LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[OceanSeaSpray]

Originally Posted by Singleprop

This link describes a way to limit alternator current (I think its a Balmar 612 regulator):

Alternator Output Management with External Regulators [Small Engine Mode]

All it does is reduce the maximum field current delivered by the regulator, which also reduces the torque absorbed at a given speed. In other words, it doesn't limit current, it just reduces the slope in the alternator output curve. The current doesn't increase as fast with RPMs, but increase the RPMs high enough and the current will keep increasing and still reach the same value in the end.

It just makes the alternator less efficient. Adding a power resistor in the field circuit achieves about the same when you don't have that functionality.

In order to be able to limit output current, the regulator needs to sense it and control to it and there is no mainstream external alternator regulator available that senses alternator output current.

I limited my 115A Mitsubishi alternator to 95A with a proper current-limiting regulator, but it still charges 50A at idle like before because it gets full field current as long as the RPMs are not high enough to reach the current limit.
It is the only way to do it if you want best performance and perfect control, there is no way around it.

[SV THIRD DAY]

Originally Posted by Singleprop

This link describes a way to limit alternator current (I think its a Balmar 612 regulator):

Alternator Output Management with External Regulators [Small Engine Mode]

We have to remember what the problem is...heat...heat that the alternator can not dissipate fast enough. So how do you limit heat...either increase cooling or limit the Amp production. Of the 6-7 LiFePO4 bank set-ups I know about and have helped out with, they are all using the Balmar 614 regulator. Specifically using the belt manager function to limit the output of the alternator, for example making a 160A into a 100A.

Next let the regulator monitor and dial back the alternator output as needed if the temperature hit the set-point. The regulator reduces current by about 10% increments until the heat dissipation and heat generation of the alternator is at equilibrium and the alt stays at a steady constant (survivable) temp.

So at heart here is an Alternator Problem, simply stated that they are not designed to run at or near their rated output for extended periods of time that could happen using a LiFePO4 Bank. So external regulation is used to hide or fix this problem...or should we just say design reality of standard marine alternators that most of us are using.

[OceanSeaSpray]

You need to remember that, no matter what, LFPs will always absorb all the current you throw at them, so any alternator is going to overheat if you rev it fast enough, unless the stator hits magnetic saturation before it can get too hot.
You can see them hitting magnetic saturation when you can increase the revs, the voltage is below the regulation setpoint and the current no longer increases: then it just can't produce more current. A heavy-duty alternator might get there before running too hot, but most don't.

All you do with external regulator "tricks" is de-rate the alternator proportionally and it can made to work. It is just that you de-rate it throughout the entire rev range, not only when it is needed, and you need to de-rate for maximum engine revs (which is very unfortunate), or you might cook it one day if you suddenly gun the engine for some reason. It is an imperfect workaround.

When you limit using temperature, you can still overheat it because short of sensing the temperature in the windings themselves, there is a significant lag between what you measure and how hot the windings really get.
Using case or stator temperature as a measurement is more a protection measure than anything else. I wouldn't want to use it for control.

Some alternators have a thermal switch in the windings for the field circuit to prevent them from burning the stator out. They should all have one really, but it cuts out pretty high and if that switch was tripped all the time, it would also fail in relatively short order.

When you already have an alternator installed and integrated with the engine, it seems easier to control it properly than replacing it - if that was actually an available option.

The other issue with Balmars and the like is that - unless you configure zero absorption time (and then they won't charge properly) - they overcharge LFPs if you start the engine with a charged bank already. No current sensing means that they can't see that there is no need to charge any more and you get your 30 minutes or whatever you have programmed at maximum voltage. They were all designed for charging LAs, not LFPs.

[exMaggieDrum]

Originally Posted by OceanSeaSpray

The other issue with Balmars and the like is that - unless you configure zero absorption time (and then they won't charge properly) - they overcharge LFPs if you start the engine with a charged bank already. No current sensing means that they can't see that there is no need to charge any more and you get your 30 minutes or whatever you have programmed at maximum voltage. They were all designed for charging LAs, not LFPs.

The need to limit alternators to not get too hot has gotten the most attention. I am certainly going to put temperature sensors on my alternators and dial them back with belt manager or such.

I think you brought the quoted part above before and it got my attention. We end up motoring a lot starting from the marina until we actually get back to the marina where the boat stays on shore power. My SP charger and my solar panels will just completely shut off at a set voltage. But not the engine alternator with current regulators (as you note). Very light fluky winds are the norm here in the summer sailing season. The batteries start out fully charged and then could get overcharged by running the engine alternator as you note, even at a voltage well under the voltage-damage part, i.e. just from continuing to pump some current at the lower voltage, which you have said can also ruin the cells. I think people have suggested manually switching the external regulator on and off, which seems like it would work, but it is always nice to have something to prevent damage from forgetfulness - which I have more than enough of already.

[transmitterdan]

Balmar regulators are easy to shut down when an HVC event occurs. Removing 12V power from the regulator will shut down the alternator and stop charging the bank. So with proper HVC sensing there is no real point in reprogramming the regulator.

[OceanSeaSpray]

Originally Posted by transmitterdan

Balmar regulators are easy to shut down when an HVC event occurs. Removing 12V power from the regulator will shut down the alternator and stop charging the bank. So with proper HVC sensing there is no real point in reprogramming the regulator.

This is senseless. If you use HVC as your end of charge mechanism, you won't get any absorption and you have no protection left. HVC should never, ever occur on a properly set up system. Besides, HVC should not be a self-resetting condition, unless the battery is about to go flat, so you would be resetting it each time after charging... great.

Besides, in general terms cutting the field is ultimately safer because all these regulators get power from more than one terminal and this way you don't make assumptions about the internal failure mode of the regulator.

[SV THIRD DAY]

Originally Posted by OceanSeaSpray

This is senseless...

Perhaps to you, but it is a simple way to make sure that when your LiFePO4 batteries have reached their maximum safe voltage the alternator shuts down.

Where some see senselessness....others see simplicity.

Take my set-up for example.
My 1300W of solar pretty much keeps my 400AH LiFePO4 battery bank charged, so when I drop the mooring ball and motor over to the pump out dock, I don't want my alternator even turning on. So I hit the simple switch that shuts down the alt and reg. No programming. No computer code or Iphone app to mess with, just a $3 on/off switch and I'm done...safe...no worries about a BMS screwing me.

Now would I want my High Voltage cut off popping on and off as standard control for my reg/alternator...no that would be pretty foolish, but I don't see where anyone was suggesting that as a normal means of regulation, but more as a fail safe approach to ensure the alternator never overcharges the bank.

Talking "Absorption" with LiFePO4 is also a mistake in my opinion. There is no such things as Bulk, Absorption, float etc, so lets not confuse people with old school "Lead is Dead battery" terms.

[OceanSeaSpray]

Originally Posted by SV THIRD DAY

Perhaps to you, but it is a simple way to make sure that when your LiFePO4 batteries have reached their maximum safe voltage the alternator shuts down.

Where some see senselessness....others see simplicity.

Take my set-up for example.
My 1300W of solar pretty much keeps my 400AH LiFePO4 battery bank charged, so when I drop the mooring ball and motor over to the pump out dock, I don't want my alternator even turning on. So I hit the simple switch that shuts down the alt and reg. No programming. No computer code or Iphone app to mess with, just a $3 on/off switch and I'm done...safe...no worries about a BMS screwing me.

Now would I want my High Voltage cut off popping on and off as standard control for my reg/alternator...no that would be pretty foolish, but I don't see where anyone was suggesting that as a normal means of regulation, but more as a fail safe approach to ensure the alternator never overcharges the bank.

Talking "Absorption" with LiFePO4 is also a mistake in my opinion. There is no such things as Bulk, Absorption, float etc, so lets not confuse people with old school "Lead is Dead battery" terms.

transmitterdan said:

"So with proper HVC sensing there is no real point in reprogramming the regulator"

It is senseless.

You are talking about a manual switch: why not. It is just another one of those human-assisted control systems however, but the potential harm is limited.

When it comes to bulk and absorption, LFPs behave exactly like old lead-acids. They are just batteries, not the kind of contraptions with supernatural properties some like to believe.
The only difference is that when they are charged, that's it. You stop. They are easier to charge because they don't need all the multi-stage, equalisation, float, temperature compensation and what-not gimmicks of the LAs. Other than that, they charge the same, just faster and more efficiently.

[SV THIRD DAY]

Originally Posted by OceanSeaSpray

When it comes to bulk and absorption, LFPs behave exactly like old lead-acids.

Originally Posted by OceanSeaSpray

They are easier to charge because they don't need all the multi-stage, equalization, float, temperature compensation and what-not gimmicks of the LAs. Other than that, they charge the same, just faster and more efficiently.

Interesting...so in other words they are exactly the same...but then different at the same time. Thanks for that clarification.

Not what I've seen in my bank in here in realityville...but hey...what do I know anyway. I've only had my bank in full time liveaboard use for 2yrs now...heck I should be dead without a BMS anyway...move along...nothing to see here....

[fallingeggs]

Originally Posted by OceanSeaSpray

The other issue with Balmars and the like is that - unless you configure zero absorption time (and then they won't charge properly) - they overcharge LFPs if you start the engine with a charged bank already. No current sensing means that they can't see that there is no need to charge any more and you get your 30 minutes or whatever you have programmed at maximum voltage. They were all designed for charging LAs, not LFPs.

I suspected this was an issue, glad I'm not making it up. Maine Sail's posted setting on his MC-614 has him running bulk for an hour and absorption for two. This makes sense when he turns the alt off at 100% and discharges down to 80% DOD before turning the alt back on.

This method can be fine and dandy for most of the time, but may not be great living aboard when there isn't going to be great sun for a few days or when passage making (where safety might dictate being as fully charged as possible).

Running two programs depending upon the situation would be a huge pain in the butt. At least for the MC-614, is the alterative to put the minimum time of 6 (or 0?) minutes and have the "calculated bulk" and "calculated absorption" take their course? If 0 minimum is possible, I wonder how it can screw things up as you suggest. Can 6 or 12 minutes of "overcharging" be that bad?

[OceanSeaSpray]

Originally Posted by fallingeggs

I suspected this was an issue, glad I'm not making it up. Maine Sail's posted setting on his MC-614 has him running bulk for an hour and absorption for two. This makes sense when he turns the alt off at 100% and discharges down to 80% DOD before turning the alt back on.

This method can be fine and dandy for most of the time, but may not be great living aboard when there isn't going to be great sun for a few days or when passage making (where safety might dictate being as fully charged as possible).

Running two programs depending upon the situation would be a huge pain in the butt. At least for the MC-614, is the alterative to put the minimum time of 6 (or 0?) minutes and have the "calculated bulk" and "calculated absorption" take their course? If 0 minimum is possible, I wonder how it can screw things up as you suggest. Can 6 or 12 minutes of "overcharging" be that bad?

The first time I set up a LFP system with a common external alternator controller (it was an Ample Power NS-2), I configured the alternator to bulk charge to 14.2V and shut off (no absorption). It was a conservative starting point, but it ended up so short of capacity on the first discharge cycles that it wasn't even funny. Absorption is essential.

With alternators, I find that after about 30-35 minutes of absorption at 14.0V, the current has tapered down to a pretty small value and I stop. It leave a little capacity on the table, but I am happy with that. I am not seeking 100% SOC.
If you have a huge bank and a small alternator, you might find that absorption is happening a little more quickly. If I have to configure an absorption time because the regulator is dumb (i.e. it has no idea of the actual charging current, standard scenario), I set it to that value.

Bulk ends when the voltage has reached 14.0V. I don't do crazy stuff trying to force current in at higher voltages before stepping back down to shave off a few minutes of charging time, it is senseless.

If the engine is started with the alternator full, it will hold 14.0V for 30 minutes with no necessity. It is not great, but as long as it doesn't happen too often... it is not dangerous... live with it?
If the mode of operation of the boat suggests that it is going to happen all the time, then you can add a manual switch or configure for under-charging by lowering the absorption voltage/absorption time. Neither are great.

When you set up a LFP system using LA charging gear, it is typically not going to do the right things at least some of the time. You can tune it more or less to suit a type of on-board power usage pattern, but deviate from this and it won't be good for the battery or require manual intervention etc.

I haven't been too happy about all that, so I have been chipping away at the problem. This is why I built an alternator controller that senses the current last year, it solved that issue. Getting the BMS to manage the solar charging solved those problems as well and we are going to do something similar about a wind generator in the next week or so.

Now, as to knowing what kind of fudging the Balmar is going to do if you set it on "calculate"... it has no idea of the battery charging current, which is the most essential piece of info with battery voltage to determine when a battery is full and there is no substitute for it.
It was only ever intended for charging LAs, those can be overcharged safely and must be kept fully charged as much as possible, so you can imagine the result!
If I was stuck with a Balmar, I would probably set the absorption voltage/time to something I can live with and accept that it won't be optimal at times. On some boats, the alternator is not a primary charging source at all and if it is of no consequence if it can't take the battery to a high SOC. In this case, I would just dial back the absorption time.

[Maine Sail]

Originally Posted by fallingeggs

I suspected this was an issue, glad I'm not making it up. Maine Sail's posted setting on his MC-614 has him running bulk for an hour and absorption for two. This makes sense when he turns the alt off at 100% and discharges down to 80% DOD before turning the alt back on.

This method can be fine and dandy for most of the time, but may not be great living aboard when there isn't going to be great sun for a few days or when passage making (where safety might dictate being as fully charged as possible).

Running two programs depending upon the situation would be a huge pain in the butt. At least for the MC-614, is the alterative to put the minimum time of 6 (or 0?) minutes and have the "calculated bulk" and "calculated absorption" take their course? If 0 minimum is possible, I wonder how it can screw things up as you suggest. Can 6 or 12 minutes of "overcharging" be that bad?

That was a very old post and I have changed settings a few times over the years. (sorry CF does not let us edit posts, or keep them current, unless you physically ask a moderator to do it).

For me that worked fine because I manually shut the reg down at voltage & current. On a long windless delivery with some friends I changed it to what you see below rather than explain to them why and when to shut down the alt..

I don't think I have changed it since that reprogramming. I fall short occasionally on bulk & absorb time but not by much. Bumping bulk from 13.9V to 14.0V helped a little bit on that. A few minutes with the magnet and this is where I left it...


bv/Bulk Voltage = 14.0V
b1c/Bulk Time = 24 Minutes
Av/Absorption Voltage = 13.8V
A1c/Absorption Time = 12 Minutes

With 12 minutes fixed the calculated absorption usually drops to float nearly immediately after the 12 minutes because the current drop is so rapid. Ideally I may go back in and tweak the absorb time to 24 minutes or so and for my bank, and charge rate, that is probably just about where it winds up at less than 10A at 13.8V...

Occasionally I will just re-boot the reg, if I really need the battery capacity, but usually I don't. At the setting above I am close to or below 10A at 13.8V by the time the 40 +/- odd minutes are up.

With the original absorb settings at 13.8V you're not going to do any major damage to the LFP bank even if you started it when it was pretty full. You'll just likely get down towards 1-2A acceptance at 13.8V. Plenty of safety margin at 13.8V absorb.

[transmitterdan]

Originally Posted by OceanSeaSpray

This is senseless. If you use HVC as your end of charge mechanism, you won't get any absorption and you have no protection left. HVC should never, ever occur on a properly set up system. Besides, HVC should not be a self-resetting condition, unless the battery is about to go flat, so you would be resetting it each time after charging... great.

Besides, in general terms cutting the field is ultimately safer because all these regulators get power from more than one terminal and this way you don't make assumptions about the internal failure mode of the regulator.

I don't see much value in absorption phase for LiFePo banks. The small additional capacity isn't worth the complexity. How much additional charge capacity do you think comes from absorption phase charge of a house bank?

Why is there no protection left? If all charging sources are off what protection is required?

[mrm]

Originally Posted by transmitterdan

Why is there no protection left? If all charging sources are off what protection is required?

Dan,

Given the cost of a LiFePO bank, a two tier protection seems sensible from my engineer point of view. As we all know, systems fail ultimately. If, for example, an alternator field disable component fails for whatever reason and there is no second tier of protection, a big cost item is at risk.

So, while your statement is true, in reality not all charging sources may be off when we want them off.

[transmitterdan]

Originally Posted by mrm

Dan,

Given the cost of a LiFePO bank, a two tier protection seems sensible from my engineer point of view. As we all know, all systems fail ultimately. If, for example, an alternator field disable component fails for whatever reason and there is no second tier of protection, a big cost item is at risk.

So, while your statement is true, in reality not all charging sources may be off when we want them off.

If one removes the field voltage then the alternator cannot produce power. And why can't there be more than one HVC circuit if redundancy is desired?

I'm not understanding the desire to reach 100.0% state of charge. With LA banks absorption is important to increase battery life but with LiFePo it doesn't seem necessary.

A simple system could charge between two set point voltages with LVC and HVC protection beyond those two voltage levels. Interfacing such a system to a traditional LA alternator regulator is not complex. If you break one wire then the alternator shuts down. Reconnect and it starts up again. A system like this will only use about the middle 80% (10-90) of bank capacity but it's hard to get any simpler or more reliable.