Alternator Size with LiFePo4

[newhaul]

Originally Posted by nebster

I dunno, not enough information.

That said, if you were to dial up a bench power supply to 13.8V and put it on an LFP cell, it will eventually overcharge that cell at any current limit you set.

Does your solar controller have an absorption setting? How do you have it configured, if so?

the controller stops charging at the point I set it to . No absorption no float pure and simple it hits ( I have it set at ) 13.8v and stops charging . While still carrying the house loads up to the amps that the solar is putting out. It drops to 13.7 and it starts charging again.
It has been working for a hear this way but I'm just making sure its not killing me prematurely.

[nebster]

Originally Posted by john61ct

I find that last term confusing.

Sorry, what I meant was "a charge profile that only has CC to a stopping voltage, without any CV."

Again, my daily cycling charge profiles are To a Voltage then Stop, no Absorb/CV at all.

Yes, this is an unusual choice but fine in the narrow context of (a) low rates, say < 0.15C, and (b) willingness to be conservative about SOC in exchange for even more longevity.

Once you get to higher rates, your strategy will not work very well, or you will be leaving a huge slice of your battery unused, sometimes.

If I were concerned about getting to Full in that context - which I am not - I could do so by varying voltage.

Actually, that's not necessarily true. At higher rates, say >0.3C, my data show that you cannot achieve any desirable SOC simply by increasing the CV threshold. The voltage will spike up, earlier and earlier, at higher rates. You will be compelled to have an absorption charge phase if you want to achieve higher SOCs in that case.

I recognize that "your" system (which you will not disclose, probably because you have not actually built a real system) might only have low rate charging. But that is not the case for some of us.

The **only** time I use a CV stage / hold Absorb V, is for benchmarking purposes, and that (of course) requires using a shunt to stop-charge based off endAmps.

Yes, unusual, but okay?

In any case I'll not dispute your claim that CC-only charging **in itself** leads to significantly lower SoC when charging stops, since I don't see that as a downside anyway.

No problem, but others will see it as a massive downside if they can't get their pack to charge beyond 70% when they fire up their big generator.

A lot of us want to stop listening to our engine and have peace and quiet. A primary objective for my system was to be able to have enough battery capacity to (a) go 3 days without charging and (b) charge to full in under three hours when it's time to charge.

I hate listening to the genset!

However the high-current rate just makes a higher voltage **safer** (less impact on longevity) with CC-only profiles.

It definitely puts a lower limit on the achieved SOC, which certainly has been shown to improve longevity at least bit. I don't think we know whether it might have any counterbalancing detrimental effect, perhaps due to local heating or something. I'm not sure that experiment has been conducted.

With an Absorb stage, higher current does make shunt-based stop-charging based on endAmps **more** important, since that point will be reached much more quickly than at low currents.

I wish you would just leave out "shunt-based." It's both incorrect (because there are other ways to measure current) and redundant (because how else do you stop at endAmps other than by measuring current?). It's also misleading, since it suggests that someone has to use a standalone shunt to achieve this kind of charging, when in fact it can be done by the charger itself if the charger is smart enough.

Well too low and you're charging much more slowly, both for extracting max power from solar, and if ICE is running for the sole purpose of charging, that is not ideal.

That's right. But too high, and you're bringing cells up into the danger zone needlessly. And if you bottom-balance, like I do, you've even more incentive to stay away, since the weakest cell will shoot up earlier.

I think my "13.6 for low rates, up to 13.8V for high" strikes a decent balance and is easy to both understand and to implement.

I've settled on 3.388V per cell for the threshold, CV to 7% of charge current one hour, whichever comes first. This yields 92-94% true SOC on my pack at any rate between 0.1C and 0.35C, with most of the variation seeming to come from what temperature the battery pack is when the charging starts.

For solar, where realistically I can't get more than about 0.04C, I use a slightly different strategy.

[Q Xopa]

Originally Posted by Maine Sail

Why......? CC/CV charging will do all this for you. Set it at 13.8V - 14.0V and have it stop at .03C - .025C & you're done.

You're over complicating something that is actually quite simple.

I think John has asked this, but how do you do the 'stop at .03 - .025'?

[nebster]

Originally Posted by newhaul

the controller stops charging at the point I set it to . No absorption no float pure and simple it hits ( I have it set at ) 13.8v and stops charging . While still carrying the house loads up to the amps that the solar is putting out. It drops to 13.7 and it starts charging again.
It has been working for a hear this way but I'm just making sure its not killing me prematurely.

It sounds kind of like you are essentially floating the pack at 13.8. I would probably nudge this down a little bit, if it were me. I dunno that it's a grave error, but that voltage is definitely high enough to fully charge the cells if that is a reading being made at the cells themselves.

(Important note: many solar controllers can't take a remote voltage reading at the cells, and if yours is one of those, this discussion so far is moot, and you'll have to go measure at the battery to see what the real value is.)

Ideally, I would have it charge to 13.8 and then fall back to a lower float voltage for the remainder of the daytime supplementing. The next morning, it can do that all over again.

[nebster]

Originally Posted by Q Xopa

I think John has asked this, but how do you do the 'stop at .03 - .025'?

Here are a few options:

1. Buy a charger that supports that mode.

2. Buy a Victron inverter/charger and hack it back in with your own code, while we wait for them to give it to us as a first-class firmware feature again.

3. Buy a BMS that calculates it, and let the BMS send a signal to turn on and off the charger.

I bet there are more...

[newhaul]

Originally Posted by nebster

It sounds kind of like you are essentially floating the pack at 13.8. I would probably nudge this down a little bit, if it were me. I dunno that it's a grave error, but that voltage is definitely high enough to fully charge the cells if that is a reading being made at the cells themselves.

(Important note: many solar controllers can't take a remote voltage reading at the cells, and if yours is one of those, this discussion so far is moot, and you'll have to go measure at the battery to see what the real value is.)

Ideally, I would have it charge to 13.8 and then fall back to a lower float voltage for the remainder of the daytime supplementing. The next morning, it can do that all over again.

ok first off I only do all of this when I'm aboard and on the hook / underway
( at the dock I run the bank down to about 40% then disconnect. And run off charger and Fla in port. )

when I hit the 13.8 that's when I fire up the watermaker and once its operational the holding plate fridge. Would be a good time to fire up the inverter running the starting Fla battery charger. As needed .
All of this is dependant on battery charge when I get up. If more than 50% then I disconnect the solar so no charging .

[nebster]

Originally Posted by newhaul

ok first off I only do all of this when I'm aboard and on the hook / underway
( at the dock I run the bank down to about 40% then disconnect. And run off charger and Fla in port. )

when I hit the 13.8 that's when I fire up the watermaker and once its operational the holding plate fridge. Would be a good time to fire up the inverter running the starting Fla battery charger. As needed .
All of this is dependant on battery charge when I get up. If more than 50% then I disconnect the solar so no charging .

Yeah, I mean you're basically being a human battery management system there, so as long as you're in the loop you can take all kinds of liberties with charge algorithms. It sounds like you are never really charging into 100% SOC for any length of time, which would be the real killer.

For those of us who have to be away from our setups sometimes for extended periods, but where the house bank still needs to perform 24/7, we have less choice in the matter. (And, unfortunately, there is no way I'm convincing my wife or dog to fool around with charge control algorithms when I'm gone!)

[Q Xopa]

Originally Posted by nebster

Here are a few options:

1. Buy a charger that supports that mode.

2. Buy a Victron inverter/charger and hack it back in with your own code, while we wait for them to give it to us as a first-class firmware feature again.

3. Buy a BMS that calculates it, and let the BMS send a signal to turn on and off the charger.

I bet there are more...

Ok sounds good, any suggestions which Chagers can do this? I dont know of any.

[newhaul]

Originally Posted by nebster

Yeah, I mean you're basically being a human battery management system there, so as long as you're in the loop you can take all kinds of liberties with charge algorithms. It sounds like you are never really charging into 100% SOC for any length of time, which would be the real killer.

For those of us who have to be away from our setups sometimes for extended periods, but where the house bank still needs to perform 24/7, we have less choice in the matter. (And, unfortunately, there is no way I'm convincing my wife or dog to fool around with charge control algorithms when I'm gone!)

yea that sums it upin quite well . Now if I would be gone for extended periods I would set the controller to stop charging somewhere well south of 13.8 likely around 13.4 would work and still keep my refer running as needed with no shore power .
And no human intervention.

[nebster]

Originally Posted by Q Xopa

Ok sounds good, any suggestions which Chagers can do this? I dont know of any.

Off hand...

I think Xantrex makes a few 12 and 24V I-C units.

Victron solar charge controllers move out of absorb at a non-configurable tail current (2A) that should work okay for most batteries.

Pretty sure the Outback MPPTs have an end amps setting.

Victron I-Cs have the hardware but not the software, currently. They are so programmable that it can be done anyway, but it's harder than it should be.

John said the Magnums can do it, but I had a Magnum with the ME-RC a few years ago, and I'm fairly sure mine didn't have that option. You might have to look at each product line to know for sure.

Some of the automotive (EV) chargers have the ability to be programmed with a termination current. I don't know how feasible this is for a single install, but I'd be curious to find out.

[Q Xopa]

Originally Posted by nebster

Off hand...

I think Xantrex makes a few 12 and 24V I-C units.

Victron solar charge controllers move out of absorb at a non-configurable tail current (2A) that should work okay for most batteries.

Pretty sure the Outback MPPTs have an end amps setting.

Victron I-Cs have the hardware but not the software, currently. They are so programmable that it can be done anyway, but it's harder than it should be.

John said the Magnums can do it, but I had a Magnum with the ME-RC a few years ago, and I'm fairly sure mine didn't have that option. You might have to look at each product line to know for sure.

Some of the automotive (EV) chargers have the ability to be programmed with a termination current. I don't know how feasible this is for a single install, but I'd be curious to find out.

Thanks, I'll see what I can find out.

[rgleason]

Nebster wrote:
"I've settled on 3.388V per cell for the threshold, CV to 7% of charge current one hour, whichever comes first. This yields 92-94% true SOC on my pack at any rate between 0.1C and 0.35C, with most of the variation seeming to come from what temperature the battery pack is when the charging starts."

Thanks, very helpful. These can be programmed into the Balmar regulators guite easily, except for the trailing amps of .07C ( is that what you meant?). The hour is programmed in 6 min intervals so use "010".

For LiFePo4 batteries what is the Battery Temperature Compensation value that should be used? LFP have low resitance but they dont like getting too hot. Is it the same value as FLA? I doubt it.

I'd like to try an alternator charge routine that does CC to about 13.588v (nebster) or 13.88v (johnct61) which appears to be about SoC75%, then have CC where the voltage is at lower by 0.2 or 0.1v and the current gradulally drops to .07C (Nebster) or .03C (johnct61) as measured by the regulator.

The Balmar uses time as a proxy for actual measurement. The correct setting for time will likely change, so an actual measurement would be better, particularly with LFP which is aensitive to overcharging due to the high slope at the shoulder. The hour seems reasonable, because I've seen testing results that say 100%Soc will take between 1 and 2 hours by measuring trailing amps down to 30mA which I believe is 0.3C for a 100ah battery being tested (Powerstream link. ).
https://www.powerstream.com/LLLF.htm

Its very helpful to have some field experience reporting that charge rates of .01C to .35C under your settings yield 92-94% of true capacity

Al Thompson has a regulator that will work, measuring current with a shunt. It will also send a signal to disconnect the ignition wire when the alternator should stop to protect the batts from overcharge and the alternator diodes when the LFP charge relay opens (which also coordinates with BMS). I believe these features are all programmed now. There is also a bluetooth app being worked on.

To protect the alternator from damage an old style zap stop or reverse diode on the alternator might also work. I believe Sterling or Balmar still has these, but with alternator capacity increasing so much these might not be as effective.

John has written Al about this application.

[john61ct]

Relying on sensing your charger current output only, as a proxy for bank acceptance, will never be accurate as long as variable loads are online.

[nebster]

Originally Posted by rgleason

Nebster wrote:
Thanks, very helpful. These can be programmed into the Balmar regulators guite easily, except for the trailing amps of .07C ( is that what you meant?). The hour is programmed in 6 min intervals so use "010".

I meant ~7% of 0.3C, or something like 0.02C.

For LiFePo4 batteries what is the Battery Temperature Compensation value that should be used? LFP have low resitance but they dont like getting too hot. Is it the same value as FLA? I doubt it.

No temperature compensation.

I'd like to try an alternator charge routine that does CC to about 13.588v (nebster) or 13.88v (johnct61) which appears to be about SoC75%, then have CC where the voltage is at lower by 0.2 or 0.1v and the current gradulally drops to .07C (Nebster) or .03C (johnct61) as measured by the regulator.

Okay, go get started. Let us know how it goes!

[john61ct]

Originally Posted by newhaul

so I will be fine with my solar maintaining 13.8v constant out of my PWM solar controller

That sounds very different from stop-charging when hitting 13.8V, call that A.

That latter is safe, but in fact means the LFP circuit voltage is always lower.

Continuing charge in CV stage requires measuring endAmps with a shunt at the bank for the stop-charge, call that B.

Unless you guesstimate using an AHT setting (C), or babysit yourself (D), or rely on sunset and constant loads (E).

I prefer A, which later you say you use.

B is required if you want to get to a consistently high SoC (as MS emphasizes not an important goal).

The rest IMO are not reliable enough.