diy solar

diy solar

Why Has My 6-Month-Old "Once Perfect" Solar System Failed?

jesfl

New Member
Joined
May 17, 2020
Messages
104
HELP, please!

I am a newbie. And, I am near to conceding defeat. My 6-month-old DIY solar system in my RV has failed.

It worked perfectly for 4 months. On heavy 120-volt appliance use days, I would consume 10% to 15% of the total battery amp hours. Then, even on cloudy days, the system would be recharged to 100% by 2-3:00 pm in December in the Arizona desert! I was ecstatic.

Then something happened and I have no idea what? I've spent weeks reading everything I can find (including here on the DIYSolarForum to a blurry-eyed state), asking questions, rewiring, replacing parts, testing, etc. But, since I cannot find the problem, I cannot fix it.

If anyone has ANY IDEAS WHATSOEVER about what might have caused such problems, I welcome any and all ideas no matter what they might be. I am truly desperate.

(1) What might cause my solar charge controller to stop working normally as it did once and now work only in "full mode" sending only 1-3 amps charge to the batteries? It worked perfectly for 4 months with normal charging rates of 20-28 amps until the batteries were nearly full, and then appropriately went into the lower amperage rate "full mode."

(Note: I installed the controller out of the box without adjusting anything, knowing the factory settings would not take full advantage of myLiFePO4 batteries, but that I could tweak it later as I learned more about the technology.)

(2) What might cause my three LiFePO4 batteries to fail simultaneously? (Note: 175 amp ANL fuses on pos/neg battery connections are OK, so not surge or short.) It seems to me the odds are astronomical of all three batteries failing simultaneously, save for some catastrophic event?

ANY IDEAS are helpful, at this point. I'm at a huge wall I can't get over.

More detail about my system and my actions is below.

__________________________________


SYSTEM:

-- 8 - 100 watt solar panels, wired series (2 sets of 4 panels) to parallel;
-- 40 amp solar charge controller,
-- 3000 watt inverter (now completely disconnected while I am testing, since it would not run any longer except minimally whenever there was incoming solar);
-- 3 - 150 Ah LiFePO4 batteries;
-- wiring is purposely short: PV roof panels to the controller is 10 AWG of 13 feet; controller to batteries is 6 AWG of 22 inches; battery wiring and batteries to the inverter is 1/0 AWG of max 14 inches but only 9 inches battery to battery,
-- 175 amp ANL fuses on pos/neg battery terminals; 300 amp ANL fuse from batteries to inverter; 50 amp breaker (extra) on controller positive output to batteries, 2-pole 50 amp 400V breaker on PV panels to controller wiring.

(Note: I’m trying to draw a wiring diagram, but I’m on about the 10th draft and having some trouble getting it understandable. It’s my first ever diagramming try.)

SOLAR CONTROLLER

-- The solar charge controller now will only function in "full mode" (per indicator lights) charging at only a 1-to-3-amp rate but never fully charging as it did previously, typically at mid-day with 26-28 amps of charge into batteries via my meters.

-- The solar controller manufacturer says it is not possible anything in the controller changed without me doing something. ( I installed it right out of the box with no settings adjustments, figuring I could tweak it later as I learned more.) The manufacturer says it must be the batteries. Is it even possible that solar controllers fail somehow or behave erratically?

-- I have checked solar volts/amps from the PV panels to the controller seems good, i.e. at 9:00 am this morning 71.5 volts & 9.5 amps incoming! The voltage output from the controller reads about 14.5 volts consistently. But, the remote meter wired to the controller and my inside voltage monitoring meter (through a shunt) read only a very low 1-3 amps (the low near-full-charge rate) going into the batteries.

-- Meanwhile, the batteries at best read 12.5 volts to 13.1 volts with solar and battery master switch off (tested with multimeter). Obviously, that's far from the full charge for lithium batteries.

-- What would cause the controller to function only in this "full batteries mode? Any of your knowledge about this type of problem will help me learn?

LiFePO4 BATTERIES (3)

-- The LiFePO4 battery manufacturer/distributor (Chinese, my bad decision) will not respond to my inquiries about failure possibilities, of course.

-- What might cause all three of my LiFePO4 batteries -- only in use 6 months -- to stop charging? As I said, aren't the odds against that? Wouldn't I know it if something catastrophic happened?

-- I have disconnected all batteries and then reconnected them one at a time to try to charge each when isolated. None will charge.

(Note: I am boondocking and do not have shore power available to try to charge the batteries via a 120-volt lithium-capable charger.)

-- Is it possible the BMS units in the batteries have failed somehow? Again, to me, it seems against the odds of all three BMS units failing simultaneously?

-- Does anyone have any suggestions of how I can "test" the batteries further to identify a problem? (Note: my electrical skills are very limited, even with my very basic multimeter.)

Thank you so much in advance for any ideas, whatsoever. I will be very grateful.

Jim S.
 
All that writing with no mention of the brand and model of your 40A SCC?
I read your previous posts to see Epever BN
Time to check the Epever settings and customize with the USER option for LFP

800 watts panels into 12V battery is 66A but you only have 40A

What is the sales link to your LFP batteries?
 
"71.5 volts & 9.5 amps incoming"
Reported by the controller? That's 700W that must be going somewhere.

"But, the remote meter wired to the controller and my inside voltage monitoring meter (through a shunt) read only a very low 1-3 amps"
at 13V, just 13 to 39W?

If the batteries have BMS which would disconnect to prevent over-charge, would it make sense to wire one PV panel directly to it, see if it charges?

People here have described Epever SCC needing to get reset, repeatedly.
 
"(Note: I installed the controller out of the box without adjusting anything, knowing the factory settings would not take full advantage of myLiFePO4 batteries, but that I could tweak it later as I learned more about the technology.)"

I wonder the betteries can be damaged due to the Equalization function (For lead Acid battery only) was not turned off?
 
"(Note: I installed the controller out of the box without adjusting anything, knowing the factory settings would not take full advantage of myLiFePO4 batteries, but that I could tweak it later as I learned more about the technology.)"

I wonder the betteries can be damaged due to the Equalization function (For lead Acid battery only) was not turned off?

That is easy to diagnose

1616215364397.png

(And it would mean BMS wasn't doing its job.)
 
Wow! First, Thanks to everyone for all the replies. Yes, I'm way over my head.

Additional facts requested:

(1) Solar controller = EPEVER Tracer 4215BN. From my solar panels' specs when wired series-parallel, I calculated a max of 79.32V and 12.52A possible. Panels are flat mounted, so I would never get that max PV output.

AFTER the failure, I spent hours and finally figured out how to change the Tracer 4215BN to the gel setting. But, I still can't figure out what settings I should use for lithium in the user settings. here and elsewhere. I found several exchanges here, but I got lost in all those numbers which do not match the settings "names" in the Tracer 4215BN/MT50 system.

(2) The LiFePO4 batteries are a Chinese BTR brand, but the seller from whom I purchased has disappeared from eBay and is not responding to the direct email they/he once gave me.

The link to one of two new eBay sellers is: https://www.ebay.com/itm/Btr-12V150...motedViewItems&_trksid=p2047675.c100752.m1982

I emailed both current sellers I found on eBay asking for parameters/specs for solar charging the battery. One responded by sending me a link to a 120V charger.

I still get sick to my stomach every time I think about that bad battery buying decision I made last August.

(3) Several of you are pointing to the BMS, which has become my greatest fear.

I have now learned a little about the charging phases of lead acid batteries vs. lithium (but not enough). And I belatedly see how just using the out-of-the-box factory setting might have damaged the BMS/batteries. But, why did it take 4 months to happen? If BMS damage/failure is what happened, didn't there have to be another trigger for that event?

I did not think about the possibility of hooking one solar panel directly to the battery as a test. I'll try to figure out how to disconnect all but one panel on the roof and do that test.

If the diagnosis is BMS damage/failure, is there any solution whatsoever?

Thanks again to all of you for all of your information. It is immensely helpful to me.

Jim S.
 
This does sound like the BMS units inside of the 3 batteries have all gone to open protect mode. If an equalization cycle did run, it very well could have pushed the batteries up to where the BMS units all turned off due to over voltage protect. Do you have a volt meter with you? Can you try connecting a 12 volt light bulb to one of the batteries? What brand are the batteries? You just answered BTR. Can you find instructions for resetting the BMS after a fault? I just found this site
Is this the battery you have?
There is not a lot of information on the page, but they do have a "Contact Us"
Try sending them your question about how to reset the BMS if it went into protect.

Can you get a hold of a lead acid battery you can connect and see if the charge controller and inverter will work with that to isolate if the problem is just the batteries? For a short term test, you could use your vehicle starter battery. But make sure to monitor the voltage so you don't hurt that.
 
Thank you so much.

I am continuing to study and realizing it is most likely the BMS, as you and others have said. BUT, perhaps there is a smidgin' of hope? I had no idea built-in BMS could be reset.

Yes, these are the batteries I have. I will contact them immediately to try to get BMS reset information and settings for the batteries for a solar charge controller.

Don't have a 12-volt light bulb. But . . .

I already bought a cheap marine SLA battery at Walmart two weeks ago to start my built-in genset. I tried to use the lithium-set 120V 10 amp battery charger that came with the batteries to charge a single (disconnected/isolated) battery. But it is too slow for using the genset. After 6 hours of run time, the voltage increased only from 13.01 volts to 13.16 volts. Or, does that prove it is the battery's BMS limiting the charging?

So, if I disconnect the lithium batteries and connect the lead-acid 12-volt battery, I also need to reset the Tracer charge controller to SLA, to be safe? Correct?
 
Wow! First, Thanks to everyone for all the replies. Yes, I'm way over my head.

Additional facts requested:

(1) Solar controller = EPEVER Tracer 4215BN. From my solar panels' specs when wired series-parallel, I calculated a max of 79.32V and 12.52A possible. Panels are flat mounted, so I would never get that max PV output.

AFTER the failure, I spent hours and finally figured out how to change the Tracer 4215BN to the gel setting. But, I still can't figure out what settings I should use for lithium in the user settings. here and elsewhere. I found several exchanges here, but I got lost in all those numbers which do not match the settings "names" in the Tracer 4215BN/MT50 system.

(2) The LiFePO4 batteries are a Chinese BTR brand, but the seller from whom I purchased has disappeared from eBay and is not responding to the direct email they/he once gave me.

The link to one of two new eBay sellers is: https://www.ebay.com/itm/Btr-12V150Ah-LiFePO4-Lithium-Phosphate-Battery-12-Volt-150-Amp-Hour-Deep-Cycle/124261623032?_trkparms=aid=555023&algo=HOMESPLICE.SIMRVI&ao=1&asc=20190711100440&meid=ac2f2bbbe6c44d10bb79a557383d0600&pid=100752&rk=1&rkt=24&mehot=pf&sd=124274829552&itm=124261623032&pmt=1&noa=0&pg=2047675&algv=SimplAMLv5hPointwiseWebNoToraCoCoViewsNoHighIdfOrRoundRobinBlenderWithPromotedViewItems&_trksid=p2047675.c100752.m1982

I emailed both current sellers I found on eBay asking for parameters/specs for solar charging the battery. One responded by sending me a link to a 120V charger.

I still get sick to my stomach every time I think about that bad battery buying decision I made last August.

(3) Several of you are pointing to the BMS, which has become my greatest fear.

I have now learned a little about the charging phases of lead acid batteries vs. lithium (but not enough). And I belatedly see how just using the out-of-the-box factory setting might have damaged the BMS/batteries. But, why did it take 4 months to happen? If BMS damage/failure is what happened, didn't there have to be another trigger for that event?

I did not think about the possibility of hooking one solar panel directly to the battery as a test. I'll try to figure out how to disconnect all but one panel on the roof and do that test.

If the diagnosis is BMS damage/failure, is there any solution whatsoever?

Thanks again to all of you for all of your information. It is immensely helpful to me.

Jim S.
I'll wager the BMS is in protect mode. I doubt all 3 batteries would fail at once. These have a 100a rating per battery and with a 3000w inverter, if one battery went into protect hitting low charge disconnect, then it could bring the other 2 down with overload if you were taxing the inverter heavy. Either that or the CC is not allowing a charge to wake up the BMS's.
 
Wow! First, Thanks to everyone for all the replies. Yes, I'm way over my head.

Additional facts requested:

(1) Solar controller = EPEVER Tracer 4215BN. From my solar panels' specs when wired series-parallel, I calculated a max of 79.32V and 12.52A possible. Panels are flat mounted, so I would never get that max PV output.

AFTER the failure, I spent hours and finally figured out how to change the Tracer 4215BN to the gel setting. But, I still can't figure out what settings I should use for lithium in the user settings. here and elsewhere. I found several exchanges here, but I got lost in all those numbers which do not match the settings "names" in the Tracer 4215BN/MT50 system.

(2) The LiFePO4 batteries are a Chinese BTR brand, but the seller from whom I purchased has disappeared from eBay and is not responding to the direct email they/he once gave me.

The link to one of two new eBay sellers is: https://www.ebay.com/itm/Btr-12V150...motedViewItems&_trksid=p2047675.c100752.m1982

I emailed both current sellers I found on eBay asking for parameters/specs for solar charging the battery. One responded by sending me a link to a 120V charger.

I still get sick to my stomach every time I think about that bad battery buying decision I made last August.

(3) Several of you are pointing to the BMS, which has become my greatest fear.

I have now learned a little about the charging phases of lead acid batteries vs. lithium (but not enough). And I belatedly see how just using the out-of-the-box factory setting might have damaged the BMS/batteries. But, why did it take 4 months to happen? If BMS damage/failure is what happened, didn't there have to be another trigger for that event?

I did not think about the possibility of hooking one solar panel directly to the battery as a test. I'll try to figure out how to disconnect all but one panel on the roof and do that test.

If the diagnosis is BMS damage/failure, is there any solution whatsoever?

Thanks again to all of you for all of your information. It is immensely helpful to me.

Jim S.
Does your battery look like this?
 

Attachments

  • s-l400.jpg
    s-l400.jpg
    16.9 KB · Views: 5
If you're lucky, BMS just needs some sort of reset.
If BMS failed but disconnected cells, you could open the packs and check individual cell voltages. You could wire the cells in series and/or parallel to fit your requirements and attach a new BMS (just as people here buy individual cells and assemble their own battery.)
 
If you're lucky, BMS just needs some sort of reset.
If BMS failed but disconnected cells, you could open the packs and check individual cell voltages. You could wire the cells in series and/or parallel to fit your requirements and attach a new BMS (just as people here buy individual cells and assemble their own battery.)

Now that is a crazy frightening-to-me thought . . . "open the packs . . . ?" Truly last resort for my limited newbie skills!

But, it does spark a curiosity. I saw an "external" supposedly high-end BMS for sale online. To use that, I would have to disconnect the (3) existing internal BMS units? Regardless, to use a new, external BMS, I suppose it still means "opening the packs?" Is it one BMS per pack? Or, worst case, if I open packs and remove the existing BMS units, is one BMS unit capable for all 3 "packs" of cells?
 
If your batteries are 12V, they contain 4 cells. BMS has to connect wires to all cells, so it is inside.

If you use four batteries in parallel, it is possible to connect four 12V batteries in parallel, all cells in parallel with a single 4s BMS. It would have to support your total current.

If you open it, do NOT short terminals of cells because they can deliver thousands of amps. Electrical tape insulating wrench handles and sockets is a good start, and remove metal jewelry.

Other people here DIY batteries, so read up on it.

But of course, figuring out how to reset BMS is the starting point.
 
It seems unlikely that all three BMS would "fail", but it's possible if one BMS failed and the other two had to handle the load, then a second failed and the third BMS gave up. This is what @Zwy was saying in his above post.

If you're not getting any support from the battery vendor, then opening up the battery isn't going to void your useless warranty. In other words, what do you have to lose? Take a look at the Will Prowse videos on YouTube where he takes apart batteries to see what they're made of. That might be helpful to you.

Did you get a voltmeter as suggested multiple times above? Poking around with a voltmeter is going to be critical to tracing where the problem is.
 
Did you see the battery pic that is packaged like an ebike battery ?
You are assuming that is 4 prismatic cells but could be cylindrical cells.

He has Three 150Ah batteries not 4.

If your batteries are 12V, they contain 4 cells. BMS has to connect wires to all cells, so it is inside.

If you use four batteries in parallel, it is possible to connect four 12V batteries in parallel, all cells in parallel with a single 4s BMS. It would have to support your total current.

If you open it, do NOT short terminals of cells because they can deliver thousands of amps. Electrical tape insulating wrench handles and sockets is a good start, and remove metal jewelry.

Other people here DIY batteries, so read up on it.

But of course, figuring out how to reset BMS is the starting point.
 
Thanks.

I am going to wait overnight to see if I get any response from the battery company about a procedure for resetting the BMS? Since it is the weekend, come to think of it, I probably have to wait until Monday to see if there is a response.

Yes, I have a basic voltmeter (multimeter), but knowing how to use it -- other than simple checking voltage and amperage -- is another problem. Never used one until recently. The one I have has only 6 possible settings.

I now have the system hooked up to a single Walmart 12V marine SLA battery. I also changed the EPEVER controller setting to "Sealed." (At least I finally learned how to set the controller. That's progress.) My solar is turned on and charging. I am running the inverter to charge some batteries and use the computer and TV. With the single SLA, I'm also able to start my genset, if necessary. So it makes boondocking more comfortable until I have another solution.

Am I correct that a 12.6 volt SLA battery is "half empty" when it is at about 12.15 volts?

Re: Will Prowse YouTube Videos . . . I actually built my system from his "basic system" video. I've also watched a few of his tear-apart-batteries videos, which is pretty much why I stopped myself from thinking about saving more $$$'s by buying cells and building some batteries. I've somehow gotten this far, but it all remains quite overwhelming to me.

Absent any other solution, perhaps I could "save my own a**" by tearing apart the batteries I have and rebuilding with new, better BMS? I am just praying I don't have to do that.

As an aside, however, I am beginning to see and wonder whether a rebuild with new BMS might get me better, longer-lasting batteries, perhaps?

With some distant hope out there, and my single SLA battery working, I'm going to try to be patient and continue learning.

Thank you, again, to all!

Jim S.
 
Using the multimeter, set it to DC volts and check the voltage at these points:

Where the solar panels come into the trailer, just before the circuit breaker and after the circuit breaker
The input port where the PV power comes into the solar charge controller
The output port of the solar charge controller
At the battery terminals

Do you have a battery monitor that tells you the state of charge of the battery bank?
Is your inverter an inverter/charger? If not, how are you charging the battery bank from the generator?
 
May be you should show us the pictures of the batteries you have in hand, and if you take it apart then we can tell at what test points you need to test and what to look for once we see the inside of your batteries.
 
Thanks to all, once again. I am so appreciative. With every bit of information and from your questions I learn new, valuable things.

I will take new multimeter readings as suggested in the (Arizona) morning when it is light again.

I am still hopefully awaiting information from the battery manufacturer about how to reset the BMS.

As mentioned earlier today, I disconnected the lithium batteries and connected a single SLA so I have some minimum battery power. (Also reset my solar controller to "sealed" battery type.) This afternoon with the solar panels turned off, I purposely ran the SLA battery down about 15% Then, I turned the panels back on and watched the behavior of the solar controller. As it should, it started recharging the battery properly -- first at a higher amperage level and as the battery got near full, the reduced rate. So, it appears the controller is working properly in the sealed battery mode.

I assume that the controller seems to be working properly on the SLA battery is another validation of the BMS problem with my lithium batteries?

To answer other pending questions:

- Yes, I have an AiLi meter for monitoring the battery, plus the wired-in MT50 meter for the solar controller also monitors voltage and charge rate.

-- But, I am still not 100% clear on what voltage = a 50% charge in a standard 12.6 volt SLA battery? One calculation example led me to 12.15 volts as a 50% SOC on an SLA battery. Is that correct, so I can watch it carefully?

-- As to how I (normally) charge the lithium batteries with the built-in generator, I purchased and installed a Progressive Dynamics replacement, lithium-capable converter for the WFCO power panel in my RV. The solar was working so well, I never used it to charge. On a couple of occasions when I "exercised" my genset, I briefly tested the converter to assure it was charging. Note: I also always turned off solar when I checked the charging for the converter as the genset was running.

-- Yet, here's just another part of my current pickle. I cannot charge the batteries now (or at least try to charge them) by running the genset and using the PD converter because there is not enough power in the lithium batteries to start the genset. I have started it and run it a couple of times (and will do so again tomorrow) using the SLA battery. But, with the SLA connected and the lithiums disconnected, I can't charge the lithiums. There must be a way to connect the SLA battery itself directly to the genset (Cummins Onan 3600 propane) for starting it (with the lithium batteries still connected). I just have not yet figured that out, amidst everything else.

-- When I tried to charge the lithium batteries previously by running the genset, I used the 120 AC power to run the three 10 amp chargers that came with the batteries. But, it did not work because of the slow charge (and I ran out of propane for running the genset).

Given that it is becoming more and more clear that I have a BMS problem, trying to charge using the 120-volt chargers is probably a waste of time, it seems to me.

Tomorrow morning, I'll review all of the generous comments here, take another round of voltage readings, and continue my learning.

I will also try to complete a wiring diagram and take some photos of the batteries, as installed.

Of course, I pray I can reset the BMS, that it is the problem and can be fixed, and that I never have to tear the batteries apart and send pix of them in that problematic-for-me state.

Thank you, thank you!

Jim S.
 
It's a bit off the wall, but you could power up all three of the 10 amp chargers and connect all of them to just one of the LiFePO4 batteries giving it a 30 amp charge, which is well below a .5C charge rate, so safe for the battery.

I think your test with the solar charge controller and the SLA battery was good enough to rule out the solar charge controller as the problem. Or perhaps the least of your problems.
wink.gif


I too have an Onan on-board generator. Mine is the 5500. I explicitly chose to NOT use my LiFePO4 battery bank to start my generator. The generator pulls a buttload of amps at startup, more than my two BMS combined are rated for. Another member tested his generator with the same BMS I have and it did fine. But if your BMS isn't rated as high as mine is, that could be part of the problem. Starting the generator off of the LiFePO4 battery may have damaged the BMS. This is all just speculation at this point, but it is something to consider.

I retained a small lead acid battery on the tongue of my trailer that I use to start the generator. It is a completely separate circuit from the LiFePO4 "house" batteries.

What are the charge/discharge specs for your batteries? This would be measured in amps.
 
Back
Top