (Low Light). This evening it is switching between FLOAT MPPT and RESTING (Low Light). The output watts is zero in both modes when this is going on. Is this normal? I'm guessing the relay has a lifetime max cycles. Any reason SSRs weren't used?
You can set the time between sweeps on a Classic.
The sound is normal so don't have it beside your bed unless you want to get up with the sun... ::)
Well I know the sound is normal, just didn't think it would be that frequent. It is in a closet far away from our bedroom so no problem there :)
My only concern is the lifespan of the relay. Would it be a good idea to increase the sweep to 5 minutes, or 10? I figured lower was better so as conditions change it can adapt quickly. But I'd also like my CC to last many years, at least longer then the 25yr warranty of the panels.
FYI, found this quote from boB from about a year ago about sweep... emphasis added
QuoteAs for the sweep interval time on the Solar mode, 3 minutes is the default time
but that is a "must sweep" time interval. There are times, in Solar mode, that the
Classic may sweep before that time interval.
I know my MX60m designed by boB, lasted for >5 years clicking away and it is now over at the neighbours... clicking on..... so I think you don't have to worry about longevity
Just a quick update; I now have 4 panels for a total of 400W hooked up in series.
This weekend I plan to work on the battery bank issues since it will be very sunny according to the forecast.
I didn't read how you have your four 12V batteries connected. This is a big deal for AGM's
See the attached drawing, ignore the NEG going to Ground, this is how RV batteries SHOULD be wired and most are not and RV owners replace batteries every year. Getting the power IN/OUT off one end of a paralleled set of batteries will kill the far ones.
Quote from: WizBandit on February 07, 2015, 04:39:04 PM
I didn't read how you have your four 12V batteries connected. This is a big deal for AGM's
See the attached drawing, ignore the NEG going to Ground, this is how RV batteries SHOULD be wired and most are not and RV owners replace batteries every year. Getting the power IN/OUT off one end of a paralleled set of batteries will kill the far ones.
You can do much better than your diagram shows:
http://www.smartgauge.co.uk/batt_con.html
However, the most important thing to know about having four batteries in parallel is that it is a poor design choice to have so many batteries in parallel, even if they are wired to perfection.
The batteries are, at best, in an unstable equilibrium. Usually they are not even at equilibrium.
Getting the current to divide equally among four batteries is like trying to balance a pencil on its point. They are both examples of unstable equilibrium.
Suppose, that for any reason, one battery gets a bit more current flow than the others. That makes it warmer. If it is warmer it gets more current flow than the others. If it gets more current flow, it gets warmer than the others... get the point? It's like balancing a pencil on its point... if it leans just a bit one way, the forces on it will pull it further in that direction which will increase the forces pulling it in that direction... get the point?
Each of four batteries in parallel should be fused. The problems I just mentioned become dangerous when a cell in one battery shorts out. Of course, adding all those fuses means adding many more connections which are all potential points of failure. Because of the issues with monitoring and maintaining four parallel batteries, it is a good idea to include some switches (or circuit breakers) to be able to switch in or out any particular battery.
The life of a battery bank is equal to the life of the shortest-lived cell. The lifespan of the cells is a Gaussian curve (bell shaped curve). With four 12 volt batteries, there are 24 cells. Hopefully, none of your 24 cells is in the lower tail of that curve, because it doesn't matter what the average lifespan of your cells is... all that matters is that one of your cells will pull down the whole bank when its time is up. If you had just 6 cells to make up your 12 volt battery, you would have a much greater chance of NOT having an outlier short-lived cell. In fact, you would even have a 1 in 64 chance of buying 6 above average cells.
--vtMaps
Quote from: WizBandit on February 07, 2015, 04:39:04 PM
I didn't read how you have your four 12V batteries connected. This is a big deal for AGM's
See the attached drawing, ignore the NEG going to Ground, this is how RV batteries SHOULD be wired and most are not and RV owners replace batteries every year. Getting the power IN/OUT off one end of a paralleled set of batteries will kill the far ones.
I did take a picture of the connection and posted it earlier on this thread, but I made this diagram to help describe it. The 4AWG cables are within 1cm of each other in length.
Quote from: vtmaps on February 07, 2015, 06:55:53 PM
You can do much better than your diagram shows:
http://www.smartgauge.co.uk/batt_con.html
However, the most important thing to know about having four batteries in parallel is that it is a poor design choice to have so many batteries in parallel, even if they are wired to perfection.
The batteries are, at best, in an unstable equilibrium. Usually they are not even at equilibrium.
Looks like I used Method 3 - but yes it is still not a good thing. However it is all I have for now until the batteries are completely shot and then I will replace them with a proper design.
Quote from: vtmaps on February 07, 2015, 06:55:53 PM
Each of four batteries in parallel should be fused. The problems I just mentioned become dangerous when a cell in one battery shorts out. Of course, adding all those fuses means adding many more connections which are all potential points of failure. Because of the issues with monitoring and maintaining four parallel batteries, it is a good idea to include some switches (or circuit breakers) to be able to switch in or out any particular battery.
I have fuses I can add. Is adding to just the positive leads acceptable?
Quote from: vtmaps on February 07, 2015, 06:55:53 PM
The life of a battery bank is equal to the life of the shortest-lived cell. The lifespan of the cells is a Gaussian curve (bell shaped curve). With four 12 volt batteries, there are 24 cells. Hopefully, none of your 24 cells is in the lower tail of that curve, because it doesn't matter what the average lifespan of your cells is... all that matters is that one of your cells will pull down the whole bank when its time is up. If you had just 6 cells to make up your 12 volt battery, you would have a much greater chance of NOT having an outlier short-lived cell. In fact, you would even have a 1 in 64 chance of buying 6 above average cells.
This is why my next battery bank will be designed using 4V or 6V SLA batteries. If I am lucky, maybe LiFePO2 batteries!
I do think I have a serious problem with the bank. I hope it isn't a bad cell, but I'm also not sure where to start troubleshooting. I just know, as I've said previous, that two batteries charge/discharge at different rates then the other two. By ruff math it is about 15% difference. Whatever is going on is causing bank voltage to drop off quickly even with minimal load. It seems better after I tried doing an "equalize" charge, at the same voltage as Absorb, for 4 hours. The bank actually stayed at 12.9 for over an hour which it used to do for several hours.
I know there are problems with my design, both performance and NEC-related. I just need to fix what I can for the least amount of money right now and try to get another year or two out of the batteries before they are scrapped for a new bank. Any advice to help make this happen would be very helpful. I really appreciate all of the information everyone has provided on here, I just don't have a solution yet unless I missed it somewhere.
Quote from: kf4hzu on February 08, 2015, 08:17:17 AM
I do think I have a serious problem with the bank. I hope it isn't a bad cell, but I'm also not sure where to start troubleshooting.
As I mentioned earlier in this thread, you need to separate the batteries and try charging them one at a time.
It would be useful if you had a DC clamp ammeter... you could easily see how the current is dividing among the batteries while they are in parallel. That will help with the diagnosis. Of course, you will still need to separate them to achieve a cure. The inevitable cure may involve amputation of one battery from the bank.
When batteries get really out of balance, you may have some batteries discharging into other batteries when you turn the charger or controller off. You can see that with a DC clamp ammeter.
--vtMaps
Quote from: vtmaps on February 08, 2015, 11:38:41 AM
As I mentioned earlier in this thread, you need to separate the batteries and try charging them one at a time.
It would be useful if you had a DC clamp ammeter... you could easily see how the current is dividing among the batteries while they are in parallel. That will help with the diagnosis. Of course, you will still need to separate them to achieve a cure. The inevitable cure may involve amputation of one battery from the bank.
I do have a DC clamp meter; that is how I figured out that I have a problem with the differential in charge/discharge rates. I think I mentioned that. Once I separate them, what should I do? Should I leave only one connected to the Classic per day for 4 days and then reconnect them all? Or should I charge them individually with an AGM charger that plugs in to regular power? It only outputs 7 amps max and these are 55AH batteries.
Quote from: vtmaps on February 08, 2015, 11:38:41 AM
When batteries get really out of balance, you may have some batteries discharging into other batteries when you turn the charger or controller off. You can see that with a DC clamp ammeter.
How did I not think about this! I have yet to try that, but it makes complete sense, thanks! I'll do that today/tonight. The batteries are already in Absorb after only 3 hours of sun (11:50AM local time here). Should I pull all load and disconnect the CC, then test immediately, or wait until Float and then pull load/CC and test immediately?
Ok I waited until the batteries were float charging to test, so full 2.5hrs of Absorb at 14.7V and then float at 13.8V for about 2 hours.
I disconnected all load, got the clamp meter ready, and disconnected the charge controller. I immediately tested each battery lead several times to get the best reading. I found the highest power flow between the batteries was from the one I just happened to have connected the temperature probe from the Classic. 200mA was flowing out of it in to the other batteries at different rates. One was absorbing 100mA and the others were both alternating between 40mA and 70mA.
I then decided to do a load test on the system and see the differences, checking the outflow of power from the batteries out to the inverter. I connected a 350W pure sine-wave inverter and ran two laptops along with a small heater which was a little over 300W of load.
The biggest issue after only about 10 minutes ended up being heating on the wiring seemingly surrounding the 80A fuses. The highest reading was closest to the fuses at 112f/44.4c which seems fairly high for only about 25A of power draw from the batteries. This is very concerning since I have never had heating of the cables before. All my connections are crimp, and were very tight when I made the cables. Now I'm questioning all of it. Should I take everything apart and re-make it, or is there a better way?
Back to the original issue, the two outside batteries of the bank were reading ~5.5A draw, where the two inside batteries were reading ~6.5A draw. So just confirming I have an issue, and it could be two batteries.
Should I disconnect all 4 batteries and charge them separately? What would be the best next steps to help the battery balance issue, even if only temporarily?
Obviously I need to fix the wiring issues now too. Luckily none of the battery interconnects were warm at all. For reference, it is 68f/20c in the room and the batter interconnect cables felt just as cold as other surfaces in the room, including the 1/0AWG cables that were not in use at the time.
KF did you check the voltages during the time of the load test?
Both V and A, taken at the same time, gives us information, as well as SG readings.
Also when doing a load test it is good to monitor the Voltage DROP over time, every half hour or so depending on the rate of load relative to the battery rating
I did some tests on my NiCd cells every 30 minutes for the first 2 hrs then at 15 minutes for the next hour, with a .25C load.
What you are looking for is a rapid V drop. Once that happens stop the test.
Then you compare all cells that have been (equally) tested...
Voltage during the load test was 12.2V and eventually down to 12.1V after a half hour at the battery interconnect junction point (this increased the amps draw to about 26). It started at 12.4V which was due to the laptops being connected to the 175W inverter for several hours after dark. They are AGM batteries so I can't give you SG readings. The load test was done late this evening; where the no-load full disconnect test was done earlier today when the batteries were at full charge.
I'll try individual battery load testing as you suggested after the Classic gets the bank back up to "full" or as close to full as it thinks it is so I am starting at a known charge level.
Any comments about the wire heating?
Quote from: kf4hzu on February 15, 2015, 12:03:07 AM
Should I disconnect all 4 batteries and charge them separately?
Which part of my previous replies to this thread are you having trouble understanding?
Quote from: vtmaps on January 30, 2015, 05:52:38 AM
What you MUST do is separate those batteries and charge them and evaluate them one at a time.
Quote from: vtmaps on February 08, 2015, 11:38:41 AM
As I mentioned earlier in this thread, you need to separate the batteries and try charging them one at a time.
--vtMaps
Quote from: vtmaps on February 15, 2015, 06:14:34 AM
Which part of my previous replies to this thread are you having trouble understanding?
Sorry, it is a lot of information from different people and obviously not all getting processed.
The sun topped off the batteries today, so I will start testing tonight and tomorrow.
I ran some tests with the batteries separated. They all charged up and stayed topped off unlike when they are charged together. They also seemed to discharge at about the same rate with a 5 amp load over 5 hours. What I am seeing with them all connected together must be very small differences in the cables/connectors or the battery internal resistance. I'll try to narrow it down as time allows. For now, I don't think the batteries are too badly damaged yet.
On a good note, I got the parts and time to install a Pv disconnect. I also bought some UL-listed wire for it. So now I am NEC compliant all the way in to the charge controller. Attached is a picture of it. I have one breaker per pole, so when both are opened the panels are completely disconnected. Not sure if this is the right way to do it.
I did have a question about the breakers. Can you install then upside down, or will that cause something in them to fail? Also it says they are rated at 150V but the label on the side says 125V - why?
Quote from: kf4hzu on March 02, 2015, 01:18:04 AM
I did have a question about the breakers. Can you install then upside down, or will that cause something in them to fail? Also it says they are rated at 150V but the label on the side says 125V - why?
Well I believe those are MN DIN rail breakers and they are polarity sensitive. So they have to be connected correctly. See Ryan's video's
http://www.midnitesolar.com/video/videoPlay.php?video_ID=32&videoCat_ID=16 (http://www.midnitesolar.com/video/videoPlay.php?video_ID=32&videoCat_ID=16)
Hi KF4HZU,
Regarding the two different voltages on the DIN-Rail Breakers.
The manufacturer -- CBI -- has the 125 VDC Rating on the breaker. MidNite paid to have those breakers tested and Listed at 150 VDC. For the 150 V Rating to apply, IIRC, the breaker must be used in a MidNite Listed box/enclosure.
Regarding the permitted orientation of the MN breakers, I should not comment. The DIN Rail breakers do use a magnet to suck-out any arc, and that function MIGHT require that the breaker be oriented in an upright position with terminals on top/bottom, and perhaps the front label right-side up.
Did read on this Forum, that many of the MN breakers could be oriented in any position, except with the breaker handle facing straight up ... will let the someone like Robin comment on these details.
FWIW, Your installation looks nice, with good signage. Have Fun, 73, Vic
Quote from: Resthome on March 02, 2015, 03:12:09 PM
Well I believe those are MN DIN rail breakers and they are polarity sensitive. So they have to be connected correctly. See Ryan's video's
They are! And yep I read the spec sheet before installing them. That is part of the reason I asked about orientation, having them wired with the right polarity using 6 AWG wire pretty much takes up all of the space in the box.
Quote from: Vic on March 02, 2015, 03:35:27 PM
The manufacturer -- CBI -- has the 125 VDC Rating on the breaker. MidNite paid to have those breakers tested and Listed at 150 VDC. For the 150 V Rating to apply, IIRC, the breaker must be used in a MidNite Listed box/enclosure.
Ah that explains it. Guess I should have done more research! Thank you! Also, that is a MN "Baby Box" in the picture so I should be fine.
Quote from: Vic on March 02, 2015, 03:35:27 PM
Regarding the permitted orientation of the MN breakers, I should not comment. The DIN Rail breakers do use a magnet to suck-out any arc, and that function MIGHT require that the breaker be oriented in an upright position with terminals on top/bottom, and perhaps the front label right-side up.
Did read on this Forum, that many of the MN breakers could be oriented in any position, except with the breaker handle facing straight up ... will let the someone like Robin comment on these details.
I'll wait for Robin or someone else from MN to chime in. For now, I don't need to put more breakers in the box so no rush.
Quote from: Vic on March 02, 2015, 03:35:27 PM
FWIW, Your installation looks nice, with good signage. Have Fun, 73, Vic
Thank you! I am trying to be NEC-compliant and overall safe with the system. Learned a lot on here recently and reading other resources. I'll replace all the cables with UL-listed cable, instead of the welding cable I bought originally, as budget permits.
Edit: The cable I used for the box is rated at 600VDC. It is not very flexible. Any recommendations on cable and where to buy it? I need more 6AWG, need some 4AWG and 1/0 or 2/0 AWG.