diy solar

diy solar

Cinergi's 28 kWh / 4 kW Solar / 10 kW inverter RV build

I posted a thread, i've let my cells 2 years without touching them at 19°C, they lost under 0.01V in this period .... 3.29V to 3.28V. So .. self discharge is pretty low.

Can you charge them up to 3.55V and report the spread between cells?

I do not know what "top ups to the knee" mean .. :D

Meaning that.
 
.01v in that voltage range can be upwards of 10-20%. There's no real gauge of SOC in that voltage range.
I agree
Can you charge them up to 3.55V and report the spread between cells?



Meaning that.
I only charge to 3.4V => 54.4V => Cells are virtually at the same voltage from each others (depending on how i parameter the BMS), i balance at 0.10V, starting at 3.25V. Tipically my cells are withing 0.02V-0.07V (all the time between 10%-95% SOC)

I do not see any advantage going higher, reason i find that top balancing is useless.... i do not have to scrap each Ah of the pack, i let them work in their comfort zone.

I use the pack in the comfort zone of the weakest cell. I charge the pack and check cells voltage... if one start to have problems to follow the others... i let the pack settle, i check the weak cell settled voltage => This will be my reference. In my case, my reference is 3.45V.
My cell 11 is the weakest and 15 the strongest, cell 11 start to "separate" from the others at 3.45V (delta with other cells increase starting at this voltage), i use this voltage x 16 to charge the pack (3.4*16 = 54.4, i take a voltage a little under the weak settled one), this way i know this cell will not be overfed.. cause i feel that increasing voltage to overfeed those cells is what make them ages faster, it make sense in my mind...
I take again the exemple i already used, in a rowing boat there are 8 guys, if you push the team to the max of the weakest guy... it will be ok, if you push the whole team to the max of the stronger one .. it will fail pretty fast... even if between those 2 "performances" there is not a lot differences. This 5% more will kill the weakest guy in no time. Whoever as done sport in competition will understand that, at high level .. those 1%, 2% improvment are insanely hard.
I apply the same rules to those cells, the weakest is the boss... :p
 
I only charge to 3.4V => 54.4V => Cells are virtually at the same voltage from each others (depending on how i parameter the BMS), i balance at 0.10V, starting at 3.25V. Tipically my cells are withing 0.02V-0.07V (all the time between 10%-95% SOC)

I do not see any advantage going higher, reason i find that top balancing is useless.... i do not have to scrap each Ah of the pack, i let them work in their comfort zone.

We would like to operate the cells in about that SoC range you mention, for longest life.

The reason to go higher would be

1) To see if the cells are actually balanced (we don't believe 10's of mV difference means close SoC at the lower voltage)

2) To get the BMS to balance the cells.

Is it possible one cell is at 95% SoC and another is at 70% SoC, showing the voltages you have? (I haven't worked with the cells, just thinking about flat middle of the curve.)
If so, that would mean reduced capacity available if discharged to low voltage disconnect.

If you were to charge well into the upper knee one time, you would find out what SoC is.
Alternatively, discharge to lower knee.
 
We would like to operate the cells in about that SoC range you mention, for longest life.

The reason to go higher would be

1) To see if the cells are actually balanced (we don't believe 10's of mV difference means close SoC at the lower voltage)

2) To get the BMS to balance the cells.

Is it possible one cell is at 95% SoC and another is at 70% SoC, showing the voltages you have? (I haven't worked with the cells, just thinking about flat middle of the curve.)
If so, that would mean reduced capacity available if discharged to low voltage disconnect.

If you were to charge well into the upper knee one time, you would find out what SoC is.
Alternatively, discharge to lower knee.
1) In fact my pack is .. kind of top balanced, cause the BMS is acting like if cells were connected in parallel.
Over 3.25V (parameter), if there is more then 0.01V delta between highest/lowest cell (parameter) then balancing occur, that the pack is charging... idle or discharging... balancing is working.
While balancing, pack voltage will decrease, then charge will start again ... and repeat.... till the pack is at 54.2V (or 54.4V in my case) AND each cells are withint 0.01V of each others. The pack will then be fully charged FOR this pack total voltage of 54.2V and perfectly (0.01V) balanced.
=> Top balancing do not seems mandatory to me, as the bms will in the end .. do the job, maybe slower.. but i do not need to buy a specific charger. And anyway i'm off grid .. so .. i could not use a charger.

2) The bms balance cells at the set voltage (3.25V in my case, thinking putting it at 3.3V), no need to go high voltage for this to happend.

At 3.4V (settled) it's unlikely that there is 20% difference SOC between cells, 3.4V is in the "vertical" part of the graph => nearly same SOC
Under 3.3V we are on the verge of the horizontal part of the graph => incertainty of the SOC (as said Dan, could be as much as 20% for the same voltage or maybe worth in some cases).

2021060339683333.jpg
 
Do you charge to 3.4V per cell? Or charge higher and let it settle to there?
From the following curves I would think charging to 3.4V gets it to 70% SoC, and would then settle to 3.25V

Looks to me like one should charge to 3.4V ... 3.55V depending on charge rate (0.05C ... 0.2C) to achieve 92% to 95% SoC.
A bit lower to hit 90%, and only reliable way to do that would be 0.2C until 3.475V (measured at the battery)
Hitting 95% should be easier steeper on the curve, so doable from 0.1C to 0.2C. These curves top out at 3.6V and 0.2C hitting that at 95% SoC.

1669390775861.png
 
Do you charge to 3.4V per cell? Or charge higher and let it settle to there?
From the following curves I would think charging to 3.4V gets it to 70% SoC, and would then settle to 3.25V

Looks to me like one should charge to 3.4V ... 3.55V depending on charge rate (0.05C ... 0.2C) to achieve 92% to 95% SoC.
A bit lower to hit 90%, and only reliable way to do that would be 0.2C until 3.475V (measured at the battery)
Hitting 95% should be easier steeper on the curve, so doable from 0.1C to 0.2C. These curves top out at 3.6V and 0.2C hitting that at 95% SoC.

View attachment 121763

Graph is for 4 years old 90Ah winston, each brand as a different graph.

See for example this one at 25°C, 0.1C/0.33C (not easy to see, blue and red line).
charge-and-discharge.jpg

I charge at 54.4V, the idea is that a particualr cell at 3.3V could be at 50 or 70% SOC, right ? It's not true if you let it charge at 3.3V enought time. And that's what i do.
Meaning at 3.3V, after a while (under charge), all cells are at their higher SOC for 3.3V .. if it make sense... so none will be at 50% SOC, all at 70%SOC (for 3.3V.. )*

Edited : Sorry i did not replied, i charge at 54.4V, and "float" to 54.0V, it charge fast till cells are at average 3.35V then it slow down, i suppose cause the delta between applied voltage and pack voltage narrow. Then my charger change to float after what i supposed being a low charging amperage for a certain amount of time, so we can say it settle at 54V. Should be something like 95% SOC.
 
Last edited:
If you operate within the middle 50% SOC, then it can take years for imbalance to appear. Most of us don't want to spend double on a pack, so we use 90% or so of the total capacity when needed.

Which is why it's a bit disingenuous/ confusing to make some broad claims when operating a system with extremely conservative, or unusual usage parameters.
 
Last edited:
If you operate within the middle 50% SOC, then it can take years for imbalance to appear. Most of us don't want to spend double on a pack, so we use 90% or so of the total capacity when needed.

Which is why it's a bit disingenuous/ confusing to make some broad claims when operating a system with extremely conservative, or unusual usage parameters.
Cause you charge your pack to 58.5V ?

My cells rest at 54V, which is not "highly conservative", this is just that i do not ram amp in them with a high voltage.
Saying that a pack that is resting at 54V is an "extremely conservative" way of playing with 48V LiFePo4 pack is a bit disingenuous.. isn't it ? isn't that a broad claim ?
Or maybe you do not even got a 48V LiFePo4 pack and never experienced with one ?

Just look at Steve post to understand : https://diysolarforum.com/threads/l...fferent-brands-of-batteries.34175/post-430831
 
Last edited:
@cinergi
IIRC, your battery cells are setting on Series 10, 1x2 aluminum extrusion without anything under the cells. Have you noticed any wear on the cell bottoms during any of your battery inspections?
 
@cinergi
IIRC, your battery cells are setting on Series 10, 1x2 aluminum extrusion without anything under the cells. Have you noticed any wear on the cell bottoms during any of your battery inspections?

I haven't checked as the only way to properly inspect would be to disassemble and lift a cell up so I can see the part that's on the resting on the aluminum. That said, the plastic piece on the bottoms of the cell didn't give me cause for concern. Maybe I'll try to do a visual inspection in its current form to see if I notice anything, especially with the cells on the end.
 
ahhh.. flat panels on an RV in the north. In the summer I'd make 22+ kWh/day. I'm lucky to make 5.5 now :LOL:
 
I wish there was a decent auto tilting mechanism out there (for RVs). Haven't found one yet at least. Even 10 degrees would make a big difference!
 
Back
Top