diy solar

diy solar

Best LifePo4 charge controller settings known to man for Maximum Service life and Minimum battery stress!!! 5,000-10,000+ cycles?

Most cycles to at least80%. DoD with well over 100 cycles to 3.0VPC /0%
<snip>
Bulk charging is approx .38C (98% by alternator (rarely even turned solar on)
<snip>
Pack still delivers slightly ovr400Ah 406.2 Ah on last capacity test

400 ah * .38 c = 152 amps
400 ah * .8 dod = 320 amps

If I understand correctly the bulk phase will be ~2 hours at ~150 amps.
What is the ambient temperature range that your cells live in?
Do you have any data on the cell temperature change during charge?

Also, have you measured alternator ripple?
 
<snip>

<snip>


400 ah * .38 c = 152 amps
400 ah * .8 dod = 320 amps

If I understand correctly the bulk phase will be ~2 hours at ~150 amps.
What is the ambient temperature range that your cells live in?
Do you have any data on the cell temperature change during charge?

Also, have you measured alternator ripple?
My company actually built this alternator. Ripple measurements we do are at a 10A load. We wont ship an alt that has more than a 0.4V ripple. The load you test at number of claws,# of rectifier diodes their ratings etc.etc.will all change the measured ripple. Into an LFP bank ripple is almost nil they act as an amazing filter. My alternator uses a massive 400A rectifier and is externally rectified which allows me to run a small frame 160A stator/rotor at full bore and have alt temp sensing turned off..The tested ripple on my alt was less than .02v if I recall, under our standard test conditions.

As for bulk charging. We only replace what we need for the next day or so then the alt gets shut off..We rarely if ever charge to 100% and the only time I dose is to re-sybncronize the Coulomb counter..(I later switched to the Balmar SG200 which tracks SoC remarkably well and far better than a coulomb counter..Our bank is in PSoC the vast majority of the time..
We are in Maine and the cells have rarely ever been over80F. Charging has zero effect on cell temp on ti bank.. Winston advises .3C but through experimentation I have found that 150A+/- is just fine. in almost 12 years of this, the cells have shown no real measurable negative impacts.. My buddy Stan has sailed from San Fran to HI multiple times then SF toAlaska then AKto the Canadian Maritimes via Panama all without solar( he is an offshore racer (I believe he is the#1 record holder in sailing history)All he uses for charging is just a big Balmar alternator. Since he installed his bank back in 2010 ish my math tells me he's done close to 100,000 nm charged 100% by alternator on CalbCA cells..
 
Last edited:
We are in Maine and the cells have rarely ever been over80F. Charging has zero effect on cell temp on ti bank.
Interesting.

My cells live at ~25C ambient and the cell temperature is ~25C(~77F) except when entering the high knee when it pops up 1-1.5C.
I'm pretty sure others have observed this with the big blue prismatic cells.
It think it happens on the low knee as well but its been several months since I had them in the low knee.
I'm charging a 280ah bank with ~20amps.
 
Interesting.

My cells live at ~25C ambient and the cell temperature is ~25C(~77F) except when entering the high knee when it pops up 1-1.5C.
I'm pretty sure others have observed this with the big blue prismatic cells.
It think it happens on the low knee as well but its been several months since I had them in the low knee.
I'm charging a 280ah bank with ~20amps.
I have noted this with some of the blue aluminum prismatic cells too. Course I don't push them into the upper knee so don't see evens a 1 degree change. Keep in mind that these blue prismatic are mostly 1C rated and my Winston cells are 3C.That is likely part of the difference
 
I usually search for the main author's name and a keyword or two. Eventually a PDF will show up....

The PDF download link worked for me.
thank you! nice idea to search authors and a few keywords, will try to remember that for future ? thank you for finding the paper and linking ?
 
I usually search for the main author's name and a keyword or two. Eventually a PDF will show up....

The PDF download link worked for me.
Added to resource section

for easy reference, the graph below comes from the above paper

DD0BDBA4-3BCF-4EF2-B3DB-3D9FA71238D3.png
more fun bits..
Storage temperature has a strong impact on capacity fade rate: cells stored at 30 C experienced less than 10% capacity loss after 450 days of storage whereas, at 45 C, capacity fade extents 20% for fully charged cells (SoCnom 100%). For fully charged cells stored at 60 C, 20% capacity loss is reached in approximately 60 days only.
30C -> 10% fade in 450 days
45C -> 20%+ fade in 450 days
60C -> 20% fade in 60 days
 
Yes, one thing that ALL research I've read agrees with, warm/hot temps is NO GOOD for LFP batteries life.
60c/140f is extremely hot. I just don't see any of us with our batteries that hot. 45c/113f is pretty hot as well, but I suppose those in hot summer climates could see that if their batteries was not stored in climate controlled space..
For those in cool climates or able to store their batteries in climate controlled environment, we fall under the 30c/86f scale. In fact many of us will be able to store our batteries even cooler than that, mine are sitting at 15c right now. The study Will linked to used 25c/77f as their coolest tested, agreeing with this study that the cooler the better.

This study found higher SOC to be worst SOC for capacity fade over time. However if stored at normal to cool temp (30c or cooler), there wasn't nearly as large of a negative affect being at 100% SOC vs 30%.

It's unfortunate LFP batteries was found in this study to degrade slightly faster at 100% SOC for those that don't cycle their batteries, holding them at near full charge for emergency backup.. I plan to keep mine in the low 90% SOC range and COOL. They do lightly cycle (5% or so) daily from buffering the selling to the grid and overnight equipment draw in standby. But they hardly ever get below 90% SOC. The only way I could make that happen is do some Load shaving in the evening to draw them down, daily.
 
It's unfortunate LFP batteries was found in this study to degrade slightly faster at 100% SOC for those that don't cycle their batteries, holding them at near full charge for emergency backup..
indeed
I plan to keep mine in the low 90% SOC range and COOL.
same here. keeping cell below 25 C is enough for me.

rather than optimize for a few percent longer runtime capacity by charging to a higher voltage, if runtime makes it to next sunlight then it’ll simply begin charging again (for solar connected UPS) is the design attitude for me.
The only way I could make that happen is do some Load shaving in the evening to draw them down, daily.
neat idea.
The point that many of the posts in this thread are trying make is that there simply isn't much Ah increase between say 3.4 and 3.65 volts.
(y)
Charging the cells beyond 3.4 volts every day puts stress on the cells.
(y)

one aspect of solar charging situations is that charging might take a variable amount of time
if charging takes a while, and a higher voltage set point is used, then the batteries would likely experience more time at the higher voltage, than if 0.5C constant current/constant voltage charged.

keep cell at or below 25 C
don’t go above 3.55 V per cell

these are the two rough guidelines i try to follow
 
Two notes about that paper that might be worth investigating: One, it was written in 2013. A lot of design work has gone on with LiFePO4 technology since then. Two, they used 15 Ah cylindrical cells for the study. Perhaps something about much larger prismatic cells could invalidate some of the findings, or even make them worse.
 
I am using my 8 EVE 280ah cells with a UPS and according to my notes have not cycled them since June 25th. of last year. I have balanced them using the BMS several times because I have one cell that tends to drift down very slowly.

We have had a few power outages that did not last long and as I recall the longest one was about 30 minutes, so the cells did not get cycled much.

I am floating the cells at apx. 3.4 volts. I will try to get around to a capacity test soon. But I will have to go by the amp hours it takes to charge the cells. I have notes and will be able to tell is there has been any capacity loss and apx. how much....if any. Last time I fully discharged I had 264.31ah's go in.

My BMS is not currently accurate because since I changed some settings is has been indicating 250ah's capacity.

I tend to avoid most papers. Some are old and LFE has improved since those tests were done. I prefer to go by real usage and am interested in what people here and elsewhere have discovered. Unfortunately, there is still much data missing.
Test completed.

DISCHARGING:
Discharge C rate is very low....apx .06C. DROK meter displayed 265ah's with full battery and 11.5ah's with an empty battery. I don't fully trust it because it uses a Hall Effect sensor.

BMS displayed zero amps remaining and still discharging because it's not calibrated but will be when I do the next cycle. Delta = .390mv's at end of discharge. Voltage at end of charge = 23.20 volts with the lowest cell reading 2.60 volts.

CHARGING:
2/28/2022 Charging completed. Charge C rate is very low....apx .06C. The only thing I did different with this test was to turn off balancing.

Total Amps in =261.49ah's. Delta = .008mv's. DROK meter indicates 265ah's as before the test. I did top balance the cells to .005mv before I tested.

CONCLUSION:
Apx. 3 amps capacity lost over 8 months from holding the cells at 3.40 volts float. I am not concerned and can conclude LFE cells are great to use with a UPS compared to Gel cells especially for those who have extended power outages. I do recommend cycling every now and then which might help prevent capacity loss. :) Also please keep in mind I am relying on the BMS for measurements.
 
Test completed.

DISCHARGING:
Discharge C rate is very low....apx .06C. DROK meter displayed 265ah's with full battery and 11.5ah's with an empty battery. I don't fully trust it because it uses a Hall Effect sensor.

BMS displayed zero amps remaining and still discharging because it's not calibrated but will be when I do the next cycle. Delta = .390mv's at end of discharge. Voltage at end of charge = 23.20 volts with the lowest cell reading 2.60 volts.

CHARGING:
2/28/2022 Charging completed. Charge C rate is very low....apx .06C. The only thing I did different with this test was to turn off balancing.

Total Amps in =261.49ah's. Delta = .008mv's. DROK meter indicates 265ah's as before the test. I did top balance the cells to .005mv before I tested.

CONCLUSION:
Apx. 3 amps capacity lost over 8 months from holding the cells at 3.40 volts float. I am not concerned and can conclude LFE cells are great to use with a UPS compared to Gel cells especially for those who have extended power outages. I do recommend cycling every now and then which might help prevent capacity loss. :) Also please keep in mind I am relying on the BMS for measurements.
Hey @Gazoo - Since it has been part of the hot-topic here lately, do you know what the approximate temperature range of the cells was for these tests?
 
Test completed.

DISCHARGING:
Discharge C rate is very low....apx .06C. DROK meter displayed 265ah's with full battery and 11.5ah's with an empty battery. I don't fully trust it because it uses a Hall Effect sensor.

BMS displayed zero amps remaining and still discharging because it's not calibrated but will be when I do the next cycle. Delta = .390mv's at end of discharge. Voltage at end of charge = 23.20 volts with the lowest cell reading 2.60 volts.

CHARGING:
2/28/2022 Charging completed. Charge C rate is very low....apx .06C. The only thing I did different with this test was to turn off balancing.

Total Amps in =261.49ah's. Delta = .008mv's. DROK meter indicates 265ah's as before the test. I did top balance the cells to .005mv before I tested.

CONCLUSION:
Apx. 3 amps capacity lost over 8 months from holding the cells at 3.40 volts float. I am not concerned and can conclude LFE cells are great to use with a UPS compared to Gel cells especially for those who have extended power outages. I do recommend cycling every now and then which might help prevent capacity loss. :) Also please keep in mind I am relying on the BMS for measurements.
I'd say with uncontrolled conditions and non calibrated equipment your cells are probably with in the margin of error of being full with no or less than you think.

Certainly nothing to lay awake fretting over. Nice report!
 
Although there is absolutely nothing wrong with what @sunshine_eggo has suggested, I'll just throw in a slight change to the charge profile I've arrived at: Bulk/Absorption to 3.5V/cell (kinda between his fast and slow charging), 30-60 minute absorption time. Truth is the current drops to almost nothing before that time. Then float at 3.35V/cell. I like the SoC way of getting to a voltage on the high end and low end, but the voltages can vary significantly for charge, discharge, and resting. Makes it hard to find the perfect spot, but the goal does have merit.


I kinda wish people would not keep posting these curves and the SoC table. It is easy to trust it too much. The thing that needs to be said is that these voltages in the graphs and in the table appear to be resting voltages, i.e. some hour or more after any charging or discharging, and with nothing going into our out of the cells. If the pack is being charged at even a relatively small current (0.2C or more) the voltages will be higher, and if it is being discharged it will be lower. I appreciate @Maintenance guy is trying to help, but it is easy for someone new to this to put too much faith into these voltages, when they shouldn't.
Are the tables not helpful when initially setting Bms?
I am puzzled with a 16s 48V (60V really, as my AIO Charger/Inverter and BMS can both accommodate that setting, and it is well under the 64V that 4.0v/cell would yield). I simply have no idea what to set the 20/40/60/80 setpoints to and would love to have a starting point besides guess -which means my BMS reports garbage on the app...
Any insight or help would be welcome.
 
Quote; " I appreciate @Maintenance guy is trying to help, but it is easy for someone new to this to put too much faith into these voltages, when they shouldn't."
Sorry to disappoint ya fella but I have been hard studying Solar for over five years and have a massive data base I have collected. I study EVERYTHING posted from Will Smith to Doctor Smith, ha, ha LOL. So i have to say I am not all that new to these experiences posted here. heat and SOC or voltage at top end is the major factor in what I have collected. and yes the Lithium batteries should be topped off at least once a month to maintain amp hour rating.
The graphs I have posted are a good starting point for many entering the field of solar storage and yes I do my own tests on systems I have built for me and other people. the info I have collected off of these systems was very valuable and along with all the other info I have collected give me a good perspective on solar battery storage.
All I am doing is sharing the info I have collected and the perspective I have from the collection of info from those systems I have built in hopes that it will help out someone new to the field. If you think the info is not for you then that is fine on your part but I am sure someone here liked the info provided just not you.
Another thing is these batteries are not designed to be held at a high SOC that is why they are being used daily. massive ahr will be lost if one is held at a high SOC for months at a time and will degrade rather fast.

PS. I have Stacks of different types of lithium batteries I am doing research on within five feet from me as I write this. I am into the study for almost a year now just so you know since I am "new" into this game.
Also just like I posted all BMS's can be adjusted to start a lower balance fore which I have done with the systems I have built and those systems are running just fine staying out of the knee's of the curve.

People need to realize each system having different parameters will behave slightly different so all the info posted from me and everyone else is an excellent starting point leading to a path of long battery life with some slight tweaking of there specific system. the tweaking is of course is the information gathered from "RESEARCH" which seems to be lacking on many peoples agenda. blindly building a system with no research is not only completely stupid but can be very, very costly and could even lead to loss of limbs or life.

Regards,
Maintenance Guy
 
Last edited:
I guess if you want to baby your batteries and let them rot to death in a slow long term torture from shelf life. Then the thread gets the they look good sitting around and doing nothing award. These are batteries, live fast and die hard!! The main reason for solar is to save money.. the more use and to sap every last ounce of energy till the equipment goes dead from heavy usages I get out of my batteries is where more of my money is save , if the batteries are rated 6000 cycles, I hope I get 1000 out of it in the end because in the long term, replacing the batteries will come out cheaper from the money you saved from using solar and being less dependent from the local utility.. solar equipment, batteries ext etc will get cheaper. The utility bill rates will always goes up and never comes down!!
 
Back
Top