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

[Hoopla]

Originally Posted by alctel

What charger/inverter are you all using?

That is the "how long is a piece of string question". Do you have a single or dual bus LFP system, what sort of BMS do you have, do you charge more than using the inverter or the reverse??

Putting all that aside do not buy a inverter/charger that can't be programmed for charging LFP at a set point suggested in this thread (ie max 14.2/28.4 pack CVCC )and does not have some ability for this charger/inverter to be turned off in the case of a high and/or low voltage event triggered by your BMS and a relay. Putting a charger/inverter in a LFP system no matter the make involves some risk as it will always be connected to the battery bank in association with say a preferred dual buss system. So the name of the game is to reduce that risk to the health of the LFP bank as well as to the charger/inverter itself by sudden opening of high current contactors.

One example is Victron's current charger/inverters (and even older versions with a firmware update) can be directed by a BMS/BMS high current contactors/ to shut down at both high and low voltage events. This is accomplished by Victrons offering of potential free auxilliary contacts tied to a third party BMS and that can be used for protecting LFP systems other than their own proprietary LFP offerings.

[pacific_voyager]

Originally Posted by Hoopla

... do not buy a inverter/charger that can't be programmed for charging LFP at a set point suggested in this thread (ie max 14.2/28.4 pack CVCC )and does not have some ability for this charger/inverter to be turned off in the case of a high and/or low voltage event triggered by your BMS and a relay. Putting a charger/inverter in a LFP system no matter the make involves some risk as it will always be connected to the battery bank in association with say a preferred dual buss system. So the name of the game is to reduce that risk to the health of the LFP bank as well as to the charger/inverter itself by sudden opening of high current contactors.

One example is Victron's current charger/inverters (and even older versions with a firmware update) can be directed by a BMS/BMS high current contactors/ to shut down at both high and low voltage events. This is accomplished by Victrons offering of potential free auxilliary contacts tied to a third party BMS and that can be used for protecting LFP systems other than their own proprietary LFP offerings.

I have been puzzling over this same question. How to use an inverter/charger that is not designed for use with LiFePO4 and/or does not have a reliable voltage sensing system?

I have TWO Outback inverter chargers: 110V and 220-240V. When I left the USA in 2008 and ended up in Australia it was easier to add a second Outback inverter/charger than mess around with transformers and mismatched hertz. I use 110vac from a Honda 2000i and 240vac when plugged in to shore power.

Now that I want to switch to LiFePO4 batteries (from AGM) I need to control the charging function of the either of the two inverter/chargers.

I.e. I want my management system to switch the charging function of the inverter/chargers OFF or ON when needed. Outback tech support recommended setting the thresholds of the absorption phase accordingly, but I was afraid that would not be adequate as the voltage sensing is not very accurate.

Mainesail wrote “Voltage sensing at the battery terminals is critical to FAST CHARGING PERFORMANCE. If you use a generator to power an AC charger proper voltage sensing means less generator run time. If a charger or inverter charger does not offer you this option BUY ONE THAT DOES!”

He recommended, for the Outback unit, adding an Outback FlexNET DC volt sensing unit to the system (I might need two Flexnet DC units and this requires a HUB-4 as well). Total cost around $500.

Then I had this idea! What about controlling the input to the chargers? I.e. when shore power is on and the upper threshold SOC is reached on the batteries, bypass the AC around the inverter/charger. If shore power is on, and SOC is reduced to a lower (programmable) SOC, restore the AC input to the inverter/charger. If shore power is not present, do nothing. This would also work for a standalone charger. It could be done with a few relays (mechanical or solid state).

Any criticisms or comments?

[john61ct]

Personally I would invest in a good programmable DCDC "battery to battery" charger, Sterling or ProMariner.

Put the more basic charge sources directly onto a reserve/starter lead bank, and then charge via the B2B from there.

Then also don't have to invest in changing alt VRs, wind/solar, whatever.

All the LFP charging gets handled by one known-good device.

[Hoopla]

Originally Posted by pacific_voyager

I have been puzzling over this same question. How to use an inverter/charger that is not designed for use with LiFePO4 and/or does not have a reliable voltage sensing system?

I have TWO Outback inverter chargers: 110V and 220-240V. When I left the USA in 2008 and ended up in Australia it was easier to add a second Outback inverter/charger than mess around with transformers and mismatched hertz. I use 110vac from a Honda 2000i and 240vac when plugged in to shore power.

Now that I want to switch to LiFePO4 batteries (from AGM) I need to control the charging function of the either of the two inverter/chargers.

I.e. I want my management system to switch the charging function of the inverter/chargers OFF or ON when needed. Outback tech support recommended setting the thresholds of the absorption phase accordingly, but I was afraid that would not be adequate as the voltage sensing is not very accurate.

Mainesail wrote “Voltage sensing at the battery terminals is critical to FAST CHARGING PERFORMANCE. If you use a generator to power an AC charger proper voltage sensing means less generator run time. If a charger or inverter charger does not offer you this option BUY ONE THAT DOES!”

He recommended, for the Outback unit, adding an Outback FlexNET DC volt sensing unit to the system (I might need two Flexnet DC units and this requires a HUB-4 as well). Total cost around $500.

Then I had this idea! What about controlling the input to the chargers? I.e. when shore power is on and the upper threshold SOC is reached on the batteries, bypass the AC around the inverter/charger. If shore power is on, and SOC is reduced to a lower (programmable) SOC, restore the AC input to the inverter/charger. If shore power is not present, do nothing. This would also work for a standalone charger. It could be done with a few relays (mechanical or solid state).

Any criticisms or comments?

When Mainesail wrote “Voltage sensing at the battery terminals is critical to fast charging performance" was he talking about LFP or LA?

I would not have anything automatically turning a charger and or inverters on or off for a LFP bank relying on SOC or pack voltage or even using the high cell voltage protection of the BMS to automaticaly turn a charger off as a day to day function. That would mean your LFP charge set point is far too high.

For inverters you also need the means to isolate high AC load devices like hotwater and standalone chargers. Most good inverter/chargers have AC pass through capability so their inverter function is always off when used to charge and for some the inverters function can be used to supplement shore or generator AC if it is insufficient. What your contemplating sounds like an automated LFP murder nightmare.

BTW having inverter/chargers automatically turn off in the unusual case of a high or low cell voltage event instead of the high current charge or discharge relay suddenly opening with them remaining/still on is a different kettle of fish.

[alctel]

Originally Posted by Hoopla

That is the "how long is a piece of string question". Do you have a single or dual bus LFP system, what sort of BMS do you have, do you charge more than using the inverter or the reverse??

Putting all that aside do not buy a inverter/charger that can't be programmed for charging LFP at a set point suggested in this thread (ie max 14.2/28.4 pack CVCC )and does not have some ability for this charger/inverter to be turned off in the case of a high and/or low voltage event triggered by your BMS and a relay. Putting a charger/inverter in a LFP system no matter the make involves some risk as it will always be connected to the battery bank in association with say a preferred dual buss system. So the name of the game is to reduce that risk to the health of the LFP bank as well as to the charger/inverter itself by sudden opening of high current contactors.

One example is Victron's current charger/inverters (and even older versions with a firmware update) can be directed by a BMS/BMS high current contactors/ to shut down at both high and low voltage events. This is accomplished by Victrons offering of potential free auxilliary contacts tied to a third party BMS and that can be used for protecting LFP systems other than their own proprietary LFP offerings.

I get that, but having trouble finding a charger or MPPT solar charge controller that even has an option to not have a float - they all seem to have a float mode, which I don't really want.

I don't really want to use a BMS for controlling the on/off of my chargers in day to day use (the BMS will be set up as an alarm/safety shutdown system - it will turn off chargers etc in that configuration)

[john61ct]

Originally Posted by alctel

I get that, but having trouble finding a charger or MPPT solar charge controller that even has an option to not have a float - they all seem to have a float mode, which I don't really want.

I don't really want to use a BMS for controlling the on/off of my chargers in day to day use (the BMS will be set up as an alarm/safety shutdown system - it will turn off chargers etc in that configuration)

That's exactly right, "bottom layer" of protection should be backup, belt & suspenders for when the daily-use stuff fails.

On the SC, next best perhaps is setting float to something low like <13.0-13.2 no actual evidence that will significantly reduce LFP longevity.

And worth trying maybe an LVD before the bank set to cutoff a little higher than that?

[Hoopla]

If you have to pick a float voltage just use same set point as absorption. If you can't something like 13.1v as john61 says is not the end of the world. 13.1v is actually same as pre-programmed for LFP in some chargers and regulators.

[john61ct]

Originally Posted by Hoopla

If you have to pick a float voltage just use same set point as absorption.

Rest of it's fine but my Vabs of 13.8-9 would not be suitable for Float, in fact definitely and significantly reduce bank lifespan.

Point is to in effect eliminate the float by setting the voltage lower than resting (in my case 13.6).

If a high-current load makes voltage drop below the float setpoint, no problem, bank's no longer pushing that full shoulder anyway.

[pacific_voyager]

Perhaps I am confusing nomenclature: BMS should be "charging source control". The number of posts is approaching 6000 and it would be nice if there were a synthesis somewhere. The best seems to be this. I am just trying to figure out what will work for a 600-700 AH bank without having to go crazy with complexity and expense.

So far it appears to me that the following is a subset of what is good for DIY LiFePO4 vessel battery banks:

1. The batteries can take as much current as you can safely put into them. This reduces expense (diesel/petrol savings) by lowering generator runtime.

2. Each current source is, usually, controlled by a device (alternator regulator, solar charge controller, wind power charge controller, mains power battery charger or inverter/charger). Older units assume a 3-stage charge cycle: bulk/absorption/float which is unsuitable to LiFEPO4 batteries. These units operate on a CC (constant current) in the bulk mode then on CV (constant voltage) in the other phases. The better controllers have custom settings which allow absorption and float to be effectively disabled. The newer devices have “lithium” settings which might not be ideal.

3. I am continually reading that these batteries are not LA and must be handled as a totally new system, yet most practical solutions appear to be voltage-driven and required staying within a safe zone in the SOC. I.e. below 95% SOC and above 25% SOC as indicated... by voltage. Different users seem to have variations on what is acceptable or safe. Is that right?

4. The "BMS" may or may not have top balancing. It is arguable whether, in a properly initially balanced bank, automatic top balancing is needed. There are good arguments for bottom balancing every year or three and not having cell balancers. It may be simpler than we think.

5. What is more important for a "BMS" (or "safety net management") is the ability to disconnect the charging source at a high voltage threshold and then to disconnect the loads at a low voltage threshold. The general thinking appears, to me, that this is your safety net. Not what I would call a “management system” but more like an emergency disconnect.

6. Ideally the loads and charging sources are on different circuits; but I don't see a huge problem with them being on the same circuit. If the SOC of the bank declines to the LVP threshold you have to start charging and this brings the batteries back on line. An inverter/charger is, by default, using the same cable for charging and inverting.

7. Inside the "safety net" range we need to control all incoming charging sources to stay below a safe voltage/SOC level. If, for some reason, the charging source fails to stop charging and the “safety net” HVP cuts the circuit this could destroy an operating alternator. Alternators need to have the regulator shut down or the exciter wire disconnected. I am not sure what an HV emergency cut-off would do to an operating battery charger. Solar isn’t a problem. .

8. Older inverter/chargers are hard to control because the battery voltage sensing is taken off the large DC cables that either power the inverter or deliver the charge. Voltage at the inverter is not necessarily voltage at the battery. If I wanted to be able to control the charging function externally while on shore power, I could bypass the AC input around the inverter charger. Despite earlier criticism, this would NOT be a problem. This is done all the time. Unplugging shore power from an inverter will shift it into INVERT mode if inverter mode is "ON". We charge with a small Honda generator and it when I turn it off (or it runs out of gas), it stops charging. Large power loads like the water heater can have a solid state relay (mine already does) to allow functioning ONLY when on shore power, which for me is rare. I disagree that this is a bad way to control the charging function on an inverter charter; it is actually a normal part of the design. At this point, however, I think a new, separate, battery charger (40A perhaps) with a Lithium program could be used. But, like I said, I’m not sure I trust the battery charger manufacturers lithium profiles.

Sorry for the long post.
Cheers,

Jeff

[Hoopla]

Originally Posted by john61ct

Rest of it's fine but my Vabs of 13.8-9 would not be suitable for Float, in fact definitely and significantly reduce bank lifespan.

Point is to in effect eliminate the float by setting the voltage lower than resting (in my case 13.6).

If a high-current load makes voltage drop below the float setpoint, no problem, bank's no longer pushing that full shoulder anyway.

john setting float to same as your set voltage is simply to keep charging at CV so it doesn't quit to early. I don't like the concept of using a low float at all in terms of max cell life/capacity and prefer to charge, stop, discharge then charge again.

[Maine Sail]

Originally Posted by Hoopla

I don't like the concept of using a low float at all in terms of max cell life/capacity and prefer to charge, stop, discharge then charge again.

We have a winner!!! See, those 5600 previous posts do work!

[john61ct]

Problem is so few even high-end charge sources allow for "just stop".

Dropping float down to 13.2 or lower? is just a workaround, not ideal but it seems necessary in many contexts.

My idea is putting all "less ideal" charge sources directly into a lead-based starter/reserve bank and from there "filtered" through a programmable DCDC charger like ProMariner / Sterling Power's newer high end battery-to-battery units.

That is, if even these allow for "stop charging" no float. Do they?

[john61ct]

Originally Posted by Hoopla

john setting float to same as your set voltage is simply to keep charging at CV so it doesn't quit to early.

Problem is most chargers float forever. We were talking about charge sources that don't allow "just stop".

To me float means "charging is finished, bank is full". All my intended charge sources DO have full programmability, sufficient to ensure Absorption doesn't terminate too early*

But I'm still looking for an Alt VR and solar controller that has "just stop".

* note this worry about stopping "too early" is actually a holdover from Lead thinking. LFP has no issue with stopping a bit early, in fact much better than keeping Vabs going too LONG.

[Hoopla]

Originally Posted by Maine Sail

We have a winner!!! See, those 5600 previous posts do work!

I needed new steak knives too :-)

Maine Sail did you see my post on previous page re your remote rectifier? Looks very nice.

I am assuming you have just connected the 3 AC phases to the stator and left everything else as is incl the onboard rectifier (but no DC out) and external field connection?

Cheers

[CGirvan]

I have a quick question as we seem to be on the subject of chargers. I haven't switched to LiFePo yet but am planning on doing so when the current set of batteries die. I was thinking of buying a benchtop power supply instead of a dedicated charger specifically the Volteq HY1550EX.

My charging system will be primarily solar, and alternator when motoring and running the Honda 2000 when the first two aren't enough. My thinking is that the power supply will be handy when initially balancing the cells and any necessary subsequent balancing, also when using the charger I will be monitoring what's going on and I'm pretty much just bulk charging.

Anybody see any problems with this scenario? Does the honda put out clean enough 120v for the power supply? Is there an issue running the full 50amps for an hour at a time or would I be better to cut back to 45?

Thanks for your help

Colin