Alternator Size with LiFePo4

[Maine Sail]

Originally Posted by rgleason

Attached are two graphs showing Charge, Cell Voltage, SOC for LiFePo4 and a proposed charge profile by John.

Those graphs suggest to me that John is enjoying theorizing about batteries not actually using, charging, discharging or actually testing them...

[Maine Sail]

Originally Posted by tanglewood

Are this illustrative numbers, or actual test results?

Based on Johns comment at the bottom:

Originally Posted by johns61ct

Of course these are just example numbers

I suspect they are theoretical.. If not, then I would love to see John show us the data.

[Maine Sail]

Originally Posted by kmacdonald

OK, some real numbers to chew on. I'm charging a 12V deep cycle LA battery, as I type. It was put in service on 5/2016, so two seasons use. In 7 hours it has gone from 3.1 Amp to 2.9 Amp acceptance @ 14.8 volts. I'll let you know what it is in the morning. It's a 90 Ah battery so that's 0.0322C or about 3% C. I said in another thread that a lot of batteries won't get to 0.5% C (magic tail number according to some) no matter how long they charge at 14.8 volts. We'll see.

Ambient temp is 75 degrees F and the battery is 85 degrees F.

If your battery is sulfated it will take a very, very long time, in excess of 8 hours at absorption, and may never drop below 2% or 1% or .5% so you charge until the tail current stops changing, usually no tail change over an hour or two is a decent place to stop as the battery is already damaged.. How long to attain this is unpredictable due to unknown levels of sulfation.

Also, at 14.8V, the tail cut should be higher than it is at 14.4V.

When batteries are not sulfated & healthy, the tail can get down to the point where it can takes just 0.1A to maintain 14.4V on a 400+ Ah bank of L16's. At 14.8V this is higher but still will drop below 1%..



I just snapped the shot below two seconds ago. It is a G of 2015 date coded Trojan SCS-225 factory rated at 130Ah.

We are currently using this battery in part of the SOH testing of the new Balmar SG200. The SCS-225's last 20 hour test delivered 103.531 Ah or a 79.64% SOH placing it at EOL by industry standards.

Even though this battery is suffering from sulfation damage it can still drop to 0.5A at 14.8V (actual voltage Fluke NIST 14.793V) during absorption charging.

Interestingly the Balmar SG200, lower middle of image, is showing an SOH of 81% or within 1.5% of actual SOH. The Midtronics analyzer puts this battery at better than 100%...

The power supply to the right is a long term float test of LFP cylindrical cells held at 3.40V...

[kmacdonald]

I just stopped the equalization charge. It was at 10.5 amps @ 15.1 volts. Looks like the previous owner killed that one. 112 degrees F.

[tanglewood]

Originally Posted by Delfin

<Sound of crickets...>

Real or illustrative, I am probably missing something, but I am not sure that posting numbers that show a 3% differential between actual Ah stored when charge voltage is high vs. low substantiates John's statement that "Since getting up that high (note: voltage, presumably) only a small fraction of AH measured at the post is actually increasing SoC". As a side note, I seem to recall seeing this data from somewhere, but I can't recall. Maine Sail perhaps?

On the main point, am I missing something TT?

I'm just trying to decide whether to pay any attention to the figures that were posted.

[evm1024]

Originally Posted by Maine Sail

SNIP!

The power supply to the right is a long term float test of LFP cylindrical cells held at 3.40V...

Interesting - I'm looking forward to your long term float results. Any ETA for your findings?

I've got a few 18650 in the shop that I was going to try out a few tests on. Looks like I can scratch the float off the list.

Regards

[kmacdonald]

I've had brand new batteries that wouldn't get below 1 amp acceptance @ 14.8 volts. 100 Ah battery. Never tested a Trojan though. The batteries I'm talking about here were severely abused.

[phorvati]

Originally Posted by rgleason

Bingo
LM74670-Q1 from Texas Instruments
full or half bridge rectifier architectures for alternators.


www.ti.com/lit/ds/symlink/lm74670-q1.pdf

What is in your Alternator?

Last time I spec-ed out a diode set for the bridge, they were rated 1000V+. Last bridge i bought has 1600V reverse breakdown on the rectifier didoes. Precisely because didoes wont last very long when ratings are violated, the more margin the better, and various loads on the boat coupled with typical boat owner can produce surprising transients.

These guys(i know they are TI) proudly claim 45V absolute maximum rating for 2mS....And I don't see anything to protect the device. Not sure if anyone is using this for a product but they better have a pretty good customer service.

[Maine Sail]

Originally Posted by evm1024

Interesting - I'm looking forward to your long term float results. Any ETA for your findings?

I've got a few 18650 in the shop that I was going to try out a few tests on. Looks like I can scratch the float off the list.

Regards

A number of months to go... The last batch had some major variations in capacity changes between cells, all lost capacity, some quite badly, and I just wanted to repeat it to be sure, due to the wide variations. Some are elcheapo/no-name and others are premium K2, A123/LithiumWerks etc...This all takes time. This is not something I am planning to publish, just for my own curiosity, and done specifically because some cell makers claim their cells can be floated without diminishing capacity. Those claims are not what I have seen so far, and this is why I am repeating the test.

[Delfin]

Originally Posted by tanglewood

I'm just trying to decide whether to pay any attention to the figures that were posted.

Since there are genuine experts hereabouts who actually own/use/test these batteries rather than pose as one, I'm not sure why. Could be correct data, could be baloney. It is all so...enigmatic.

[evm1024]

Originally Posted by Maine Sail

A number of months to go... The last batch had some major variations in capacity changes between cells, all lost capacity, some quite badly, and I just wanted to repeat it to be sure, due to the wide variations. Some are elcheapo/no-name and others are premium K2, A123/LithiumWerks etc...This all takes time. This is not something I am planning to publish, just for my own curiosity, and done specifically because some cell makers claim their cells can be floated without diminishing capacity. Those claims are not what I have seen so far, and this is why I am repeating the test.

I hear you there. TTT - things take time. That is the reason I bought some A123 cells. Just for personal interest in the results.

Just waiting on the cell holder to arrive and then to select a test to do here.

[Delfin]

Originally Posted by Maine Sail

The power supply to the right is a long term float test of LFP cylindrical cells held at 3.40V...

Rod, do you think that there would be a difference in capacity changes between LFP batteries on float with a load vs. no load? Do you detect any heat buildup in the floated batteries?

[nebster]

Hmm, yes, interesting. I would expect a 3.40V float to pretty much "overcharge" LFP. 3.375V, not so much. What made you pick that voltage?

[rgleason]

A gentle reminder to all folks, this thread is about "Alternator Size with LiFePo4". It appears FLA and Equalization is taking over this thread in a huge thread contamination.



I am trying to get as far away from FLA charged by ICE/Alternator as possible and I am not eager to watch trailing amps creep down to .05C as the clock turns 3-5 times. I would like to forget about equalization, so there would be no need for it. So I guess those wayward posts are a good example of what we are trying to avoid by getting the right size and type of alternator with good regulation, cooling, adequate continuous hot condition output, under the expected loads, with a properly sized LiFePo4 that does not break the my bank so to speak.

Post #416 has two graphs. I should have explained them a little better.


Graph 1 (Left)

Is an overlay of the Powerstream graph How to Charge LeFePo4 Batteries -Lund Engineering linked to in Post #367


On top of this graph I overlaid the information from the SOC vs Charging graph at the bottom of Post #367 which is referred to in Post #352
and the actual page can be found here Charging LFP (LiFePO4) batteries to 90% SOC, or not. For these plots see the lower right corner of the graph.



I took that graph data and plotted it on the Powerstream graph so I could understand the charge characteristics better and determine an adequate size for the batteries, so that I could then get a reasonable size for the alternator.


The conclusion that I came to was that I should be able to use 62% of the battery by charging with the alternator without any "long tail" like FLA.



If my assumptions are whacked out or too conservative or too aggressive, please let me know.



The top left of this graph has my calcs for the size of the battery needed.
It appears for 100ah daily, with 62% use I will need 162ah of batteries. I am going to actually get a 200ah bank. Assuming this size, and using .35C I came up with a charge time of 1.5 hours for running the alternator. I could reduce that by charging at higher rates. My initial goal was 100ah in 1 hour.


As I have stated many times, I regard batteries as consumable (just as CNB states), they get damaged by users and mistreated inadvertantly. If I were to get 2000 cycles out of them I would be happy, that would be many years of use.


Should I increase the charge rate given those terms?
Should I have different assumptions for the 62% of C is available to the alternator in normal use?
I can then pick a rational size for the alternator. Hopefully.


Note: I do not have elaborate testing equipment or access to LiFePo Batteries so I am reliant on data available and experiences of others at this point.

--------------------------------------

Chart 2 (Right)

This is a theoretical charge regimen for using tapering constant current (using multiple increasing voltage change points) in an effort to reduce the time spent charging and get a good high SoC to utilize the battery capacity while not damaging it. John's suggestion is found in this thread.


This is purely theoretical and has not been tried to my knowledge. I would like to know what the %SoC is when 13.88v is reached, and how long it actually takes to charge from 20%SoC.



Low soc
.5C tapered to
.4C at 13.52v
.3C at 13.57v
.2C at 13.67v
.1C at 13.72v
.05 at 13.77v
0 at 13.8v, stop
As an example only: start off at low SoC at .5C, taper down to .4C at 13.52V, .3C at 13.57, .2C at 13.67, .1C at 13.72, .05C at 13.77, as you approach 13.8V, just stop charging. Note I just WAG'd those setpoints as an example, feedback welcome on what you think they should actually be.
http://www.cruisersforum.com/forums/...ml#post2795180



Sorry, I should have written this last night.

[tanglewood]

Originally Posted by Maine Sail

A number of months to go... The last batch had some major variations in capacity changes between cells, all lost capacity, some quite badly, and I just wanted to repeat it to be sure, due to the wide variations. Some are elcheapo/no-name and others are premium K2, A123/LithiumWerks etc...This all takes time. This is not something I am planning to publish, just for my own curiosity, and done specifically because some cell makers claim their cells can be floated without diminishing capacity. Those claims are not what I have seen so far, and this is why I am repeating the test.

Looks like you are floating at 3.4V and the cells are accepting 0.5A? (I have that same power supply ) Just curious if you are trying other voltages such that there is zero current? I think zero current is the only way to float without damage, and that means a voltage at or below the cells resting voltage at whatever standby SOC you want to maintain.


I just ran my bank through it's first charge cycle (not a full cycle, only to a touch over 3.40 Vpc), and am observing the voltage afterwards with about a .02C load. They are sitting at 3.325 Vpc which is lower than the 3.35Vpc I have seen recommended for float. I'm searching for the voltage that I'll use, and it's looking like 3.30Vpc might be a good fit.


I'd be curious how much you need to lower the voltage to get zero current. At that point electrically it's the same as being disconnected, and therefore I can't see how it would have any effect on the cell. But if you keep pumping in current, well, it has to go somewhere and so something, and it it probably won't be good once the cell is fully charged.