diy solar

diy solar

Charging profile for eco-worthy LiFePO4 8Ah battery

Archerite

New Member
Joined
Sep 22, 2021
Messages
102
Location
Netherlands
Because of the price and small formfactor I have got a few Eco-Worthy 12V 8Ah LiFePO4 batteries from Amazon. Here is the listing on Amazon UK for the same battery since I got it in NL and all text is in dutch there. ;)

I have a Victron BlueSmart 5A charger and it charges the battery just fine, except when it reaches the "fully charged" voltage of 14.5V it's current draw drops to zero! I have tried tweaking the settings but it happens no matter what! I have put four of them in parallel and matched the voltages as much as possible before putting them together. Even let them sit a few days after charging, and then using the ones that were "too high" for something to get it closer together.
That 4 battery bank works fine...but it has the same charging issue near the end. And the problem is that the smart shunt is not showing 100% yet at that time. At least 5-10% still remaining to be charged.

To upgrade my solar system from 2 7.2Ah SLA batteries to lithium I got two more of them. Matching them after full charge and a night resting and getting the voltages close together. Now this smaller bank of two is connected to a victron MPPT 75/15 and while charging works (given there is enough sun ofcourse) it has the same issue!

There is also another thing that puzzels me about the specs of this battery. They say in series you may connect up to 4 identical ones to get 48V and I get that. The BMS might have a voltage limit. But in parallel they say you may connect an unlimited number of batteries!! Before really thinking about it I ordered another 4 of them to upgrade the two in the solar array into a 6 battery 48Ah battery bank! Yeah, in the end that was more expensive but I like how small they are and it allowed flexibility in placing them. Also...I have two 60Ah victron super cycle batteries that are intended to be hooked to the MPPT instead...and then these can go in smaller "powerbanks" or something.

The batteries each have their own build-in BMS without an option to read it's status! It just does internal balancing and a bunch of other protections like: over or under-voltage, over-current, high temperature or short-circuiting. My guess is that the BMS kicks in for over voltage and stops charging, but the victron smart shunt is showing that not all Ah have been recharged! So I am a bit confused...am I charging the batteries the wrong way or is the shunt configured incorrectly?

I have tried finding a datasheet on the battery but could not find anything useful! Only for a close relative from eco-worthy but those have yellow labels and I have the blue ones!

Hopefully the mess above makes some sense, sorry. ?
 
I have a Victron BlueSmart 5A charger and it charges the battery just fine, except when it reaches the "fully charged" voltage of 14.5V it's current draw drops to zero!
It's supposed to drop to 0A when fully charged.

And the problem is that the smart shunt is not showing 100% yet at that time. At least 5-10% still remaining to be charged.
Post the Battery settings you have for the SmartShunt. Make sure you have the correct capacity. Make sure the "Charged voltage" setting is 0.1V less than the "Charge voltage" setting in your SCC.

Now this smaller bank of two is connected to a victron MPPT 75/15 and while charging works (given there is enough sun ofcourse) it has the same issue!
Post the Battery settings you have for the SCC. Did you choose LiFePO₄ as the battery type? Make sure you do that first.
 
As far as the Victron shunt, I notice the same thing. Here are my shunt settings for my bigger 24 volt 5690 ah battery:

34FA6E63-FEDE-4A99-B84F-4FD140A5A93E.png
When I changed the Charge efficiency factor up to 100%, it got better. Not completely, but less of a factor. The documentation is a little lacking, but I believe this is where lead acid batteries are 85% efficienct so you need to charge 100 ah from the panels to get 85 ah and the missing 15 turn to heat. I read that lithium was 90% - 95% efficient so I tried that, but did not help That.

Somethign different about my batteries is they are made from cells, so I don’t have a sec for charging parameters except don’t exceed 3.65 per cell, which is actually quite bad. On my Victron SCCs I started at 27.6 and to prevent that negative balance you mention, I bumped this up by .2 amps.

Because you have manufacturer charging specs, I would not exceed 14.5. You charge at 3.625 per cell when the max is 3.65. I personally from a beginner level thing 14.5 is too high and should be around 3.45 per cell or 13.8 total. I charge mine at 3.475, 27.8 total.

I read posts about people’s BMSs and Shunts not agreeing and I did not really understand it until I switched to Lithium. I think we’ll get this dialed in eventually. For lead acid since they always had a charge going in, even on float, that negative balance would go up.
=====
Post the Battery settings you have for the SmartShunt. Make sure you have the correct capacity. Make sure the "Charged voltage" setting is 0.1V less than the "Charge voltage" setting in your SCC.
Reading this, my settings onthe picture may need to be 27.7.
=====

=======
The 0 amps after charged sounds normal to me. I think what you’re saying is you have charged the battery to 14.5 volts when solar starts in the morning and when it is fully charged amperage drops to 0.

I have found unlike my old flooded lead acid Batteries where when fully charged goes to float and has a slight amperage charge still, unlike the lithium that have little or no charge when floating. Even if the absorption voltage is still on, I’ll get 0 or next to 0. My smaller 50 ah battery pack Is what I considered fully dialed in that after its fully charged and goes to float voltage, I will have a 220 watt load on it, and the shunt shows 0 amps from and to the battery, but the SCC is producing 8.3 amps, exactly what the load needs.
 
It's supposed to drop to 0A when fully charged.
Maybe I just got used to a slower curve at the end from SLA batteries. But it's a real sudden change from 3-5A down to 0A! I am thinking it's the BMS inside the batteries but you mean that this is "normal and expected" behaviour right? If so than, great. Thanks! :)

Post the Battery settings you have for the SmartShunt. Make sure you have the correct capacity. Make sure the "Charged voltage" setting is 0.1V less than the "Charge voltage" setting in your SCC.
Ah...that might be it! I set the "charged voltage" at the same setting and not below that! Here are the settings (but I constantly keep tweaking them):
1644430344367.png1644430375799.png
I have two battery banks of the same type: Left is using 4 x 8Ah and the right is 2 x 8Ah. And like I said I keep tweaking the settings in an attempt to get it "right" ;). Only the smaller 16Ah bank is on the SCC at the moment but the issues is on both.

Post the Battery settings you have for the SCC. Did you choose LiFePO₄ as the battery type? Make sure you do that first.
I keep tweaking stuff so it's not at default anymore. But I did ofcouse started with the LiFePO4 profile and adjusted the voltages as given by the manufacturer. This one is from the SCC:

1644430567899.png

And here from the BlueSmart, also custom profile that started from the lithium:
1644430717352.png
 
When I changed the Charge efficiency factor up to 100%, it got better. Not completely, but less of a factor. The documentation is a little lacking, but I believe this is where lead acid batteries are 85% efficienct so you need to charge 100 ah from the panels to get 85 ah and the missing 15 turn to heat. I read that lithium was 90% - 95% efficient so I tried that, but did not help That.
That is an interesting idea and I could try that. I think it's also the peukert settings that needs to be something else than it is for SLA right?

Because you have manufacturer charging specs, I would not exceed 14.5. You charge at 3.625 per cell when the max is 3.65. I personally from a beginner level thing 14.5 is too high and should be around 3.45 per cell or 13.8 total. I charge mine at 3.475, 27.8 total.
I wish I could see the individual cell voltages, but I need to trust the internal BMS and only see the external total voltage.

The 0 amps after charged sounds normal to me. I think what you’re saying is you have charged the battery to 14.5 volts when solar starts in the morning and when it is fully charged amperage drops to 0.
Exactly, with the exception that this won't happen until early afternoon in my area with all the cloudy days at the moment. ;)

I have found unlike my old flooded lead acid Batteries where when fully charged goes to float and has a slight amperage charge still, unlike the lithium that have little or no charge when floating. Even if the absorption voltage is still on, I’ll get 0 or next to 0. My smaller 50 ah battery pack Is what I considered fully dialed in that after its fully charged and goes to float voltage, I will have a 220 watt load on it, and the shunt shows 0 amps from and to the battery, but the SCC is producing 8.3 amps, exactly what the load needs.
Maybe this is the issue, and I just got used to the small current an SLA battery takes in at float. maybe it's normal with lithium that this does not exist and am I looking for something that is not supposed to be there. I am still learning and this is my first time using Lithium with solar.

I also had no "continuous loads" yet and kept turning the MPPT load output on/off when I thought it was getting to low by hand. I am charging powerbanks and an ecoflow river and using those to "distribute" the solar power around the house for later use. This morning I changed that and hooked the rpi with VenusOS onto battery power. Until now that was still connected to the grid power since my SLA batteries could only run it for 18 hours ;). Right now with the sun gone for at least 4-5 hours already I still have 37 hours left! :)
 
Your SCC looks good except for maybe the max charge current. If you have 2 8Ah batteries in parallel then that's of course 16Ah. A 6A charge current is 0.375C. That may be OK but some may suggest a lower C rate for charging for longer life. But that can be a later discussion.

For your shunt, change the "Charged voltage" to 14.1V to better align with the 14.2V charge voltage of your SCC. For LiFePO₄ you want a Peukert exponent of 1.05 and a charge efficiency of 99%. The rest of the values look good.

For the battery charger the settings look fine. You don't want to regularly charge to 14.5V but doing so to each battery once in a while to get them all to the same initial voltage is just fine. The 14.2V of the SCC is what you want for daily charging.

Once you get the settings updated, connect the batteries and use them. Get them down in the 50-80% range (about 13.05V - 13.30V) and then let the SCC fully charge them. Once the SCC completes the charge the shunt should show 100%. As long as you fully charge the batteries with the SCC at least 2 times per month the shunt should be accurate.
 
But it's a real sudden change from 3-5A down to 0A! I am thinking it's the BMS inside the batteries but you mean that this is "normal and expected" behaviour right?
A sudden drop to 0A doesn't sound right. If it takes a minute then that's fine. If it goes from 3-5A to 0A in a second or two then that sounds wrong.
 
Your SCC looks good except for maybe the max charge current. If you have 2 8Ah batteries in parallel then that's of course 16Ah. A 6A charge current is 0.375C. That may be OK but some may suggest a lower C rate for charging for longer life. But that can be a later discussion.
I have read a lot about charge rate too, but these specific batteries are specified to support 10A charging and 10A discharging by the manufacturer. It even says so on the label. There are a lot of fancy numbers and charts on the amazon page and also on their own website. Except for a real datasheet unfortunately....
The 6A charge rate was more to prevent another issue with the fuses and wires actually :giggle:...I had a cheap inline glass fuse that melted away at 9A...it was rated for 10A so not sure what went wrong there. Luckily I keep watching everything and smelled something was not right...literally! :giggle:;)

I have measured the currents both while charging and discharging and it's literally split between them. So each battery should only "see" 3A at this setting. I have a clamp multimeter to check currents going through individual wires. Not sure how accurate it is but it should be enough to tell the difference between 1A and 5A I think ;)
For your shunt, change the "Charged voltage" to 14.1V to better align with the 14.2V charge voltage of your SCC. For LiFePO₄ you want a Peukert exponent of 1.05 and a charge efficiency of 99%. The rest of the values look good.

For the battery charger the settings look fine. You don't want to regularly charge to 14.5V but doing so to each battery once in a while to get them all to the same initial voltage is just fine. The 14.2V of the SCC is what you want for daily charging.
I have changed both shunts to these settings and I will also try them on the charger. The separate batteries in these banks were already synced so I was not planning on charging them individually that much. Is that recommended anyway? It was a thought I had as to how do I know when a single battery failed, you know? I am going to expand that 2 x 8ah with four more...to make it 6 x 8ah. Tomorrow is a rainy day from what the forecasts tell me here...so not a lot of sun to do a full recharge to 100% on those two I think. It's on 73% now with 1d 9h on the clock :) The only load is a rpi running venusOS using 4W continuously.

The 32Ah (4x8Ah) battery bank I don't even charge daily since I only used it for testing various high current devices. I can charge my bosch 12V batteries about 6-7 times with it...then it was at 15%. I charge those through a cheap-ish pure sine wave inverter as I could not find another option. Also have a laptop car charger and a car vacuum cleaner. And a water kettle that uses like 12A to boil a litre of watter in 35 minutes, hahaha. Not usefull...except for my purpose of having a high current load to test with. Making sure wires and fuses stay cool to the touch...that sort of things.

Once you get the settings updated, connect the batteries and use them. Get them down in the 50-80% range (about 13.05V - 13.30V) and then let the SCC fully charge them. Once the SCC completes the charge the shunt should show 100%. As long as you fully charge the batteries with the SCC at least 2 times per month the shunt should be accurate.
I'll let the SCC do it's thing for a day or two. Tomorrow is mostly rain like I said but maybe Friday or Saturday seems to be sunny. I'll let you know if these settings helped or not. I can try it on the 32Ah battery bank though with the charger or another SCC that has "simulated solar power" as it's input. ;)

A sudden drop to 0A doesn't sound right. If it takes a minute then that's fine. If it goes from 3-5A to 0A in a second or two then that sounds wrong.
That's indeed what it does. Charging at 3-5A and then in less than a few seconds it drops to 0A. The charger also clicks at that moment so at first I though it was doing that and not the battery BMS. But now that the SCC is doing it too I was thinking it might be the BMS doing an overvoltage protection or something. It does not sound good to me either.

Thanks for all the suggestions.
 
While I am preparing my new battery cabinet which will have 6 x 8Ah in parallel at 12V, I have also expanded my solar array a bit. It's now 6 x 30W outside with one of them at a different angle to catch the last few hours of usable sun rays. I have moved my 4 x 8Ah batterybox in place of the older 2 x 8Ah and have used the settings you suggested for the last few days.

Besides recording a record p max of 177W it has also been quite sunny so that helped a lot to test this out. Getting that batterybank down to 50-80% though was not that easy! My raspberry pi monitoring the SCC and shunt is using about 3-4W and takes 10% out of the battery, and it reports 3-4 days of power on average which is fine by me. Quite an improvement from the barely 18 hours when I was using two 7.2Ah SLA batteries!

As for the StateOfCharge value I have noticed it did improve a lot already but still it does not go to 100%. yet! It get's really close at 99.5% most times and only a small 0.2-0.4Ah left in the consumed value. But I did noticed that I chaged the voltage limits a little to high again...so changed them back now.

Hopefully in a few days I can say my state of charge is perfect! :)
 
A sudden drop to zero when charging is normally due to the BMS in the battery detecting over volts on a cell.
To prevent this, lower the charge voltage. The voltage of 14.5 volts is on the highside. Lower to13.8 volts as a starting point and if the battery charges OK, try slightly higher, 14.0 or14.2 volts is high enough.
Float at 13.4 or 13.35, storage should be lower by 0.05 volts.
With the settings on the Victron smart shunt/BMV, the charge voltage should be 0.1 volts less that the charger setting. Even the best SOC monitor with ideal settings will accumulate errors.

Mike
 
A sudden drop to zero when charging is normally due to the BMS in the battery detecting over volts on a cell.
To prevent this, lower the charge voltage. The voltage of 14.5 volts is on the highside. Lower to13.8 volts as a starting point and if the battery charges OK, try slightly higher, 14.0 or14.2 volts is high enough.
Float at 13.4 or 13.35, storage should be lower by 0.05 volts.
With the settings on the Victron smart shunt/BMV, the charge voltage should be 0.1 volts less that the charger setting. Even the best SOC monitor with ideal settings will accumulate errors.

Mike
Off course is won't be without errors over time. But my main issue was the sudden drop when it was near 98%-100%, which was also my expectation this was triggered by the BMS inside the batteries. I have now lowered the absorption voltage to 14.25V and float to 13.40V in the SCC.
In the shunt I have set the charged voltage to 14.2V..which I know is a small margin with the absorption voltage. But like I said, still tweaking the settings to be just a little more accurate in the consumed Ah value to be closer to 0 when it says 100% SOC.

I had discharged my batteries yesterday to about 50% (4 x 8ah bank) and it was 19.5Ah consumed. Within 2 hours of really high yields from what I am used too the batteries were charged to 100%. The curve at the end was much more smooth already than it was before. So the lower charge voltage helped a lot.

@rmaddy Thanks again for the suggestions.:)
 
Back
Top