Hi,

Has this been posted before? : a very lengthy article about charging LiFePO4 batteries.

https://marinehowto.com/lifepo4-batteries-on-boats/

Argues for:

•  No floating of batteries
• Long term "storage" only half charged or so.  Or cycle the battery.  But don't let it sit there at or near 100% SOC
• Don't push the cells into the "knee" trying to get the voltage of the cell up much above 3.3v.  It stresses the cell and hardly pushes any more energy in.

Now I realise that the Pylons have integrated BMS but all the reading is interesting.

What do the experts make of it?

 

5 minutes ago, Elbow said:

Don't push the cells into the "knee" trying to get the voltage of the cell up much above 3.3v.  It stresses the cell and hardly pushes any more energy in.

Oh yes, I concur with that. Especially where the control loop might be slow or AC-tied PV in an off-grid system (eg Fronius tied to output of Multi). They overvolt really quickly. I've seen a pylontech go from below 53V (charge state was still bulk) and shoot right over 53.2V and over 54V (alarm, shut down) in the space of less than 10 seconds.

1 minute ago, plonkster said:

Oh yes, I concur with that. Especially where the control loop might be slow or AC-tied PV in an off-grid system (eg Fronius tied to output of Multi). They overvolt really quickly. I've seen a pylontech go from below 53V (charge state was still bulk) and shoot right over 53.2V and over 54V (alarm, shut down) in the space of less than 10 seconds.

So 53.2 - which is what Pylon suggests - is 3.547v per cell which is quite a bit above that 3.3.  

This article shows this chart and comments:

So on that basis the 53.2 should rather be 15 * 3.45 - 51.75V,?

I think you said that you did reduce from 53.2.  What did you use?

Thanks

Very, very interesting. I have got 4 3,2 50 Ah LiFePoy batteries. Any suggestion to test it??

2 hours ago, Elbow said:

I think you said that you did reduce from 53.2.  What did you use?

When you turn on DVCC in a Victron system, and you have pylontech batteries, it automatically uses 52V. I picked that value after a discussion with someone at Pylontech who suggested at least 51.8V, which apparently is still high enough that cell balancing can happen (that is the concern... if you hold them lower than a voltage at which the balancers can work, then you must periodically take them higher).

2 hours ago, Javi Martínez said:

Very, very interesting. I have got 4 3,2 50 Ah LiFePoy batteries. Any suggestion to test it??

Hi, 

If you are patient enough you can charge to different final voltages and then discharge into a known load down to a fixed final voltage and see how much difference it makes. This guy says you get very little extra for pushing the battery that last couple of hundred milli volts which is not good for the battery. 

See Nissan Leaf and preferably not charging past 80%

 

16 hours ago, Elbow said:

Hi, 

If you are patient enough you can charge to different final voltages and then discharge into a known load down to a fixed final voltage and see how much difference it makes. This guy says you get very little extra for pushing the battery that last couple of hundred milli volts which is not good for the battery. 

See Nissan Leaf and preferably not charging past 80%

 

Well, I´ll test charging to 3.45, 3.50, 3.55 and 3.60 V. Ok?

Here's a discharge from 100% down to 23% for my pair of Pylons - as seen by the Pylons and by my inverter:

 

You can see the very rapid drop from 53.2 to 49.8 as the charger went off when I killed the grid input.

Which is consistent with the idea that there is no energy to be had.

You can see that the Pylons were still reading 99-98% SOC but the inverter went to 73%.

The battery voltage holds very flat as the Pylons report a steadily decreasing SOC.  The inverter with its lead acid programming think these batteries just don't go down.

There's that interesting drop in voltage as the batteries go through 50% SOC which I seem to see each time but don't have an explanation for.

You can see the rate of decline of the battery voltage increasing as they go under 30% SOC.  Meantime the inverter still says the batteries are at 67% or so.

I turned the power on again at about 23% SOC with the voltage at 48.3 or so.  I think by the time they get to 48V they are at about 20% and that is pretty much as low as I dared to go.

 

Another chart - what happens if you turn the charger off after "floating" the batteries for a while at the recommended 53.2V.

 

You can see that the voltage is gradually declining - looks like it will settle about 51V maybe - I'll tell you tomorrow.

The inverter still says the batteries are 100% though the Pylon's themselves say 99%.

Anyway - I'll look again tomorrow to see how it settles.  I suppose running the BMS must take some power so they should gradually go down as that power has to come from somehere.

 

 

19 hours ago, Elbow said:

Anyway - I'll look again tomorrow to see how it settles.  I suppose running the BMS must take some power so they should gradually go down as that power has to come from somehere.

 

 

Well here's how it looks today.

So with the charger off the battery voltage dropped from 53.2V to 49.9V over 20 hours or so.

The inverter is in bypass and isn't loading the battery as far as I know (my BMV isn't fitted yet).  Of course the batteries need to power the BMS inside the packs.

It looks like a massive drop if you are used to Lead Acid - and indeed the inverter says that batteries are at 86%.

But actually the batteries report that they are at 97% and still report >2.4kW of power stored.

Another interesting point - as the battery got pushed to that 53.2V the imbalance betwen the cells across both batteriies increased.

Then, as the batteries dropped back down towards that 49.9V then imbalance dropped to almost nothing.  So at the highest voltage there was about 100mV difference between highest and lowest cells in each battery.

As the battery voltage dropped to the 49.9V point the differences smoothed out and now the highest to lowest cell voltage difference is only 10mV in the one battery and 0 in the other.  - 29 cells read 3.3V and one reads 3.2V.

So the batteries seemed to balance beautifully by themselves over those 20 hours even with no charge voltage applied.

 

 

Edited January 1, 20197 yr by Elbow
clarification

5 hours ago, Elbow said:

Well here's how it looks today.

So with the charger off the battery voltage dropped from 53.2V to 49.9V over 20 hours or so.

The inverter is in bypass and isn't loading the battery as far as I know (my BMV isn't fitted yet).  Of course the batteries need to power the BMS inside the packs.

It looks like a massive drop if you are used to Lead Acid - and indeed the inverter says that batteries are at 86%.

But actually the batteries report that they are at 97% and still report >2.4kW of power stored.

Another interesting point - as the battery got pushed to that 53.2V the imbalance betwen the cells across both batteriies increased.

Then, as the batteries dropped back down towards that 49.9V then imbalance dropped to almost nothing.  So at the highest voltage there was about 100mV difference between highest and lowest cells in each battery.

As the battery voltage dropped to the 49.9V point the differences smoothed out and now the highest to lowest cell voltage difference is only 10mV in the one battery and 0 in the other.  - 29 cells read 3.3V and one reads 3.2V.

So the batteries seemed to balance beautifully by themselves over those 20 hours even with no charge voltage applied.

 

 

Reading what yoy say, makes me  think is better to do tests without balancers. ¿Do you agree?

 

Edited January 1, 20197 yr by Javi Martínez

14 hours ago, Javi Martínez said:

Reading what yoy say, makes me  think is better to do tests without balancers. ¿Do you agree?

 

In the longer term its important that the cells stay balanced.  But for testing you are watching closely so you can stop your test if they do get out of whack.

I have started the first test. It is a 3.2 X 4 50 Ah battery. It was charged to 3.4x4=13.6 V full.

 

Hi,

Here's a little test on my side - I discharged my two Pylon US2000B batteries to about 48% SOC, and then put them on charge.  My charger set to a max on 25A (12.5A per battery) and then a max voltage of 51.5v,  There are 15 cells in the battery, so 51.5 v is 3.433v per cell.

The charger was programmed to turn off once the charge current had been <1A for 30 minutes if I remember right.

It came out like so:

The battery ended up charged to about 91% SOC.  As you can see when the charger went off the voltage dropped to about 49.9v, but the state of charge still around 90%.

The batteries each reporting about 2275Wh stored - actually 100% SOC on these batteries is a bit more than 2400Wh.

At this point there is no problem with the cell balance but it might change over time of course.

Elbow

 

 

 

It has just finished first test:

Full charge 3.4 V per cell

Discharge cut : 3.125 V

Ah discharge : 42.85     Ah 85.70%

Temperature: 5ºC