diy solar

diy solar

DIY 48-Volt LiFePO4 Battery voltage dumb question?

OffGridForGood

Catch, make or grow everything you can.
Joined
Nov 3, 2021
Messages
3,350
Location
Canada, NW Ontario
When we buy a factory built "48-volt"server rack battery it will note right on the front 51.2 v (nominal voltage)
But what does this really mean?
Does the battery average 51.2volt output, running between 48 to 54.4 volts?
If sitting idle with no load at 51.2v is the battery approx 50% state of charge?
or is this just a convention, use 51.2volts x Amp-hours to approximate the rack capacity?
I did a search and didn't see a clear answer to this anywhere.
 
51.2 / 16 = 3.2. The nominal voltage of lifepo4 cells.
I've always correlated 51.2 to a 16s pack.
Nominal voltage meaning...at 50%SOC?
Meaning half-way between low voltage (48V) and high voltage (54.4v)
I am wondering what it is supposed to indicate?
Why is is 51.2 and not 51.4? where does this value actually derive from?
 
Nominal voltage meaning...at 50%SOC?
Meaning half-way between low voltage (48V) and high voltage (54.4v)
I am wondering what it is supposed to indicate?
Why is is 51.2 and not 51.4? where does this value actually derive from?
Waiting for experts. I have no idea.
You can't really use voltage to determine soc unless you're at the bottom or top of the curve, from what I understand.
 
Waiting for experts. I have no idea.
You can't really use voltage to determine soc unless you're at the bottom or top of the curve, from what I understand.
yes agree,
I am just curious where the "51.2 v" comes from.
as you say, it may be just how we describe the LiFePO4 cells, nominal 3.2 x 16 cells = 51.2v but I am trying to dig a bit deeper.
(I started this journey long ago with 18650 cells for eBikes, and these are described as 3.7v per cell, Li-ion cells, and can charge up to 4.2 typically)
 
My Growatt server rack battery operates between 48 - 52V. This makes me think that it only contains 15 cells. Previously when running on my home-brew 16s battery it would charge to 55v - extra 300W capacity ?
 
3.2V is the Mid Point of the Working Voltage range which is from 3.000-3.400. Every LFP Manufacturer uses the same values.
If adhering to the Allowable & Working voltage ranges, then one can consider 3.000 as 0% SOC while 3.400 is 100% SOC.

Remember, that when we test cells that are charged to 3.650 (while not allowed to settle) down to 2.500 this is a LIE as that is the Allowable Range which causes no harm but not the working range, as this includes the steep curves from 2.50-3.00 & 3.40-3.65. Actual Grade-A Capacity tests should be delivering the rated AH from that 3.00-3.40 range, not from 2.50-3.65. Also, remember that you can charge LFP to 3.65 until Amps Taken drops to 1.0A and letting that sit for 2 hours with no charge, the cells WILL settle closer to 3.45 +/-, which is perfectly natural & expected. The "KNEES" from 2.50-3.00 (Bottom) & 3.45-3.65 (Top) generally only add a few percent more AH but NOT significant, so a 280AH cell would rate at 290AH +/- "Gross Capacity" which is NOT the deliverable AH which comes from the working range. *Note that some vendors use that for trickery... we have all seen this.

Using a 48V/100AH example;
16 Cells X 3.2V nominal = 51.2V. 100AH cell capacity X 51.2V = 5120Wh or 5.12kWh
Charge (Bulk/Absorb) to 54.8 (3.425Vpc) or 55.2 (3.450Vpc) MAX as there is no gain above that only stress on cells and runner cells reaching HVD.
Float at 54.7 or 55.1
Active Balancer ON @ 3.300 *

* Active Balancer MOVES power from Hi to Lo Cells. It uses NOTHING when not initiated. It also DOES NOT WASTE energy as it moves it and there are too many misconceptions about that but it is observable & measurable (so much for wives' tales assumptions eh). Starting @ 3.30 (the last 25% of SOC) allows the cells to keep up at the higher side when cells will begin to deviate (usually above 3.350 and keep ahead of higher imbalances. This will occur on ANY Grade of cell and is relative to the amount of incoming charge, the higher the charge rate the faster & higher the deviations will occur.

There are people here that think Active Balancers need not be operating below 3.400 which to be honest is "silly" as there is no gain and you are leaving heavier lifting to the end by which time you are likely triggering HVD due to a runner cell which in the end is handicapping your pack. I have been running different active balancers for years and have done an obscene amount of tests that proved this out.

Case in point: I run 6 packs, 2 Bulk 280's, 1 A Grade 280, 1 A Grade 105, and 2 174AH (Used EV cells)
They reach float by noon and there will be deviations on the Bulk cells up to 100mv, By 1 pm with float & active balancing, all cells are under 0.010mv deviation while taking in minimal Amps while topic off & balancing. The amount of deviation varies depending on the Charge Input, the higher the input the more deviation but it all does balance out within an hour or so.

For the record, I have used QNBBM Active Balancers, Heltec Capacitive (the 2 types) and am now using the Active Balancers in the JK-BMS'...

Hope it helps, Good Luck.
 
3.2V is the Mid Point of the Working Voltage range which is from 3.000-3.400. Every LFP Manufacturer uses the same values.
If adhering to the Allowable & Working voltage ranges, then one can consider 3.000 as 0% SOC while 3.400 is 100% SOC.

Remember, that when we test cells that are charged to 3.650 (while not allowed to settle) down to 2.500 this is a LIE as that is the Allowable Range which causes no harm but not the working range, as this includes the steep curves from 2.50-3.00 & 3.40-3.65. Actual Grade-A Capacity tests should be delivering the rated AH from that 3.00-3.40 range, not from 2.50-3.65. Also, remember that you can charge LFP to 3.65 until Amps Taken drops to 1.0A and letting that sit for 2 hours with no charge, the cells WILL settle closer to 3.45 +/-, which is perfectly natural & expected. The "KNEES" from 2.50-3.00 (Bottom) & 3.45-3.65 (Top) generally only add a few percent more AH but NOT significant, so a 280AH cell would rate at 290AH +/- "Gross Capacity" which is NOT the deliverable AH which comes from the working range. *Note that some vendors use that for trickery... we have all seen this.

Using a 48V/100AH example;
16 Cells X 3.2V nominal = 51.2V. 100AH cell capacity X 51.2V = 5120Wh or 5.12kWh
Charge (Bulk/Absorb) to 54.8 (3.425Vpc) or 55.2 (3.450Vpc) MAX as there is no gain above that only stress on cells and runner cells reaching HVD.
Float at 54.7 or 55.1
Active Balancer ON @ 3.300 *

* Active Balancer MOVES power from Hi to Lo Cells. It uses NOTHING when not initiated. It also DOES NOT WASTE energy as it moves it and there are too many misconceptions about that but it is observable & measurable (so much for wives' tales assumptions eh). Starting @ 3.30 (the last 25% of SOC) allows the cells to keep up at the higher side when cells will begin to deviate (usually above 3.350 and keep ahead of higher imbalances. This will occur on ANY Grade of cell and is relative to the amount of incoming charge, the higher the charge rate the faster & higher the deviations will occur.

There are people here that think Active Balancers need not be operating below 3.400 which to be honest is "silly" as there is no gain and you are leaving heavier lifting to the end by which time you are likely triggering HVD due to a runner cell which in the end is handicapping your pack. I have been running different active balancers for years and have done an obscene amount of tests that proved this out.

Case in point: I run 6 packs, 2 Bulk 280's, 1 A Grade 280, 1 A Grade 105, and 2 174AH (Used EV cells)
They reach float by noon and there will be deviations on the Bulk cells up to 100mv, By 1 pm with float & active balancing, all cells are under 0.010mv deviation while taking in minimal Amps while topic off & balancing. The amount of deviation varies depending on the Charge Input, the higher the input the more deviation but it all does balance out within an hour or so.

For the record, I have used QNBBM Active Balancers, Heltec Capacitive (the 2 types) and am now using the Active Balancers in the JK-BMS'...

Hope it helps, Good Luck.
Thanks for taking the time Steve,
I am glad to know these details, and hear the 3.0 to 3.4 range lead to the '3.2 nominal'
Like you, I run a number of DIY rack batteries and perfer the bulk=3.45 ish, and float =3.35 settings (actually a bit higher adjustment to accound for some voltage-drop between the inverters and the batteries but you get the idea)
I have played with the active balancer settings, and monitored the effects for most of a year, but settled on 3.3 as well.
My first DIY racks had Daly BMS, but after one failed over the winter, I tried out a JK and really like how this BMS works and sets up so easily. I have ordered a couple more JK's for future DIY racks, and a spare/backup to the other Daly I still have running today.
great stuff, thanks again.
 
Nominal voltage meaning...at 50%SOC?
Meaning half-way between low voltage (48V) and high voltage (54.4v)
I am wondering what it is supposed to indicate?
Why is is 51.2 and not 51.4? where does this value actually derive from?
It's exactly the same as when you buy a AA Duracell battery and the package says they are 1.5V.
 
Thanks for taking the time Steve,
I am glad to know these details, and hear the 3.0 to 3.4 range lead to the '3.2 nominal'
Like you, I run a number of DIY rack batteries and perfer the bulk=3.45 ish, and float =3.35 settings (actually a bit higher adjustment to accound for some voltage-drop between the inverters and the batteries but you get the idea)
I have played with the active balancer settings, and monitored the effects for most of a year, but settled on 3.3 as well.
My first DIY racks had Daly BMS, but after one failed over the winter, I tried out a JK and really like how this BMS works and sets up so easily. I have ordered a couple more JK's for future DIY racks, and a spare/backup to the other Daly I still have running today.
great stuff, thanks again.
I started out with Chargery BMS' with QNBBM Active Balancers then moved onto JKBMS but after JK & Hankzor abandoned ship and left us all adrift in the sea of tech I'm waiting on a new BMS that is coming which will be similar to Seplos/Pace/TDT as far as capability, compatibility plus a few extras presently not available by those mentioned... the 1st prototypes are just going into test now, so 3-6 months before ready is my best guess.
 
I started out with Chargery BMS' with QNBBM Active Balancers then moved onto JKBMS but after JK & Hankzor abandoned ship and left us all adrift in the sea of tech I'm waiting on a new BMS that is coming which will be similar to Seplos/Pace/TDT as far as capability, compatibility plus a few extras presently not available by those mentioned... the 1st prototypes are just going into test now, so 3-6 months before ready is my best guess.
I hope to see a reveiw then! {maybe you will be able to try out the new BMS on a 16-S 540Ah cell pack!}
it seems the hole in the BMS-space is all in comms. so many options, so many work-arounds, so much to learn...love it.
 
I will review things and post here as usual. I may even lend a hand in the documentation but not sure... been down that rabbit hole only to get crapped on by some... As I am no fan of Paralleled cells I won't be testing that. I will be converting my 8S Packs to 16S so I will end up with 3x 48V/280AH and 1x 48V/174AH as I will be pulling the 24V/105AH and shifting that to the Toolshed Bank.
 
This page also answers your question very well:
 
Allowable Voltage Range varies by the chemistry and is relative to the voltage curves. LFP does have the flattest curve (at this time) but now Sodium Ion seems to be similar, so we will have to see.

Allowable Range is just that, it means that within that Voltage Window there is no harm or damage caused to the cells. This does NOT constitute the Working Range where the cells perform optimally, in fact they are NOT optimal outside of that narrow window.
The Working Range is where the cells will perform optimally & consistently. With Regards to LFP, this is where the Manufacturers do their Capacity Testing & Validation which is between 3.000Vpc-3.400Vpc. They do not test from 2.500-3.650.
 
Back
Top