what happens between OVD and CLV values of a controller?

[Sunking]

1st step - connect cells in parallel over night, the lower the soc currently is, less time it takes, over night it should be much more time than realy needed. however, this will only balance them on unknown level of soc, right? the usefull point of this step is that at least we will be sure there is not a big diffference in soc of cells and it will be much more safe to hold them in planned 30-80soc. right? and also it will be much more safe to proceed to botom balance, next step...

No you are not quite understanding what is happening. You are correct when finished you will not know the exact SOC, but there is no need to know exactly what the SOC is as that is not why you are doing it. The point is to get every battery at the same voltage or SOC if you want to call it that. They will not have the exact same capacity. I assume you have 4 cells maybe 8, does not matter. Your low cell maybe 3.2 volt, and high cell might be be 3.3 volts. When you connect them all in parallel the voltage will meet in the middle like 3.25 volts. Low cells will be charged by the high cells, and high cells will be discharged into the low cells. When done will be MIDDLE BALANCED. Sensij gave you the link to David's site how to BULK BALANCE the cells. Make sure you understand what David is telling you. Very first step is connect all the cells in parallel. You walk away to allow them to equalize to the same VOLTAGE. You could careless what voltage that is. You have Winston Cells aka Thundersky batteries and it will take them a few hours to EQ because of the high resistance time constant.

2nd step - discharge them to the point, where NOW balanced cells (as described in 1st step) will become slightly different voltage. of course, with low discharge power and attended mode, as no balancer/bms/alarm are available. in fact, lover voltage will be shown, the lover should be discharging power to catch the moment some cell will fall from the hockey stick curve and start getting under lets say 3 volts. then disconnect it from the battery and similary discharge the remaining cells to the similar voltage. then, as soon as all fall out of the hocky curve on same voltage - very slowly discharge every(or maybe all together in serie if same voltage lets say 3vpc???) to the final 2.5vpc, where is the treshold 0%soc considered... then, connecting them again together in paralel, letting them auto-balanceat to absolute soc 0.01%... then start to charge them slowly and measure the ah, or kwh real capacity until we will get to the wanted 80%soc top disconnect value. then measured the resulting voltage of the serie and THIS will be reliable and secure value to be set as the charging limit for chargers of the particular one battery i have... the 30% value should be measured by the same way during the process... right?

Again does not sound like you have your noodle wrapped around what is happening. You leave the cells in parallel. When in parallel every cell has the exact same VOLTAGE no Ifs ands or buts. You leave them connected in PARALLEL. At this point you either TOP or BOTTOM BALANCE with the cells in parallel. You either discharge them to 2.5 volts or charge them up to 3.65 volts. As of now I do not think you have any equipment to implement either method. I do not recall what AH rating your cells are. But for argument sake let's say they are 100 AH. Once you connect them in parallel to Equalize voltage the SOC ends up being roughly 50% so each cell is around 50 AH. ( I am just picking a number out of the sky) With them in parallel times 4 cells is 200 AH to discharge or charge. So let's say you have a 3.65 volt 20 amp power supply it will take 200 AH/ 20 Amp = 10 + hours to charge until current tapers off to 8 amps.

To discharge is going to take a high wattage Power Resistor, For 20 amps will require 0.15 Ohm 75 watt Power Resistor. Again if the SOC is roughly 50% is going to take 10 plus hours. You want the finish rested voltage to be no greater than 2.5 volts and no lower than 2.4 volts. By rested I mean allow them to sit 1 hour.

You can speed things up a bit with your cells connected in Series. If you decide to Bottom Balance connect them to an Inverter and put a large load on the Inverter and discharge the batteries down to 3 or so volts. Them connect them in Parallel, walk away, then come back and discharge them down to 2.5 volts. For Top Balance reverse the process. Charge them up to 3.4 volts, connect them in parallel, walk away, come back and finish charging with a 3.65 volt power supply until charge current tapers to 8 amps.

pls note - the video link of jack rickard doesnt work, so the process above is just my guess of how that may be done, based on what i learned from this thread and the fine liionbms.com site you provided. so its a kind of test, if im getting into the picture finally at least a bit...

My bad, try again with this link., then scroll down to close to the bottom of the ARCHIVE PAGE to November 13, 2009.

3rd step we are there, safely cycling 30 to 80 soc, all we need to do is from time to time repeat the paralleling autobalance at lower-better soc to balancing happen in shorter time...right? anything else? is there some best practice of repeating to 0% balancing process again after some number of cycles? or anything else to do...?

If you Bottom Balance, you take the cells apart and put in parallel again, If you TOP BALANCE the BMS will keep them Balanced if you run the voltage up to 14.4 volts

Again you need to understand the difference between BOTTOM BALANCE and TOP BALANCE. With Top Balance the cell voltages are only equal at the TOP. All you know for sure is at the TOP is the cells are 100% SOC. You do not know what the capacity is. Capacity is dictated by the lowest capacity cell. With Chi-Com cells capacity is +/- 10%. So if you have 100 AH cells one maybe as low as 90 AH and one maybe 110 AH. The weakest cell dictates the PACK capacity and in this example would be 90 AH, not 100 AH.

With Bottom Balance voltage of the cells is equal at the BOTTOM, not the TOP. However at 2.5 volt you know the SOC is 0% and you know the capacity is 0 AH. When the cells are wired in series, and charged you will know the Capacity, and the capacity will be whatever the lowest cell capacity is. So if the weakest is 90 AH and you charge to 90% is 81 AH for the pack. When you charge the first time you monitor each cell voltage so you can locate the weak cell. The weak cell will be the highest voltage when charged up. When you see th efirst cell reach 3.6 volts, note pack voltage and stop charging.

[sensij]

anything else? is there some best practice of repeating to 0% ballancing process again after some number of cycles? or anything else to do...?

I would suggest that you attempt to articulate *why* you are choosing to top or bottom balance (or rejecting the other approach), with justification more substantial than "so and so told me to do it that way".

From the site that was linked:

http://liionbms.com/php/wp_parallel_balance.php

You might also want to also read this thread:

https://www.solarpaneltalk.com/forum...prismatic-bank

[racmaster]

so - finally and hopefully we are getting forward a little bit

what i should do to get there has to be:

1st step - connect cells in parallel over night, the lower the soc currently is, less time it takes, over night it should be much more time than realy needed. however, this will only balance them on unknown level of soc, right? the usefull point of this step is that at least we will be sure there is not a big diffference in soc of cells and it will be much more safe to hold them in planned 30-80soc. right? and also it will be much more safe to proceed to botom balance, next step...

2nd step - discharge them to the point, where NOW balanced cells (as described in 1st step) will become slightly different voltage. of course, with low discharge power and attended mode, as no balancer/bms/alarm are available. in fact, lover voltage will be shown, the lover should be discharging power to catch the moment some cell will fall from the hockey stick curve and start getting under lets say 3 volts. then disconnect it from the battery and similary discharge the remaining cells to the similar voltage. then, as soon as all fall out of the hocky curve on same voltage - very slowly discharge every(or maybe all together in serie if same voltage lets say 3vpc???) to the final 2.5vpc, where is the treshold 0%soc considered... then, connecting them again together in paralel, letting them auto-balanceat to absolute soc 0.01%... then start to charge them slowly and measure the ah, or kwh real capacity until we will get to the wanted 80%soc top disconnect value. then measured the resulting voltage of the serie and THIS will be reliable and secure value to be set as the charging limit for chargers of the particular one battery i have... the 30% value should be measured by the same way during the process... right?

pls note - the video link of jack rickard doesnt work, so the process above is just my guess of how that may be done, based on what i learned from this thread and the fine liionbms.com site you provided. so its a kind of test, if im getting into the picture finally at least a bit...

3rd step we are there, safely cycling 30 to 80 soc, all we need to do is from time to time repeat the paralleling autobalance at lower-better soc to balancing happen in shorter time...right? anything else? is there some best practice of repeating to 0% ballancing process again after some number of cycles? or anything else to do...?

[Sunking]

You should not let any of the cells get below 2.5V.

Wrong, dead wrong.

Do you want to believe a pretender (Karrak) or manufactures? 2.5 volts is a CYA spec some manufactures use. The real voltage is 2.0 volts.

There is a brillant Engineer by the name of Davide Andrea and is known as the God Father of BMS systems. I know him fairly well and even interviewed with several years ago when he started his company Elithion. He designed his own IC circuit he uses on his BMS systems. On his website I am going to turn you onto has everything you want to know about Lithium Ion Batteries. Every Lithium battery, BMS, and IC made are listed with selectors. There is also a tab called White Papers with a wealth of information. One in particular you may not like, Karrak will hate it. . It is Fast Discharge about 1/3 down the page that list Discharge Times for a wide variety of popular Lithium Ion batteries, Super Caps and a couple of Lead Acid batteries. Now what Karrak will hate and you may not like is Winston is second from Last with all the Chi-Coms. GBS is dead last joining all the Chi-Coms. Fast Discharge Time is a direct correlation Internal Resistance. It allows a designer to quickly size up batteries to see if they are suitable for high current applications. It is theoretical calculation of time in seconds to discharge a cell with a dead bolt fault. The longer it takes, means the higher the resistance. Does not matter what the capacity of the cell is. It is the chemistry and quality of the cell design, so it applies to any capacity. Scroll down further and Winston is dead last in Power Density expressed as w/L and w/kg. It proves Karraks claims are false.

Lastly there is this White Paper on David's site that will make Karraks blood boil. Hint 2.0 Volts. Read through the website and you will learn a lot. Or buy his book if you want to know just about everything there is to know about BMS.

[karrak]

3 - balancing topic - i understood that even if new battery, i cannot be sure it is prebalanced by factory until i take the battery on top or bottom limits or best both... at least for the first time... as i dont know what is real the internal capacity of cells and if im using them between 30-80 soc, in fact 1 cell can be cycled 20-70, another 40-90 and so on... maybe worse scenarios... right?

Do you have the Winston single12V volt battery or do you have the 4 individual 3.25V cells? If it is the battery it should be prebalanced at the factory.

4 - as i currently dont have a balancer, all i can do is reduce the interval of cycling to lets say 40-70 SOC to reduce the risk of going over the line of no return. and of course watching it by hand measurements as often as possible . until my oredered balancer will arrive.

What balancer have you ordered?

5 - isnt there really any other workaround of the process of discharging the batteries downto 0 soc 2.5vpc just to be sure they are ballanced? im afraid this discharge will also reduce their life in reasonable manner...

If you are bottom balancing do not just leave the cells in parallel with a load across them unless you monitor the voltage and disconnect the load when it gets to 2.5V. If you don't do this the load will drive the cells down to nearly 0V which will damage them. You should not let any of the cells get below 2.5V.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[karrak]

At some point, the internal construction is going to influence the total life of the battery, and as the internals age, the calendar is kill the battery, regardless of cycles. Finding THAT point should be easy, if all the battery mfg's had realistic charts for their battery.

I agree, as with all batteries it is a case of use it or loose it. I am not so sure that it is easy to work out the balance between usage and lifespan as there are so many factors that influence this.

I am now very confident that with my usage I am going to get at least ten years use out of my battery and think it could be as high as twenty years.

For LFP batteries I think it is better to think in terms of amount of energy cycled through the battery during its life rather than the number of cycles. So you might get 2500 cycles at 100% discharge but 5000 cycles at 50% discharge. The amount of energy cycled through the battery is the same in both cases.

Because the batteries have a fixed lifespan regardless how much you use them you want to cycle as much energy through the battery as possible during its lifetime to get the best value for money.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[Mike90250]

karrak - what i want to reach is higher cycle life. i dont need the battery to be used for higher capacity,........ cyclelife

At some point, the internal construction is going to influence the total life of the battery, and as the internals age, the calendar is kill the battery, regardless of cycles. Finding THAT point should be easy, if all the battery mfg's had realistic charts for their battery.

[Sunking]

Lots of questions.

1 - the dammed ac smart leadacid charger question - if i understood your point about chargers, they in fact have not so much phases, just a - "charging" or b - "stand by and get ready for load to suppor by power". lets forget the equalizing by overvoltage for this time... this promises a opportunity for me to use this old leadacid charger for lifepo, of course in attended mode and not let it charge higher then mentioned safer 70soc, which could be about 13.5 or 13.6. then turn it off by hand. this could result in pushing some 10,20ah into battery without any serious risk. isnt it? or is it not so easy and other threats are there?

When you program a 3-stage charger, all you are doing is programming Voltage Set Points. Example Bulk/Absorb you set for 14.4 volts or whatever you want. Additionally with Solar Charge Controllers you set a Time Period for Absorb phase typically 2, 4, or 6 hours. So when you start charging a battery with an OCV of say 12 volts, the controller pumps in all the current the panels can possible deliver for the given conditions. This is the Constant Current of Bulk. With Solar it is really Constant Power, what ever power the panels can generate. . This continues until the voltage rises to whatever set point you programmed. When Set Point Voltage is reached begins the Absorb Phase and the Timer Starts. After the Absorb Timers times out, the Absorb phase ends, and the Controller now lowers the voltage to say 13.6 volts. The loads will bleed of the batteries or drain them to 13.6 volts. At that point if there is any Sun left the panels will supply power.

Did you catch two points? Absorb is a timed event, and Bleeds off power? That is just dandy for lead acid, but not lithium. That would be a serious over charge on lithium. For Lithium batteries, Absorb is NOT A TIMED EVENT. It is a Current Event. You hold 14.4 volts until the charge Current Taper down to 3 to 5% of C where C = the battery AH capacity. Example a 100 AH battery would be 3 to 5 amps. Absorb time with a Lithium battery is measured in minutes not hours. I am not familiar with your controller but I would bet money you cannot set Absorb to stop on a preset current. I bet you anything it is a TIMER right? I also bet your Pb battery charger is a timed event. This is the difference between a charger made for Pb and one for Lithium. Now some of the newer more expensive AC battery chargers can set Absorb to be either Time or Current. Those can be used for either battery. You can use you rPb charger by either lowering the voltage or manually monitor it.

Second point here is Bleed-Off power when the voltage is lowered for Float. That means you went from over charged, bleed off some of the energy back down to safe levels. So ask yourself why would you want to over charge the battery causing damage, then bleed off the over charge down to safe limits. I would not do that. Would you? Why go there in the first place.

2 - i counted several scenarios i have and had here with your internal resistance formula in vice versa ways. seems like it perfectly sits and answeres some of my long time opened questions...
just to be sure, today the sun was there even if not predicted, co i was able to test those limit values in live... what happened is, that the battery reached my 13.85 charging limit set and then the controller switched to float, voltage went down, current also, no charging, standby, waiting for load. after load happened battery was untouched, all power came from panel. everything works as designed...

the problem occured was: voltage i measured on battery was -0.2V in comparison with battery voltage measured by controller. so the value 13.85 which stopped the charging was in fact only about 13.65. charging current was cca 10a, so based on this formula it gives exactly 0.2v difference if i pressume my new battery has ri 0.02. is this this case or its a coincidence and problem of inaccurate voltmeters?

1 of 2 things is wrong. Either your controller metering is off, or your DVM. 95% chance it is the controller. Ohms Law has 3 equations for Voltage. In this case Voltage = Current x Resistance. Wire and connectors have resistance. How much depends on the length or the wire and the size of the wire. Example 12 AWG stranded wire has 1.24 Ohms per 1000 feet. So if you had say 10 feet loop (5-feet one way) would be .0124 Ohms. So if you had say 10 amps of current flowing the wire would drop 10 amps x .0124 Ohms = .124 volts. So if you measured 13 volts at the controller, you would see 12.876 volts at the battery. So what would the voltage at the battery with 0 Amps flowing assuming the controller voltage was 13 volts? This is a Trick Question I always ask students. Same question as: Who is buried in Grant's Tomb?

0 x Anything = 0. No current = no voltage loss. You had better measure 13 volts at both the controller and battery with no current flowing. You would be surprised how many people cannot figure out who is buried in Grant's Tomb. Send that student home with a Dunce Hat and Stupid Shirt.

then i redefined charging limit to 13.95 and charging went back, but clouds occured, so not fully back. however, charger status was shown boost all the time before and after, what i do not understand, as it should be bulk based on my openion and limit values. however, based on point 1 it should be just a marketing name for the same, as in fact battery was still charged at full amperes provided by controller/sun and the voltage was growing slowly... so maybe only interface error, or misunderstanding of the process by software author...

3 - balancing topic - i understood that even if new battery, i cannot be sure it is prebalanced by factory until i take the battery on top or bottom limits or best both... at least for the first time... as i dont know what is real the internal capacity of cells and if im using them between 30-80 soc, in fact 1 cell can be cycled 20-70, another 40-90 and so on... maybe worse scenarios... right?

4 - as i currently dont have a balancer, all i can do is reduce the interval of cycling to lets say 40-70 SOC to reduce the risk of going over the line of no return. and of course watching it by hand measurements as often as possible . until my oredered balancer will arrive.

OK STOP RIGHT NOW, you are playing with fire. Your batteries are not balanced, and there is no such thing as a Balancer. All a BMS or Balancer can do is keep a Balanced pack Balanced. They are not capable of Balancing a set of new batteries. The Initial aka Bulk Balance is done by connecting all cells in parallel. Walk away for a few hours, and then either Top or Bottom Balance. That takes a Power Supply with a precision Voltage Regulator to Top Balance, or a load to discharge the cells to 2.5 volts for Bottom Balance.

If anything STOP what you are doing and MIDDLE BALANCE the cells by connecting them all in parallel and let them set over night. Then you can connect them back up in series and play. Having said that LiFeP04 cells do not do well Middle Balance. Any of the higher voltage types can be and is exactly what EV's use. No commercial EV manufacture would ever Top Balance. Otherwise they could not offer any warranty and would go bankrupt with warranty claims.

5 - isnt there really any other workaround of the process of discharging the batteries downto 0 soc 2.5vpc just to be sure they are ballanced? im afraid this discharge will also reduce their life in reasonable manner...

No more risk than Top Balance. In fact less risk. First point is 2.5 volts is not the danger zone, it is 2.0 volts. Manufactures say 2.5 volt to cover their arse. Second point is over discharged cells get damaged by the adjacent cells with energy left in them. Example say in a 4S configuration one cell reaches 2.5 volts, while the other 3 cells have 3 or more volts and some charge left in them. They will drive the discharged cell into reverse polarity, and destroy it instantly. If you put all cells in parallel, walk away over night, it is impossible for that to happen. All cells are at the exact same voltage.

Do your self a favor, get some popcorn and a few cold beers and watch this 1-hour video on Bottom Balance by Jack Rickard. Jack is the premier custom EV builder in the USA. He is a retired Electrical Engineer with a passion for EV's and started his own biz.

[Sunking]

There is not a problem with my battery, there is a problem with Sunking's understanding of battery chemistry.

There you go again making a fool out of yourself. You cannot even read your own graphs. Your graph is telling you exactly what your Ri is and you do not know how to read it.

[racmaster]

screnshot and graph. take a look on that charging state boost bull****..?? this is there all the time since the beggining, no matter what happens... however, the other things are more important, not the status shown i think...

2hrsofsun.jpgcahrging day1.jpg

[racmaster]

karrak - what i want to reach is higher cycle life. i dont need the battery to be used for higher capacity, as i dont need so much power. therefor im focusing for control parameters to be set so that this goal will be reached... cyclelife

[racmaster]

ok sunking, brainstorming session is helping me bit by bit, here is how your info landed in my brain, pls check, if thats right...

1 - the dammed ac smart leadacid charger question - if i understood your point about chargers, they in fact have not so much phases, just a - "charging" or b - "stand by and get ready for load to suppor by power". lets forget the equalizing by overvoltage for this time... this promises a opportunity for me to use this old leadacid charger for lifepo, of course in attended mode and not let it charge higher then mentioned safer 70soc, which could be about 13.5 or 13.6. then turn it off by hand. this could result in pushing some 10,20ah into battery without any serious risk. isnt it? or is it not so easy and other threats are there? which i didnt realize yet...

2 - i counted several scenarios i have and had here with your internal resistance formula in vice versa ways. seems like it perfectly sits and answeres some of my long time opened questions...
just to be sure, today the sun was there even if not predicted, co i was able to test those limit values in live... what happened is, that the battery reached my 13.85 charging limit set and then the controller switched to float, voltage went down, current also, no charging, standby, waiting for load. after load happened battery was untouched, all power came from panel. everything works as designed...

the problem occured was: voltage i measured on battery was -0.2V in comparison with battery voltage measured by controller. so the value 13.85 which stopped the charging was in fact only about 13.65. charging current was cca 10a, so based on this formula it gives exactly 0.2v difference if i pressume my new battery has ri 0.02. is this this case or its a coincidence and problem of inaccurate voltmeters?

after the charging was stopped at controller measured 13.85(13.65 by hand multimeter), battery sit back on 13.4 in few minutes. with small 2a load went to 13.3 under load.

unfortunately, i dont have exact logs, just the graph attached and my meantime hand measurements...

then i redefined charging limit to 13.95 and charging went back, but clouds occured, so not fully back. however, charger status was shown boost all the time before and after, what i do not understand, as it should be bulk based on my openion and limit values. however, based on point 1 it should be just a marketing name for the same, as in fact battery was still charged at full amperes provided by controller/sun and the voltage was growing slowly... so maybe only interface error, or misunderstanding of the process by software author...

3 - balancing topic - i understood that even if new battery, i cannot be sure it is prebalanced by factory until i take the battery on top or bottom limits or best both... at least for the first time... as i dont know what is real the internal capacity of cells and if im using them between 30-80 soc, in fact 1 cell can be cycled 20-70, another 40-90 and so on... maybe worse scenarios... right?

4 - as i currently dont have a balancer, all i can do is reduce the interval of cycling to lets say 40-70 SOC to reduce the risk of going over the line of no return. and of course watching it by hand measurements as often as possible . until my oredered balancer will arrive.

5 - isnt there really any other workaround of the process of discharging the batteries downto 0 soc 2.5vpc just to be sure they are ballanced? im afraid this discharge will also reduce their life in reasonable manner...

[karrak]

so to get back to conclusion, my limit voltages could be fine for my purpose 30-80soc? no matter of what charging current will flow from panel? (0-25a)

I think it is impractical to keep your battery in the 30%-80%SOC range by using voltage as there is so little voltage difference between an SOC fo 30% and 80% and the fact that the battery voltage is dependent on the current going into or out of the battery which is variable.

If you wanted to keep in the 30%-80%SOC range I think you would have to use an SOC meter to control your solar charge controller and provide the LVD (low voltage disconnect) for your inverter. I think an SOC meter is a very handy thing to have with LFP batteries as they are very accurate because of the very high charge efficiency of LFP batteries of greater than 99%. Due to inaccuracies in measuring the current and charge efficiency of the battery the SOC meter has to be reset on a regular basis by charging the battery to around 99%-100%.

BTW my battery is just fine, when it was new a 200A (~0.55C) load on it dropped the overall battery voltage by ~1.00V which includes the wiring losses , this equates to an overall battery resistance of ~5mOhms (0.005Ohms). Today the same 200A still produces a ~1.00V volt drop. There is not a problem with my battery, there is a problem with Sunking's understanding of battery chemistry.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[Sunking]

When I first commissioned my system I used the accepted wisdom that you shouldn't charge them up to 100% as it would dramatically lessen their lifespan. I tried to charge the battery to around 90% and not float charge it but found this impractical as it required a charge voltage of around 3.375V which meant that on sunny days the charge controller would go into the Absorb/CV phase and start limiting the charge current at less than 70%. This meant that it took for ever to complete the charge and I would run out of sun.

After further research and seeing what other people are doing I opted to charge at 3.45V/cell (27.6V for my 24V battery) and terminate the charge when the voltage had reached 27.6V and the charge current had dropped below C/50 and then drop to a float voltage of 3.35v. Contrary to what Sunking is saying there are a number of solar controllers that will terminate the charge when the voltage reaches the CV voltage and the current drops below a certain value, you can also approximate this by setting the absorb time, around 1/4-1/2 an hour is a good time for LFP batteries. This charge regime has worked very well for not only me by numerous other people that are using it and if there is enough sun results in a battery that is around 98% full at the end of the day.

Complete BS and make believe science.

Makes no fricking difference what battery or voltage you use. If there is no sun on cloudy days, there is no fricking power except in never never land of Karrak. Please explain to the world how you get power where there is none to be had. What laws of physics change in your make believe world? What idiot would turn off his solar system if the battery is charged except you.

The rest of of want our panels to supply power when the sun is shinning and save our batteries until after dark and for the next cloudy day if that happens. Your way starts you off behind the 8-ball. Any off grid solar system must have a generator for those cloudy spells or you go dark and wait for a few bright sunny days to recharge.

Karrak your logic is completely moronic and lacks even the most basic common sense.

[karrak]

battery is used on my cottage as offgrid, to power some led lights, notebook, small 5amp water pumps and so on... nothing big. charging current about 25a max, but currently its snowy weather, so up to 1 amp )

This being the case I assume like us that you want to store as much power in the battery as possible to carry you through the cloudy days?

When I first commissioned my system I used the accepted wisdom that you shouldn't charge them up to 100% as it would dramatically lessen their lifespan. I tried to charge the battery to around 90% and not float charge it but found this impractical as it required a charge voltage of around 3.375V which meant that on sunny days the charge controller would go into the Absorb/CV phase and start limiting the charge current at less than 70%SOC. This meant that it took for ever to complete the charge and I would run out of sun.

After further research and seeing what other people are doing I opted to charge at 3.45V/cell (27.6V for my 24V battery) and terminate the charge when the voltage had reached 27.6V and the charge current had dropped below C/50 and then drop to a float voltage of 3.35v. Contrary to what Sunking is saying there are a number of solar controllers that will terminate the charge when the voltage reaches the CV voltage and the current drops below a certain value, you can also approximate this by setting the absorb time, around 1/4-1/2 an hour is a good time for LFP batteries. This charge regime has worked very well for not only me by numerous other people that are using it and if there is enough sun results in a battery that is around 98% full at the end of the day.

Because this scheme takes the battery to around 99% full you have to make sure that the battery is balanced at the top end. Unless all your cells are the same capacity you cannot have them balanced at the top and bottom end. For me it is more important to have the battery balanced at the top end because that is where my battery spends most of its time. As I said I hardly ever go below 30%SOC. If you have something that is monitoring the individual cell voltages all the time and will turn off the load and/or give an audible alarm if any of the cell voltages go below a safe level there is no way that you can damage your battery by over discharging it. You can do this with something as simple as a Cellog 8 which will only monitor the individual cell voltage, does not have any balancing function and cannot make your battery go out of balance, even if it malfunctions.

btw, next week no sun predicted, so i plan to charge it with classic leadacid 20a "smart" charger as there is nothing better. what i suppose is the charger recognizes it as almost fully charged, so maybe bulk part will be skipped and some absorption or float will occure. i plan to sit there and monitor it by hand to prevent overcharging, hoping to get some 10-20% higher soc and then turning it off. have no better idea currently and some power would be usefull to have. charging will go from petrol generator for about a hour or so. any hint?

The smart charger will most likely try to charge your battery to ~14.5V and then float charge it at ~13.8V which is only OK if your battery is top balanced. If you had a cellog 8 you would get an audible alarm if any of the cells goes too high. If you are checking manually you have to be really diligent, the voltage can go too high in a matter of minutes.

was also thinking to use the old lead acid charger as an input to the solar controller mppt 30a, which could be theoreticaly nice as it could take care for the charging process without my fulltime attention, but as im not sure what that could make, im not going to risk that. would be nice if that would be working, but things will not be probably so easy... do you have any experience or ideas regarding charging with ac leadacid charger/solar controler and lifepo4?

Not sure that the solar controller will work with this. What will work is of you buy one of the RC chargers and use it to charge the LFP battery. I have a Reaktor 300W which is not expensive and works very well.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller