Lithium Batteries (for the rest of us)

[ebaugh]

Originally Posted by roetter

I am now assembling the cells to a 2P4S battery matching the weakest with the strongest cell, second-weakest with second-strongest, and so on. I am hoping to get a well balanced pack doing it this way.

I wish I could have done that with mine, but with 48 cells, it was too much work. I don't think it will impact "balance" as much as it will maximize capacity.

[T1 Terry]

Originally Posted by hellosailor

Terry-

Originally Posted by hellosailor

First you say that charging over 3.4 volts causes damage, then you say to charge to 3.6 volts. And Balqon say their "12" volt batteries have an operating range from 11-16 volts, which is a full 4 volts per cell. In normal operation, no mention of damage.

Maybe you can explain those apparent contradictions to me? Maybe "16" was just a typo for "13.6" and two characters got left out of the middle, huh?

"This would indicate the batteries will cycle for at least 10 yrs " {The Mars rovers]
Well, that would indicate that NASA had a working BMS ten years ago, along with a good battery chemistry. Perhaps someone could find out what they did, so we could just follow that example and ignore all the rubbish being passed around today.

There is a lot of understanding required about how these cells react when fully charged before random figures mean anything really.
If you constant current charge at a 1C to 3C rate, stop charging at 4v per cell, if you charge at a constant voltage down till the last little bit can be squeezed in, stop charging at 3.45v. If the charge rate is a few amps and a cell climbs to 3.6v the internal heating is 0.2v over charge x 2 amps = 4w, the time taken to trickle charge from 3.45v full to 3.6v would be in mins, 4w applied for a few mins = stuff all cell internal heating, if you stop charging for a few mins, that heat can dissipate and any load will burn away the slight over voltage so no harm is done. If you hold the cell at 3.6v then it's 4w x how ever long you hold it there, given enough time you will damage the cell. Charge the cell to 4v and hold it there = 0.6v at what ever amps for how ever long = one stuffed cell.
Your call, simply stating charging voltages without any reference to charging rate is meaningless with these batteries, just snatching little bits of information without understanding and floating over the bits you don’t understand will result in a lot of pain to the wallet and then the battery chemistry getting the blame, not the operators lack of understanding or care to do it right in the first place.

Seriously, if people are not going to take the time to understand just what it is they are doing, please stay with lead acid batteries, Li batteries don't need a bad rap due to poor practices.

T1 Terry

[chala]

Originally Posted by goboatingnow

never seen an standard alt, with a float cycle , interesting
dave

If you call a marine alternator a standard alt then I have a Valeo 24v that float, it is about 25 years old.

Originally Posted by Maine Sail

She can pump out 100A at engine idle....

Why idle? Idle speed as never been the best for diesel or belts.

[roetter]

Originally Posted by ebaugh

I wish I could have done that with mine, but with 48 cells, it was too much work. I don't think it will impact "balance" as much as it will maximize capacity.

You are right, all I get is a tiny bit more in potential capacity.

In the middle of SOC (10-90%) it makes no difference if one cell at a SOC 5% lower or higher than another one. The charge and discharge curves are almost horizontal. It is only at the very end that it shows.

Looking at my graph showing the end of the discharge cycle, the cells start spreading at time index 2:30 (this by the way is 24 hours plus 2:30 since I started the logging). To the left of it it is mostly inaccuracy in the celllog and noise that we sees.

I terminated at 3:45 (+1:15). At that point the lamps were drawing about 6A. So it was the last 7.5Ah that they started spreading out in a 185Ah set, or 4%. So if I define that when I stopped at 2.95V the cells were empty, (probably another 2-3 Ah down to 2.8V) then the cells stay completely aligned down to 4%SOC.

The same is probably true for the upper end. I wish I had another set to test that. If so, that would mean that there is no significant imbalance of cells measured as a voltage difference unless any of the cells gets below 5% SOC or over 95% SOC. So if you limit yourself to SOC 10%-90% in terms of voltage cutoff, than it would take any cell a 5% drift in any direction to show up in your logging. A good case to say - BMS with automatic balancing - not required.

Anyway, I will keep looking at those logs and should be able to pick out any cell that goes off in on direction. If that is the case, out come the lightbulbs with the alligator clips, and they are balanced again in an hour or so.

[roetter]

Interesting LFP testing underway as we speak.

I have a friend with a 38- foot mono. His 4 LA Exides got pretty much toasted several months when somehow when his boat got unplugged accidentally with the fridges on. A few days later the batteries were way down (I have no data on that) and since then they are unable to provide much capacity. So he was ready to install a low voltage cutoff to protect any new batteries, LA or LFP.

He was looking over my shoulder a lot in the last few months and decided to go for LFP as well. Except, he really wants to do it in a minimalistic way. I could not convince him to install any fancy gadgets. This gives us a good way to follow this as an experiment.

The starting specs:
- 5 year old boat used for weekend cruising and a few short cruises. So probably exactly what MainSail described - 150-200 cycles on the batteries.
- 4 Exide "Cutting Edge" golf cart batteries - 226AH, so a total of 452, with about 135Ah usable under cruising conditions (50-80%SOC, no solar)


- By the way I noticed these stickers on the Exides. Looks pretty dangerous to me. I failed to find any similar stickers on the LFPs, so they should be 'Mostly Harmless'.

- Hitachi 80A alternator(not 100% sure, as difficult to read off the label) regulated to 14.2 measured at the alternator. We have not found a split diode set yet or a combining relay. Had no time for that.
- Charger older model Promariner 1230.
Charger.jpg Views: 181 Size: 428.7 KB ID: 66672" style="margin: 2px" />
- No solar
- Xantrex Link10 battery monitor

This is what he is going to add:
- Low voltage disconnect. Still investigating which one. Has to be simple for him ($200)
- Junsi CellLog ($25)

This is what we changed:
- Shore charger, flipped the switch from LA to GEL (30 seconds)
- LA - 14.7V absorption (set to the minimum of 1 hour), 13.5V float
- GEL - 14.1V absorption (set to minimum 1 hour), 13.5V float

Just to make sure this sinks in. On this LFP install, all we did so far, we flipped the switch from LA to GEL on the shore charger to lower absorption voltage from 14.7V to 14.1V. NOTHING ELSE!!! The low voltage disconnect will be installed later. He is currently out for an overnighter with the admiral. I forced him to hook up a Junsi claiming that he owes me data. That will hopefully happen when he arrives at the destination later this afternoon.

The new LFP batteries are sitting in the bilge - see picture below - where two of the Exides used to to be. The other two Exide were under a bench. That space is now valuable storage space for the admiral.




Installation took us about 30 minutes. We hooked up shore power and verified that the charger was outputting 30A into the batteries, which were at 75%SOC when we installed them.

We then started the engine, and the alternator charged at 65A with about 13.7V and fairly quickly reduced to around 50A. The Hitachi has a temperature-sensitive resistor in the regulator, which cuts back with rising temperature. About 30 minutes later we were at 13.8V with 30A going in.

The admiral was really pushing to go out with the boat, so we did not have any more time.

P.S. Just got message from him that the monitor is showing 13.9V with 9A going in, so they are 95+% full with charging almost stopped. Interesting to see what I can read off the Junsi tomorrow. I asked him not to plug in overnight, to get to see a good charge cycle from the alternator tomorrow. I hope the winds are not good for sailing back.

[T1 Terry]

Originally Posted by EddieS

Hey Folks

Originally Posted by EddieS

I wonder if I might get some help. Im enamored with the idea of the Lithiun Ion batteries. Not because of any one factor, but I want a system that is dependable, safe, and doesnt need a lot of care and feeding. I'm considering the Lithionic or the Smart battery. Yes, I know, huge expense, but I'm a big fan of plug and play. My time to R and D things is at a premium. So I suppose my question is; are these types of LI batteries good, reliable purchases?
Thanks
Ed

If people want to buy a plug and play set up and want to throw the $$ at it, I can't see an issue with that, DIY isn't and probably shouldn't be for every body.
What plug and play systems can you trust? I guess the company tells you that. Their honesty in what products they use, then you can research into the quality but more importantly, the reliability of the product use from each of these suppliers.
If you are satisfied at that level, what R & D can the plug and play supplier show you, real data, not some made up stuff and endorsements from the their family members and friends, actual graphs showing cell voltage logging and how the system reacts to high and low cell voltages, does the system monitor SOC%, how does it auto reset and verify it's readings. If you don't understand the answers given, maybe start another thread and put the questions and answers forward, there is quite a bit of experience and knowledge building up within this group on this forum and they will quickly tell you if it's factual or gobbledegook mumbo jumbo.
Next, how confident are they with their product and how confident are their backers? How long is the guarantee/warranty, repair/replace or pro/rata? Who backs that warranty, an insurance company with strong roots or the flashy shop front of a leased building and a company that was establish a week ago behind it?

I started learning as much as I could back in 2008 about any other form of battery after too many let downs by lead acid and being frustrated with their very limited capabilities. I learnt the most important bits from the man who designed the battery packs for the Mars Rover missions, including the latest one, if NASA keep using him and he can build a system that outlasts anyone’s wildest dreams he must know what he is talking about.
Unfortunately, commercial interests only started looking at these batteries in the last couple of yrs, most of them don't even seem to know there are different chemistries within the lithium battery group, some of them use the fear factor to try and scare you into using their product so big names doesn't necessarily mean they really know what they are talking about. Although it has recently been revised and some of the huge truth distortions have been tamed, this document still has nonsense mixed with fact, you would think they were big enough to trust, http://www.victronenergy.com/upload/documents/Datasheet%20-%2012,8%20Volt%20lithium%20iron%20phosphate%20batt eries%20-%20rev%2004%20-%20EN.pdf I wonder how many here can spot the nonsense or distorted truth?

It is a mine field, there are some companies that have got it sorted to a fair degree, their systems aren't cheap but they would have spent big $$ paying their top people to learn, they are going to get those $$ back plus some and the people who want someone else to do all the research and testing for them are the ones who will pay them for doing it, some are just there to cash in and get out quick, the $$ sparkle I their eyes.

T1 Terry

[vRUN]

Originally Posted by T1 Terry

... DIY isn't and probably shouldn't be for every body.

While I do agree to that for most of my life, electronics is the part where I do like to tinker

Up to now I had little need, now I'm late to the LiPo_on_board-party
...and most likely haven't read all available material on the forum (mainly "...for the rest of us"). Forgive and just point me if I fail with my 1st question...

Charging my LiPo House bank
I was wondering what the pragmatists among you do use for
"everyday charging with land power"-machinery.
I've seen that MaineSail uses his regulated Lab-PowerSupply ...
which is a method my first mate would politely reject ;-)

So let's assume new equipment:
What do you use for _optimum_ charging ? (please don't forget to mention Cell type)?


DIY Electronics - Maxim
I looked and there is much charger material in the form of ICs on the market.
Maxim describes a charger for A123 Systems "ANR26650m1".
I've used them and think we all know these cells:

That Cell accepts a standard charge mode at 3A (1.3C) and
A123 ANR26650m1 can be fast-charged at 10A (4.34C)
with a charge-termination voltage of 3.6V.
Here you guys may finde the application note: No.4520 (pdf)
Abstract: By adding a dual op amp and some other components to a lithium-battery-charger IC, you can adjust the charging voltage to any level within the range for standard lithium batteries (3.6V to 4.2V). (Generally the MAX1737 charger is internally set to switch from constant current mode (CC) to constant voltage mode (CV) at 4.2V ±0.8%)

Boat Batteries - Maxim IC based Charger
Now your (our?) on board batteries are prismatic like:

I would be quite confident building a Maxim IC based landline charger using above application and an off the shelf switching PS.
3A(1.3C) - fast charge 10A (4.34C) termination voltage at 4.0 Volts (can be 3.6V to 4.2V -selectable).
What can I expect with Winston batteries like: WB-LYP40AHA (pictured):
4hours charge from empty to full charge for a LYP40AHA 4S setup ?
(To T1 - thanks for all the detailed comments but I'm still unsure )
Do I set termination voltage at 4.1 Volts - safely?

Regards




P.S.:
Those that read the application Note and have (electronics) questions: I'd be happy to answer what I can!

P.P.S.:
Maxim integrated also has written an Application Note (No.676)
New Developments in Battery Chargers
http://pdfserv.maximintegrated.com/en/an/AN676.pdf
which I do recommend reading to all that would like to have
a view on what's available in IC Form (from the market leader).

[goboatingnow]

Sorry. I meant to say HVC should be set above the charge termination point


Dave

[ebaugh]

Originally Posted by vRUN

P.S.:
Those that read the application Note and have (electronics) questions: I'd be happy to answer what I can!

P.P.S.:
Maxim integrated also has written an Application Note (No.676)
New Developments in Battery Chargers
http://pdfserv.maximintegrated.com/en/an/AN676.pdf
which I do recommend reading to all that would like to have
a view on what's available in IC Form (from the market leader).

I'm not sure any of the batteries specifically listed in the application note are LiFePO4 or what we call LFP. There are very few "mainstream" applications using this lithium chemistry yet. The voltage for LFP is usually quoted at 3.2 or 3.3V.

Many marine battery chargers and inverter chargers are programmable to work well with LFP, most users are looking for 100A plus charge rates, building one not so easy.

On a nominal 100Ah cell, fully discharged, it takes maybe 101.5Ah to recharge, from my testing.

[senormechanico]

These are the cells I have on my boat.
Note the voltage recommendations.

http://www.evsource.com/datasheets/B...wer%20cell.pdf

So far in the last year and a half, I have been charging with 14.4 bulk/absorb and 13.8 float with no shore power use at all.

All this talk about cell damage has caused me to reprogram my BlueSky
controller to 13.8 everything.

I have a Clean Power BMS, but under these conditions I doubt it will ever get a chance to balance anything.
Cell # 4 has always been the last one to shunt.

Suggestions?

Steve


Steve

[ebaugh]

Originally Posted by senormechanico

These are the cells I have on my boat.
Note the voltage recommendations.

http://www.evsource.com/datasheets/B...wer%20cell.pdf

So far in the last year and a half, I have been charging with 14.4 bulk/absorb and 13.8 float with no shore power use at all.

All this talk about cell damage has caused me to reprogram my BlueSky
controller to 13.8 everything.

I have a Clean Power BMS, but under these conditions I doubt it will ever get a chance to balance anything.
Cell # 4 has always been the last one to shunt.

Suggestions?

Steve

Steve

There are no absolute answers. Your original configuration is within the normal operating range. Some of us are hoping to extend life by not pressing to 100% every cycle. There is some research that suggests that, but from a practical perspective installed for marine cycled at low C rates, we don't know for sure if there is any impact or not. We do know from testing that 13.8 will get you at least a 95% charge (some new cells I tested 98%) and we have observed that once balanced, in normal operations, balancing is not needed often, if ever.

I think your new configuration is better. If on a sunny day you are reaching 13.8V for several hours, I would try and get the solar controller to reduce even further after reaching 13.8, to say 13.4. If you rarely see 13.8, it's probably not worth worrying about.

Plugged into a marina, 13.8 is almost certainly too high for 24x7 floating.

[goboatingnow]

Plugged into a marina, 13.8 is almost certainly too high for 24x7 floating.

I would not recommend floating Li anyways

dave

[vRUN]

Originally Posted by ebaugh

I'm not sure ... what we call LFP.

...most users are looking for 100A plus charge rates, building one not so easy.

Thanks for looking at the data.
- How do YOU charge your LPF House bank ? <smile>
- Where can I find a "marine charger" LPF charge algo ? <GRIN>

Up to now I thought A123 was once THE LiFePO4 battery maker.
Isnt' the 18650 also a "LPF" which A123 'improved' with Nanophosphate?

My real need for a 400Ah house bank is still a few (working) years away (not cruising yet),
but 100A shore charge doesn't seem an unsurmountable difference/problem to me,
if at the circuit core I can use an off the shelf 24V switching PS which is readly available:
ICs I looked at are all 90%+ duty cycle and thus we look at FETs
and Shottky Diodes that have to handle 100A but dissipate very little at fairly low frequencys - not a problem.
(Don't let the App Note fool you. Shasta Thomas uses a Dual Fet Fairchild FDS6990 the size of a halfed peanut.
The whole charger part, now at 3 A, is maybe 10 US$ and will scale to 10 A for maybe 5 US$ easily (DIY)).

From reading all the comments in this thread, specially T1s and yours I believe
that charging with curves/algorithms like this is optimum for LPF:

3A Battery Charge current vs. Battery Voltage
(can scale to 10A++, Battery ShutoffVoltage selectable between 3.6V to 4.2V)

Algorithm with a 2A Battery Charge current vs. Battery Voltage


Is this correct?
What "termination voltage" would you choose for your "marine""LPF Cell"?

Building my Model Bank
If I am going to build a 1/10 scale model of say 40Ah 4s1p,
assume I am always at port for 10hours+ shore-charge...
What max charge current would you set ?
What termination voltage should I use ?
- For 4 x WB-LYP40AHA LiFeYPO4 (3.2V/40Ah) what should I pay (ex.shipping ex.VAT) ?

[ebaugh]

Originally Posted by goboatingnow

I would not recommend floating Li anyways

dave

Dave,

I need to find it again, but I did see a paper that suggested floating, especially near full charge was a bad idea. If I remember correctly, it basically suggested there was little difference in the capacity loss of cells stored at 100% and cells continuously cycled.

But if you are a 100% LFP boat without solar, what do you suggest when at a marina for long term? I have some options, but this is a huge design issue for marine banks.

I can with my existing equipment:

1) Switch off the LFP (at 50% SOC) and Float on my engine start batteries (AGM), but a power failure would leave those dead in 2-4 hours from the refrigeration load. I suppose I could leave the inverter set to not invert, but then the freezer thaws.....

2) By accident, sort of, the charger has the ability to cycle between say 13.4 and 12.8. This is really designed for FLA to limit overcharge and the upper limit is adjustable, the bottom is not. After reaching the upper limit for a few hours, it switches the charger off until reaching the lower limit and then switches the charger back on. Still only a couple of hours of juice left if they power fails when the LFP are at 12.8V.

[ebaugh]

Originally Posted by vRUN

Thanks for looking at the data.
- How do YOU charge your LPF House bank ? <smile>
- Where can I find a "marine charger" LPF charge algo ? <GRIN>

Up to now I thought A123 was once THE LiFePO4 battery maker.
Isnt' the 18650 also a "LPF" which A123 'improved' with Nanophosphate?

My real need for a 400Ah house bank is still a few (working) years away (not cruising yet),
but 100A shore charge doesn't seem an unsurmountable difference/problem to me,
if at the circuit core I can use an off the shelf 24V switching PS which is readly available:
ICs I looked at are all 90%+ duty cycle and thus we look at FETs
and Shottky Diodes that have to handle 100A but dissipate very little at fairly low frequencys - not a problem.
(Don't let the App Note fool you. Shasta Thomas uses a Dual Fet Fairchild FDS6990 the size of a halfed peanut.
The whole charger part, now at 3 A, is maybe 10 US$ and will scale to 10 A for maybe 5 US$ easily (DIY)).

From reading all the comments in this thread, specially T1s and yours I believe
that charging with curves/algorithms like this is optimum for LPF:

3A Battery Charge current vs. Battery Voltage
(can scale to 10A++, Battery ShutoffVoltage selectable between 3.6V to 4.2V)

Algorithm with a 2A Battery Charge current vs. Battery Voltage

Is this correct?
What "termination voltage" would you choose for your "marine""LPF Cell"?

Building my Model Bank
If I am going to build a 1/10 scale model of say 40Ah 4s1p,
assume I am always at port for 10hours+ shore-charge...
What max charge current would you set ?
What termination voltage should I use ?
- For 4 x WB-LYP40AHA LiFeYPO4 (3.2V/40Ah) what should I pay (ex.shipping ex.VAT) ?

I think the 18650 is a physical size, like AA or D batteries, not a chemistry spec. A123 is the only US made LFP cell I'm sure about, although there is at least one other US company marketing prismatic cells with their name on them, mostly to the military. Everyone here is using Chinese made cells.

All you need to do below .5C rate is 2 steps, charge to 3.45V, then terminate or drop to like 3.34V to maintain charge. This gets 95% charge. Change the upper limit to 3.6V, gets 99% or better. Many of the battery specs call for CC/CV, but at low rates, it's just not needed. Rate can easily be up to .5C for most all cells, many will charge at 1C, finishing in an hour. If you want to test cycle, estimate your loads and final bank size and scale for your model.

A rough rule of thumb cost is $1.25 US per Ah, so about $200 for 4 40Ah cells. For that small a cell, maybe $250.

The Victron, Mastervolt, Outback, and Magnum inverter chargers can all be programmed to do this. So can many shore power chargers. But it's a not always a pre-programmed profile. You have to custom set the parameters you want in the chargers configuration.