diy solar

diy solar

Signature Solar LiFePOWER4 batteries charging/discharging differently. My fault?

notevennothing

New Member
Joined
Mar 28, 2022
Messages
8
I received and installed four Signature Solar EG4 LiFePOWER4 batteries, along with the six-battery rack, in January and was happy as a clam. I noticed pretty soon that they don't charge/discharge identically. Judged purely through the four-LED SOC indicator on the battery, one battery would achieve full SOC much later than the others. Since this battery was the furthest down on the bus bars, I figured that was the problem, and didn't think much about it because I was working on higher priority stuff.

Nine days ago, we moved in, bringing new loads.

As we added additional loads, I quickly noticed that the bottom two batteries were slower to hit full SOC. In fact, the bottom-most battery no longer hits full SOC at all. All the other batteries fill up, and the inverter/charger switches to float charge, but only three of the SOC indicator LEDs are lit on that one battery. A bit of Googling led me to Will's recent battery enclosure Youtube video where he mentioned switching one of the leads on the bus bars to the bottom. I did that on Saturday morning thinking it would fix all my problems, but haven't noticed an impact.

One factor may have been that I played with my inverter/charger's settings to maximize the life of the batteries according to a post here on the forum. That might have been a mistake. My thinking is that the battery's BMS handles this, and that my modified inverter/charge controller settings may be part of the problem I'm seeing, if not the whole problem.

I'm looking for suggestions. My plan was to hook a Raspberry Pi running Solar Assistant up to the batteries so that I could better see what is going on, and to adjust my inverter/charger's charging curve to delay its switch to float charging. If the problem persists, I will reach out to Signature Solar, but I want to eliminate my own mistakes as the cause first.
 
I received and installed four Signature Solar EG4 LiFePOWER4 batteries, along with the six-battery rack, in January and was happy as a clam. I noticed pretty soon that they don't charge/discharge identically. Judged purely through the four-LED SOC indicator on the battery, one battery would achieve full SOC much later than the others. Since this battery was the furthest down on the bus bars, I figured that was the problem, and didn't think much about it because I was working on higher priority stuff.

Nine days ago, we moved in, bringing new loads.

As we added additional loads, I quickly noticed that the bottom two batteries were slower to hit full SOC. In fact, the bottom-most battery no longer hits full SOC at all. All the other batteries fill up, and the inverter/charger switches to float charge, but only three of the SOC indicator LEDs are lit on that one battery. A bit of Googling led me to Will's recent battery enclosure Youtube video where he mentioned switching one of the leads on the bus bars to the bottom. I did that on Saturday morning thinking it would fix all my problems, but haven't noticed an impact.

One factor may have been that I played with my inverter/charger's settings to maximize the life of the batteries according to a post here on the forum. That might have been a mistake. My thinking is that the battery's BMS handles this, and that my modified inverter/charge controller settings may be part of the problem I'm seeing, if not the whole problem.

I'm looking for suggestions. My plan was to hook a Raspberry Pi running Solar Assistant up to the batteries so that I could better see what is going on, and to adjust my inverter/charger's charging curve to delay its switch to float charging. If the problem persists, I will reach out to Signature Solar, but I want to eliminate my own mistakes as the cause first.
I would get the data cable and look at what is going on with the batteries using a Windows laptop.
 
I'm a Linux nerd. Any Windows laptop becomes a Linux laptop in short order. I was going to grab Solar Assistant anyway, as I have half-a-dozen Raspberry Pi sitting around.

But I get your point: Look under the hood of the batteries.
 
I'm a Linux nerd. Any Windows laptop becomes a Linux laptop in short order. I was going to grab Solar Assistant anyway, as I have half-a-dozen Raspberry Pi sitting around.

But I get your point: Look under the hood of the batteries.
No not solar assistant. I am talking about the free battery monitoring software for that battery pack that is on Signature Solars website.
You just need a $19 USB to RS485 converter and you can see what is happening with each pack down to the individual cells.
 
he mentioned switching one of the leads on the bus bars to the bottom.

If I'm reading this right, you had the top battery connected to charge/loads and the others "cascaded" after it?
Not the proper way to do it, is it? You should draw/charge from the positive of one and the negative of the one furthest away from it.

I did that on Saturday morning thinking it would fix all my problems, but haven't noticed an impact

Yeah well. It might take a while... but it's probably best to re-balance them. Meaning, charge them individually to full SOC, the re-parallel them.

As to Linux and Windows programs... COM ports in Wine are a bit of a bitch, but with some effort they can be... guessed :·)
Add yourself to the dialout group. Logout/in.
Look in ~/.wine/dosdevices. Open terminal, ls -la. See what com links to ttyUSB0, or whatever you have your cable connected to.
-
 
I have one Trophy - 220ah, 9 - 280aH batteries ...132kWh. None charge and discharge equally. Absorb now 55.0/Float 54.0V. Any higher and my Trophy and a couple of other batteries will disconnect from cell ov.
 
Last edited:
No not solar assistant. I am talking about the free battery monitoring software for that battery pack that is on Signature Solars website.
You just need a $19 USB to RS485 converter and you can see what is happening with each pack down to the individual cells.
I understand what you mean. And thanks for the advice. I just don't have any laptops that run Windows, nor do I want any. Solar Assistant does the same, and more, just in a way that works for me. (Solar Assistant runs on a Raspberry Pi, talks to the batteries over RS485, and gives me a web interface to graphs, stats, and configuration.

When I get some data, I'll share it here.
 
If I'm reading this right, you had the top battery connected to charge/loads and the others "cascaded" after it?
Not the proper way to do it, is it? You should draw/charge from the positive of one and the negative of the one furthest away from it.



Yeah well. It might take a while... but it's probably best to re-balance them. Meaning, charge them individually to full SOC, the re-parallel them.

As to Linux and Windows programs... COM ports in Wine are a bit of a bitch, but with some effort they can be... guessed :·)
Add yourself to the dialout group. Logout/in.
Look in ~/.wine/dosdevices. Open terminal, ls -la. See what com links to ttyUSB0, or whatever you have your cable connected to.
-
You aren't quite reading me right. The Signature Solar battery rack comes with two bus bars, a negative and positive. All the batteries' positive leads are connected to the positive bus bar, and the negatives to the negative bus bar, making a big parallel battery. (I'm not talking down here, just trying to fully illustrate.)

I had the leads from the solar MPPT charge controller running to the top of the bus bars. Both leads, the positive and the negative, at the top. So the batteries had different distances from the charging leads. The correct way is to have the charging leads connected on opposite ends, "diagonally connected".

You know what? I'll just link to the relevant part of Will's video. He explains it well, with pictures:

Thanks for the Linux tips. I'm not actually trying to run Signature Solar's provided monitoring software through Wine. If I do, and I might, I'll use your suggestions.
 
I just fired up 5x EG4's in the same battery rack with the positive and negative cable at the top of the buss bar. My amp draw is pretty low but you can see the different discharge rates with this design shown below.

Going to monitor for a week and then move the negative cable to the bottom of the rack on the buss bar to see the difference is.


dsff.JPG
 
Both leads, the positive and the negative, at the top.

Well, this is what I don't understand. You either have the positive at the top and the negative at the bottom or the other way around.
Which is what the video capture shows (negative at bottom) - and what I have on mine.

What does "both leads at the top" mean, then?
-
 
One factor may have been that I played with my inverter/charger's settings to maximize the life of the batteries according to a post here on the forum. That might have been a mistake. My thinking is that the battery's BMS handles this, and that my modified inverter/charge controller settings may be part of the problem I'm seeing, if not the whole problem.
What if you use the signaturesolar recommended values? Unless you also changed the settings in the BMS i would keep to the values they recommend.
 
I just don't have any laptops that run Windows, nor do I want any.
I feel the same way, but I have a few windows-only apps I want to run, so I give in and run dual boot and avoid as much as possible connecting to the network with win booted. I've had some success with WINE. Some win apps work, and some don't.
 
Well, this is what I don't understand. You either have the positive at the top and the negative at the bottom or the other way around.
Which is what the video capture shows (negative at bottom) - and what I have on mine.

What does "both leads at the top" mean, then?
-
It means I was doing it wrong. However, I've since corrected it.

If you watch ten seconds of the Youtube video I linked to, you will see what I did, before and after.
 
I think I figured out the problem(s).

First, not having the charging leads (from my MPPT charge controller) connected at opposite ends of the bus bars (ie. diagonally), wasn't helping anything.

Second, I had set the MPPT charge controller to bulk charge to 56.0V, absorption to 56.0V, and float to...something lower. This was cutting off the batteries well before they had reached full SOC, and circumventing the battery's charge controllers. I've now set my MPPT charge controller to 57.5V/56.5V/56.5V (B/A/F) according to the battery spec sheet. I actually set it 0.5V higher yesterday to get all the batteries to full SOC. The charge controller on the battery should prevent any issues with that. (Right?)

Sort of an interesting story:

Like I explained in my opening post, I could only get three of the four batteries to full SOC, one never went over 75% . I don't even think I was getting those three to 100% SOC before the MPPT charge controller cut them off. In any case, my dear wife decided to do laundry late in the afternoon three days ago. This was the first time she had attempted this on our solar system, and did so without my knowledge as I wasn't on-site. So I get home, realize what she's doing, and see that the dryer draws about 5.5kW. Combined with everything else in the house, we were over our inverter's 6kW rating by about 250W. I turned off a couple of non-essentials to get us below 6kW and let the load of laundry finish drying. The system held up fine, but two of the batteries were now at 75% SOC and the other two were at 50% SOC.

She then decides to bake and cook, using the electric oven despite our having a masonry heater that had been fired (so its oven was at baking temperature). I tell her that isn't a good idea given the state of the batteries and the fact that the next day (two days ago) is forecast to be cloudy with snow, dreary enough that we would continue to draw from the batteries for most of the day. She ignored me. (Which is par for the course.)

The next day was exactly as forecast. By the evening, we had two batteries at 50% SOC, and two at 25%. By the time we turn in, the two at 25% had entered low voltage shutdown, but the others were holding at 50% SOC. She is freaking out. I'm not worried because the forecast for the next day (yesterday) is forecast to be sunny, and as long as my wife is asleep our electrical load is minimal. Despite my consoling her that we aren't going to lose power overnight, and that the food in the fridge and chest freezer are fine, she continues to worry late into the night. I sleep soundly.

The next day is clear and sunny. I leave for work with the batteries in the same state (50%, 50%, empty, and empty). By noon, we were at 75%, 75%, 50%, and 50%). I was able to get home in the early afternoon and witness three of the batteries hitting 100% SOC, but I had to raise the bulk charging voltage to 57.5V and force a charge cycle on the MPPT charge controller to get the "problem" battery to charge beyond 75% SOC.

The funny thing is that the problem battery jumped from 75% SOC and charging, to 100% SOC in a few seconds. I think that the battery had never been fully charged and the firmware finally got to set what 100% SOC was.

Anyways, we had a normal evening, even putting a couple of loads of laundry through the washer and hanging them to dry inside. By morning all of the batteries had only just hit 75% SOC, and all within a few minutes of one-another. My wife has calmed down, and realized that hanging laundry isn't a big deal. She is looking for an efficient electric dryer. I don't really think such a thing exists, in the sense that the low power models probably take much longer to dry, so that the total energy use per load is the same. (I would love to be corrected.)

I'll keep an eye on the batteries, and should be able to hook up Solar Assistant sometime over the next couple of days, but at this point I believe the problem was mostly me. The only thing I could blame on Signature Solar is the design and instructions for their battery cabinet. It should be designed so that the charging leads (from the solar unit, Growatt, MPP, or whatever) attach at opposite ends. At the moment, connecting it properly would involve drilling and tapping. I've used some stainless steel washers to make it work, but it isn't perfect.

So at this point, I'm content with the batteries. That may change when I have a deeper look at the batteries, and have had more time with them, but I'm good for now.
 
I have the same eg4s and cabinet from signature, are you saying that I should reverse the negative buss bar in the cabinet so that the negative battery cable from the inverter attach at the bottom of the buss bar rather than at the top ?
thanks
 
You aren't quite reading me right. The Signature Solar battery rack comes with two bus bars, a negative and positive. All the batteries' positive leads are connected to the positive bus bar, and the negatives to the negative bus bar, making a big parallel battery. (I'm not talking down here, just trying to fully illustrate.)

I had the leads from the solar MPPT charge controller running to the top of the bus bars. Both leads, the positive and the negative, at the top. So the batteries had different distances from the charging leads. The correct way is to have the charging leads connected on opposite ends, "diagonally connected".

You know what? I'll just link to the relevant part of Will's video. He explains it well, with pictures:

Thanks for the Linux tips. I'm not actually trying to run Signature Solar's provided monitoring software through Wine. If I do, and I might, I'll use your suggestions.
exactly as will stated inthe video from 3:14;onward.

drill a hole at one of the busbars and connect the line going to the inverter there
 
I have the same eg4s and cabinet from signature, are you saying that I should reverse the negative buss bar in the cabinet so that the negative battery cable from the inverter attach at the bottom of the buss bar rather than at the top ?
thanks
Yes, if you are seeing problems. One lead from the inverter attached at the top of a bus bar and the other lead attached at the bottom. The bus bars are not set up for this. You can make it work by adding a stainless steel washer, and carefully tightening a bolt, but it would be better to drill and tap a proper hole so that the bigger bolts at the top of the bus bar can be used. I don't think you can just flip one bus bar, because the holes for mounting the bar to the cabinet aren't spaced in such a way that it would work. (I'm pretty sure that is the case, but I'll check it again tonight.)

In my case, I think I was seeing an interaction between my inverters charging settings and the non-optimal (ie. non-diagonal) way the inverters battery leads were connected to the bus bars. Had my inverter been properly configured for the batteries, the issue might have been much less severe.
 
A quick update: Since changing my inverter's charging settings, I haven't noticed the batteries being in a different state of charge...as viewed with the four charging indicator LEDs on the front panel.

I'll know more when I hook up to the batteries via RS485, hopefully this weekend. (We've just moved in to our house. I'm juggling a lot of tasks. Get off my back! ;))
 
Back
Top