Solar input drops in half after a few hours

[ggh]

Hi everyone,

I built a DIY system with the help of Will's youtube video. Its a 1012LV-MS 12v 1000 W alliin-one inverter with 12v 4s overkill BMS, and four 300 Ah CATL Lifep04 cells. I just finished building it and during the first couple days of testing, I have noticed that the unit starts off charging my battery at 500 watts (this is the solar charge controller max input for the unit) and then after a few hours of that, it only takes in about 200 watts from the panels for multiple hours. The sun conditions are the same, if not better during this time, so its not an environmental issue, and this doesn't happen on my Bluetti unit, so I assume its not a panel issue. This drop in input seems to happen at 60 or 70% charge voltage, which is quite a bit lower than the 80% battery charge that I have read can cause a drop in charge current. I know these things aren't an exact science and that readings can be off, but it's hard to use my full capacity when it takes multiple extra hours to even get to 80% or 90%. My bluetti ACP200 takes the full solar wattage all the way up to "100%" (according to the digital display). Is this an issue with settings, my own ignorance, or one of the components? Thanks!

 

[Ampster]

What is the actual voltage when it tapers off?

 

[sunshine_eggo]

What is your absorption voltage?

What is your battery voltage when you observe the output decrease?

What is the size of your array?

Is it facing South tilted at your latitude?

Does your array experience any shading whatsoever between sunrise and sunset?

 

[ggh]

@sunshine_eggo, my absorption voltage setting is at 14.6v, the array consists of 3 x 200watt panels. Total array voltage hovers in the mid 50s when the battery is taking in the full 500 watts, but I need to check what the output is when the current drops. I’ll check that tomorrow. No shading and yes it’s all tilted properly. I’ve watched the current drop and is immediate, without a cloud in the sky, no other obstructions, and with panels tilted just right.

@Ampster, I’ll do my homework and check voltage of battery next time I can observe the drop off.

As for battery voltage during low charging rate, do you want the voltage after the battery has settled for a few minutes of not charging, or the voltage while the battery is actively charging at the decreased rate? And should I give you the battery voltage reading from the BMS or from the mmp inverter? Is one more accurate than the other?

I appreciate the help

 

[sunshine_eggo]

14.6V requires that your cells be PERFECTLY balanced to 0.01V. Recommend 14.4V.

Given the above setting, it's possible that your BMS is entering OVP due to a single cell hitting the limit. If that's the case, the array output would immediately drop to what is required by loads and the unit's self-consumption.

Panel voltage not relevant.

@Ampster and I want battery voltage at the time it transitions to low output.

Also worth checking the overkill at that time to confirm it's not in protection mode.

How did the top balance go? Did you follow the guide?

 

[ggh]

@sunshine_eggo I’ll check the BMS status when it happens next time I charge the battery with solar, but I’m almost positive I checked last time and there was no over voltage error or anything like that. I top balanced according to instructions. I was able to zero out the input watts eventually and get the cells almost identical (within 2 mv i think) but it was strange how long it took to top balance. I don’t remember what the voltages of the cells were when I started the top balance, but for how long it took to charger, they would have had to start at 20% charge for the time frame to make sense. The input watts also stayed at a pretty low level for a long time during the end of the top balance.

I’ll post the battery voltage readings at the time input drops off from both the inverter and the BMS when I charge them tomorrow.

 

[sunshine_eggo]

@sunshine_eggo I’ll check the BMS status when it happens next time I charge the battery with solar, but I’m almost positive I checked last time and there was no over voltage error or anything like that. I top balanced according to instructions. I was able to zero out the input watts eventually and get the cells almost identical (within 2 mv i think) but it was strange how long it took to top balance. I don’t remember what the voltages of the cells were when I started the top balance, but for how long it took to charger, they would have had to start at 20% charge for the time frame to make sense. The input watts also stayed at a pretty low level for a long time during the end of the top balance.

Did you fabricate new higher quality leads for the power supply?
Did you charge in series with the BMS installed before paralleling the cells?

The above two things are stated in the top balancing guide.

Per hazmat regs, it is illegal to ship LFP batteries at > 30% SoC though they often are. Cells are typically received at 30-50% SoC. My nine 280Ah EVE cells were received at 28-41% SoC. I discharged them to 2.5V before charging them and then compared the as-received discharge to the total tested cap.

If you simply paralleled all cells as-received and then charged them to 3.65V with the stock leads, I wouldn't be surprised if it took a couple weeks to top balance.

If you left the cells sitting for 1-2 weeks or more before deploying the system, it's possible the top balance was partially lost.

I’ll post the battery voltage readings at the time input drops off from both the inverter and the BMS when I charge them tomorrow.

Yes please.

 

[ggh]

I connected the batteries to the inverter and BMS within a day or two if top balancing, and I did not charge in series first. I must have missed that step.

 

[ggh]

Makes sense now why they took so long to top balance. They must have arrived really low.

 

[Supervstech]

I recommend you post pictures of your build, include closeups of the bms sense wires and terminal connections.

 

[ggh]

Will do!

 

[ggh]

Alright I figured it out yesterday. The inverter/charge controller default setting for battery type is flooded acid. I thought I changed it, but I didn’t, so the controller was bulk charging the battery to a certain voltage, and then slow charging it the rest of the way. Once I changed the setting to custom battery, it started charging at the full solar wattage up to full charge. Thanks for pointing me in the right direction everyone!

Follow up question. When discharging, my BMS reads a different current draw than the inverter, and the discrepancy is far more than the stated inverter loss of 10%. For example, I hooked an appliance up that is supposed to use 160 watts. The inverter had a reading of 190 watts, and the bms had a reading of 240 watts. When I did a capacity test, it seemed like the most accurate reading was the really high one from the BMS, which is over 25% higher than the inverter reading, and 50% higher than the stated appliance power draw. Is this a horribly inefficient inverter, is something wrong, or is this normal? Thanks

 

[sunshine_eggo]

Recommend you get confirmation from yet another device, like a kill-a-watt meter.

Inverter consumes an additional ~20W that you'll never see on the inverter output. Additionally, the rated inverter efficiency is at about 30% (300W). Below that, it can get VERY inefficient. Above 30%, it's a gradual reduction in efficiency to 100%.

This chart is an example of a typical inverter: