[goingbush]

Now that iv'e finished my conversion I need to think about what needs to be done if I go away for 3-4 months .

If I fully charge the LifePo4 pack 45 x 200Ah Sinopoly cells would I expect the cells to be healthy after 4 months or do I need to devise some kind of 150Volt trickle charge ??

 

[Duncan]

Hi
My tuppenceworth

Don't fully charge them - they live longer at about 1/2 charge

Disconnect them! take out the main fuse or switch off the contactors
Whatever is required to ensure that they cannot discharge

Then they will sit perfectly happily for a couple of years if necessary

This is one of the reasons that I don't want a BMS - more batteries have been killed by BMS's than saved

 

[goingbush]

Thanks Duncan.

That is great news . Do you mean don't fully charge them ever , or just when laying up ??

As it happens I have a Zeva BMS , i preordered the new 3rd Gen & probably one of the first to have one installed .

I think its really terrific , Iv'e been watching my cells as they charge they are in perfect balance up to 3.4 volts , well all within .01V anyway but once over 3.4 V one cell skyrockets & goes past 3.6 within a few minuites while the rest of the pack is on 3.41 or 3.42 , then within 10 seconds it jumps to 3.8 & the BMS disconnects the Charger , after a few minuites that cell has stabilised back down to around 3.41 . Each charge It seem to be a random cell that skyrockets.

So in effect my BMS is saving my pack , it stopping the charger going to 165V where it will internally cut off & would have that high cell way overcharged. The most I'm getting per charge is about 152V so I guess I won't worry & be happy that my cells will last longer .

(If I installed a battbridge I'd now be pulling my hair out !!)

Am I looking at this the right way ??

cheers Don

 

[Duncan]

Hi Don

I meant 1/2 charge for the winter layup

Re your cells - did you top balance them first?

And what does the charge/voltage graph look like?

I've got a Volt pack now - I charge to 4.05v per cell - and they are always within 1/100th of a volt of each other

I would have said that if some of your cells are jumping up the voltage curve then you have already charged then a little too much
If when some of them are at 3.41v - you get some on the upramp then I would be shutting off while they are all at 3.40v - or a little earlier

That upkick does no good and should be avoided!

I got the impression that you had more than enough range? - so maybe 3.35v?? - how much range will you lose?

I worry about BMS's because some of them can drain the cells down when left unattended - any small current adds up over months

With mine the contactors open and then the main pack is isolated - no current at all

When I had Headway's I did have some failures

They were sold as 16 Ah - so I used 80%
But the buggers weren't 16 Ah - they went as low as 12 Ah - so I killed some

Some just died!

The BattBridge was great for telling me that a cell had died - so I knew NOT to try and charge that string until I had fixed it

In your case the BMS is stopping you from massively overcharging - but IMHO you are treading on the ragged edge
You would be better a bit further away from the edge

 

[goingbush]

Thanks mate good to know, I'll back off the upper limit to 3.4 and see how we go.

I do have adequate range as it stands , I was just curious what range I could actually get from the vehicle when push comes to shove . In one test I did 55km mixed driving with the battery showing 49% , I can extrapolate to 100km so happy with that,

The cells were all balanced at 3.3 +/-.01 before first charge ,

The Zeva BMS only balances whilst charging, it turns itself off after an hour and consumes 3mA whilst sleeping, but I have a red button to totally isolate the system via a maintenance contactor so no drain while laid up.

 

[Duncan]

Hi Don

Problem
You can't balance at 3.3 v

You must go up into the uphook to balance - so 3.5 v or 3.6 v

Top balance is in the top uphook
Bottom balance is in the bottom downhook

Trying to balance in between is pointless as the charge/voltage is not steep enough

Your Red button - how do you know that it stops the Zeva from drawing current?
It probably does - but you need to talk to somebody who knows electronics (NOT me) to confirm that it does

3mA - is 9 Ah over 3 months - which should be OK

As long as it is evenly spread there was one BMS which only drew down one of the eight cells in it's string - and taking 72 AH out of one cell could cause issues

 

[john61ct]

Do you mean don't fully charge them ever , or just when laying up ??

LFP should never be fully charged except when you are just about to start discharging. Sitting for more than a few hours at even 3.4Vpc will reduce cycle lifetime.

For longevity, storage should be at the lowest SoC that ensures self-discharge will not approach the bottom voltage knee. Also cooler the better.

If I knew I could check on the bank every month or so, I'd discharge to 2.9Vpc then add .05C if I think it might get down to 2.7Vpc before the next visit.

If I was really leaving them alone for 4 months, I'd leave at 50%, if for lots longer maybe set up a low current 3.1V charge source to turn on for an hour or two once a week, or ideally try to match the self-discharge rate more precisely.

Obviously talking ideal coddling conditions here, have to balance with what's practical for your setup.

Sitting at high SoC is certainly not an issue compared to risking dead flat.

 

[goingbush]

Thanks fellas, I'm learning a lot here.

Having just spent $10,500 (AU) on my LFP pack I don't want to abuse them.

I don't mind sacrificing top end for longevity , especially when the pack is still 50% when I get home after a typical drive.

Its been over 35c and into the low 40's (95F -104) every day here since Iv'e completed my LandRover , I don't suppose the heat is being kind to the charging either.

 

[john61ct]

the pack is still 50% when I get home after a typical drive.

Keep that up you'll get 3-5x the lifespan.

Or at least help compensate for the heat, if that's the norm a real longevity killer.

Little PC muffin fans are pretty efficient, the ones designed for racked servers move a lot of air

 

[Duncan]

Keep that up you'll get 3-5x the lifespan.

Or at least help compensate for the heat, if that's the norm a real longevity killer.

Little PC muffin fans are pretty efficient, the ones designed for racked servers move a lot of air

Not if he is charging to nearly 100% and then discharging to 50%

Now charging to 75% and discharging to 25% - or something similar would give more life and the capability to go to 100% if you really needed to do it occasionally

 

[john61ct]

No.

As long as your definition of 100% Full is conservative, well short of the shoulder,

(e.g. mine is "charge to 3.45Vpc then stop, no Absorb, never Float")

there is no harm at all in going there as long as some level of discharging follows shortly thereafter.

If normal use stops before 50%, the lifetime cycles with low C-rate usage might be say 4500 cycles, while going from 75%-25% might be 2500.

Obviously the actual numbers will vary but every LFP makers' charts show radically steep curve ratios between DoD and lifetime cycles.

 

[Duncan]

No.

As long as your definition of 100% Full is conservative, well short of the shoulder,

(e.g. mine is "charge to 3.45Vpc then stop, no Absorb, never Float")

there is no harm at all in going there as long as some level of discharging follows shortly thereafter.

If normal use stops before 50%, the lifetime cycles with low C-rate usage might be say 4500 cycles, while going from 75%-25% might be 2500.

Obviously the actual numbers will vary but every LFP makers' charts show radically steep curve ratios between DoD and lifetime cycles.

Disagree - and the most expert manufacturer (Tesla) appears to agree with me

A Tesla is normally charged to a lower percentage - with the ability to select "road trip" that charges it fully

But you don't do that all of the time or it reduces the battery life

From a basic chemistry point of view you need to avoid either extreme - and I'm pretty sure that "wear" occurs more at the top end than at the bottom end

 

[john61ct]

Are you under the impression that all "lithium" batteries are the same?

It is possible what you say is true for Tesla packs, but that is not relevant here.

I know that Tesla doesn't use these big prismatic cells. I don't think they even use any sort of LiFePO4 chemistry.

In any case what you say is not true for LFP.

The protocol I outlined for the top end is plenty safe, far below the shoulder, in fact most EV users would claim sacrifices too much usable capacity, but in my experience is within 3-4% of rated AH.

The critical aspect there for longevity, as I said, is to not let the bank *sit* at a high SoC.

And again, if you get the spec charts for these cells showing cycling DoD vs cycles lifetime, the shallower you discharge, you get *radically* better longevity.

Obviously leaving 40% as a "reserve" for contingencies and occasionally longer trips is a good thing.

 

[john61ct]

Tesla has been using lithium nickel manganese cobalt oxide (NMC)

Model S has evolved that to a nickel cobalt aluminum oxide (NCA) flavor.

Neither of these have anything to do with LFP, other than the fact they (and dozens of other widely different chemistries) all wear the "lithium ion" family label.

 

[Duncan]

Other that the fact that like all lithium battery chemistries they age more at the top of the charge

All of them

 

[john61ct]

Then please resolve the inconsistency with this

?

Which every LFP large-prismatic chart matches. Note the logarithmic scale on the Y axis.


Again, defining "Full" as "stopping charging at 3.45Vpc" is well below any definition of "at the top"

and far enough below the voltage shoulder to avoid issues.

Hundreds of boats set up like this for daily House bank use have now passed 2-3000 cycles with little to no loss of AH capacity, we fully expect 10,000+ cycles to be the norm.

 

[Duncan]

Your chart
Like all of the others simply ASSUMES a 100% start point

Which is a hangover from Lead acid where it was useful to charge to 100% to reduce stratification

The problem is that charging to a lesser number works better - but is not included on that chart

 

[john61ct]

Yes, and I've already specified the charging to a lower point.

The differences between each if the point on the curve there are where you *stop discharging*, and everyone with a clue knows, the shallower that is the longer the lifespan.

If you're actively cycling using say 60% of capacity each time, it would be crazy to only charge to 80% so that every cycle you had to draw all the way down to 20%.

Obviously we're not talking "damage" here, but if a bank can last 15 years, why settle for 5-10? Especially without getting any benefit in exchange.

 

[goingbush]

Thanks for the discussion, Im learning a lot here . I certainly would be in trouble if I left my LFP charging up to the charger. The Vendor wanted to know the amount of cells & chemistry when I ordered my Elcon / TC 3.3kw charger so they set the Constant Charge voltage at 3.65 x 45 = 164 v.

Im glad I have the Zeva BMS which Ive now set to 3.45 x 45 = 155
and now instead of 3.8 (as it was programmed) I have the BMS set disconnect the charger if one cell reaches 3.6 V .

Today is the first day since I finished the car that its been below 40c , I drove 70Km & the 20% BMS warning suggested I go home & charge, but the voltage was still at 143V , Im impressed . What voltage would you not go below ?? I set the minimum voltage at 2.4 but now after the discussion think thats waay too low . Not sure why the BMS thought 143 was low , the lowest cell was 3.14 V .

Now in the evening its a cool 28c so have the car on charge , interesting to see the SOC in the morning !!

 

[kennybobby]

That's partly why we reversed the TCCH chargers--the 3.65 was too high to hold the cells and they would overcharge and bloat.

From testing cells 100Ah cell we found that there is no useful energy left below 3.0--and likely very little available at 3.05-3.1, so that region could be considered empty on the fuel gauge. It's not "0% SOC" on the graphs and datasheets, but for all practical purposes it is.