LFP FLA hybrid

[Delfin]

Originally Posted by rgleason

Delfin wrote:
I do that with the BMS and an ACR. When the LFP is "full", as in CAR < 5% of C, the BMS gets switched off and this isolates the LFP bank from charge sources. I then use a Blue Seas ACR to manually combine the LA starter bank and LFP bank, which, since the LFP bank is offline means all charge current goes to the LA bank.

This seems practical. Do you trust the BMS to shut off properly? I assume so because your BMS is programmed by you.
How do you know it is shut off, so that you can manually switch the ACR? Also what switch do you use for the LFP BMS.

The BMS doesn't shut off automatically unless a high or low voltage condition exists. It manually disconnects and isolates the bank by a disconnect switch on the BMS. The trigger to disconnect is pulled by me, since the time at which the LFP bank needs to be disconnected because it is full is easily predicted as a point in time. Set an alarm, note the CAR and take the bank off line. IMO most of the perceived complexity of managing LFP banks stems from the desire to use a completely automated charging system a la an FLA bank. Again, IMO, LFP should be treated like a fuel tank - fill them, then turn off the spigot.

[evm1024]

OK - you all got me thinking about this. I'll have to go back and re-read the thread and see if I want to reconfigure the charging setup.

Having the charging sources programmed for lead acid (modified as needed) and using a combiner relay to parallel the LiFePO4 bank in when 1) there is a charging voltage, 2) the LiFePO4 bank is at or below the charge trip point makes a lot of sense.

When the LiFePO4 bank is "charged" the combiner breaks the parallel and the charge regulators go on their merry LA way charging the start battery fully.

A BMS monitors the LiFePO4 house bank independently from the charge source and isolated it when too high or too low.

[Delfin]

Originally Posted by evm1024

OK - you all got me thinking about this. I'll have to go back and re-read the thread and see if I want to reconfigure the charging setup.

Having the charging sources programmed for lead acid (modified as needed) and using a combiner relay to parallel the LiFePO4 bank in when 1) there is a charging voltage, 2) the LiFePO4 bank is at or below the charge trip point makes a lot of sense.

When the LiFePO4 bank is "charged" the combiner breaks the parallel and the charge regulators go on their merry LA way charging the start battery fully.

A BMS monitors the LiFePO4 house bank independently from the charge source and isolated it when too high or too low.

Other ACRs may work differently than mine, but for the Blue Seas I have, you wire the house bank (LFP) to the default destination for charge current. When that bank reaches a fixed setpoint voltage, a latching relay closes and combines it with the starter bank. So the combiner, in my case anyway, doesn't "break the parallel" connection when the LFP bank reaches a setpoint voltage, it makes the connection.

I use the BMS to isolate the LFP bank when full, and the ACR in that circumstance is "Manually combined", resulting in charge current going to the LA bank. This would be the normal state when motoring for quite awhile - the LFP bank gets filled, then isolated and the continuing output of the alternator then goes to the starter bank.

Make sense?

[evm1024]

Originally Posted by Delfin

Other ACRs may work differently than mine, but for the Blue Seas I have, you wire the house bank (LFP) to the default destination for charge current. When that bank reaches a fixed setpoint voltage, a latching relay closes and combines it with the starter bank. So the combiner, in my case anyway, doesn't "break the parallel" connection when the LFP bank reaches a setpoint voltage, it makes the connection.

I use the BMS to isolate the LFP bank when full, and the ACR in that circumstance is "Manually combined", resulting in charge current going to the LA bank. This would be the normal state when motoring for quite awhile - the LFP bank gets filled, then isolated and the continuing output of the alternator then goes to the starter bank.

Make sense?

Yes, quite, thanks. COTS stuff is good but a kickstart project would be nice too for a "full" system. (voltage, tail current, etc)

[rgleason]

I get concerned about the voltage setting for FLA bulk charging at 14.7 to 14.8v. This is way too much for safe LFP charging. That is why the alt regulator needs to be a dual mode regulator. Mode 1 safe settings for LFP, when FLA can also be charged, and Mode 2 for when LFP is not connected and FLA can be fully charged past bulk and abosorprion.

I think it is a programming thing. When that ACR is closed, the regulator must be set to Mode 1.

[Delfin]

Originally Posted by rgleason

I get concerned about the voltage setting for FLA bulk charging at 14.7 to 14.8v. This is way too much for safe LFP charging. That is why the alt regulator needs to be a dual mode regulator. Mode 1 safe settings for LFP, when FLA can also be charged, and Mode 2 for when LFP is not connected and FLA can be fully charged past bulk and abosorprion.

I think it is a programming thing. When that ACR is closed, the regulator must be set to Mode 1.

It is not too high for LFP charging unless you expect to set that as bulk and walk away. Before the battery reaches that voltage, the CAR will drop off to near zero and so the charge source can be disconnected from the LFP. As I have repeated more than once and as MaineSail has noted, it is the duration at a given voltage that matters when charging LFP, not the voltage.

[newhaul]

Why is everyone trying to make this so complex. Here is an idea that would work great for many if not most installs.
1) a small sacrificial 12v Fla battery in the loop with the Lfp as a safety for the alternator.
2) run a 10 amp battery charger to the Fla start battery off of a small (150 watt)inverter.
3) wire up the inverter to draw power from the Lfp bank via a relay that draws control power off of the ignition switch.
Cost of this entire setup will be less than a single acr .

If this wouldn't work someone please explain why not.

[CatNewBee]

Originally Posted by rgleason

I get concerned about the voltage setting for FLA bulk charging at 14.7 to 14.8v. This is way too much for safe LFP charging. That is why the alt regulator needs to be a dual mode regulator. Mode 1 safe settings for LFP, when FLA can also be charged, and Mode 2 for when LFP is not connected and FLA can be fully charged past bulk and abosorprion.

I think it is a programming thing. When that ACR is closed, the regulator must be set to Mode 1.

this is not an issue, the voltage drops immediately when the LFP are connected due to low resistance of the battery, the charger cannot keep up the voltage, when the voltage rises above safe values because the LFP cells are fully charged, the BMS would disconnect the source to the battery.

[rgleason]

CNB thanks, I guess that provided the alternator is 1C or less that the voltage will drop down and the BMS must be very reliable for the disconnect. I'd like the disconnect to be automatic to avoid issues related to being distracted.

Newhaul, that sounds possible, but we are on a mooring. I suppose the 10a charger could be PV.

[newhaul]

Originally Posted by rgleason

CNB thanks, I guess that provided the alternator is 1C or less that the voltage will drop down and the BMS must be very reliable for the disconnect. I'd like the disconnect to be automatic to avoid issues related to being distracted.

Newhaul, that sounds possible, but we are on a mooring. I suppose the 10a charger could be PV.

on a mooring you just need a 50 watt flex panel, small PWM controller and alligator clips to hook to the Fla start battery when you leave the boat .
I only said a 10 amp charger because that seem to be the smallest commonly available marine charger.

[rgleason]

50a charger would be good on a mooring and would help while cruising, but cruising with daily charging needs, with my old alternator requires too much engine time.


I do have old 14 watt and 30 watt panels which do work at the mooring for FLA.

[newhaul]

Originally Posted by rgleason

50a charger would be good on a mooring and would help while cruising, but cruising with daily charging needs, with my old alternator requires too much engine time.


I do have old 14 watt and 30 watt panels which do work at the mooring for FLA.

then you have the charging of start battery covered. .
Just run a 10 amp charger off of an inverter when charging your Lfp bank .
It doesn't take much juice to start an engine.
Depending on engine likely less than 5 to 7 ah worth of power.

[evm1024]

OK, I went back and read the thread and see that I was posting here till (oh, um, ah) something happened to make me stop posting. I am still interested in the topic.

So let's see if I can throw out some thoughts.

1st, start batteries in cars last quite a while and don't have any fancy charging profiles. Just blast some amps to the start battery and then quickly derate based on temp or state of charge. A small number of amp hours are needed to be replaced.

2nd, 3 stage charging (or 4) is based on overcoming the shortfalls of automotive regulators when replacing lots of AH in a house bank. I'm not sure how much more life you would get from a start battery that was charged with a 3 stage regulator when the battery was in start service.

3rd, LiFePO4 batteries do not need 3 stage regulation. They appear to do fine with 2 stage and even better with 1 stage then disconnect charging.

4th, Typical LiFePO4 voltage at the desired max SOC (13.8 or 13.9 volts) is within the bulk voltage range for typical Lead Acid start batteries. (Hey, let's not quibble the words, just understand the desired meaning)

5th, LiFePO4 BMS are only for bank protection. They cut the charge when over voltage and cut the loads when under voltage.

With these thoughts in mind I'm thinking that I might rewire my boat such that the alternator feeds the AGM start battery directly and reporgram the MC-612 to be more like a bulk/float regulator for AGM voltages.

Then put a contactor in that joins the start AGM to the house LiFePO4 bank.

This of course it where the rub comes in. The contactor needs to have a bit of "smarts" in it.

The contactor needs to close when there is a good charge voltage, disconnect when the LiFePO4 reaches (what you call ) full. Disconnect when the charge voltage is low so as to prevent the LiFePO4 bank from "charging" the AGM start battery and have a number of failsafes worked in. How about measuring the charge into the house bank and disconnect when the LiFePO4 bank reaches a target voltage OR when 80% of the bank capacity is restored (as a failsafe....).

I've been banging around to come up with a state machine for the contactor. Lots of state transitions and reporting modes to think about.

Just thinking out loud, nothing to see here, move along.

[CatNewBee]

start batteries are very shallow discharged, they have plenty of capacity purely to deliver a decent current for the cranking. They are immediately recharged when the engine runs, you also will not notice any capacity loss until they fail to crank.

it is a completely different use pattern to a deep cycle house battery, where the main goal is cycles and capacity in contrast to high Amps for some seconds in cold conditions. The plates of a deep cycle battery are much thicker and less structured to allow more cycles and deeper discharge.

So are the requirements for charging different, a 3 stage charger is great for deep cycle batteries, a start battery staying at 90...100% SOC is happy with absorption / float to get full again, and the degradation of cycling is not important to her.


A good BMS keeps the cells balanced, keeps an eye on the temperature and on the resistance, it is not only voltage protection on the upper and lower end.

[john61ct]

See Cpt Pat's logic, link #251 that resurrected this thread. Paraphrasing, feel free anyone to refine if too rough.

Assume charge sources are "lead side" A. LFP circuit B, combiner between.

No nitpicking over V setpoints, adjust to suit your taste, assume delay / hysteresis to avoid chatter.

State 1 no charging, low volts say under 13.5V, A and B are isolated.

Charging starts on the A side circuit, when it gets up to 13.55V, State 2, combiner closes, LFP starts charging.

Voltage keeps climbing, when it gets to 13.85V, LFP is Full (enough), that's State 3, combiner opens (isolates) only lead bank(s) on A circuit keep charging.

When charge source stops, returns to State 1.