Yet another charging Lithium/Lead hybrid setup thread

[WE9V]

Okay, lots of threads on this already, I know. A lot of good ideas, but as one person said, it's all dependent on your system. So instead of a generic question about "what's best", let me pose a specific question: Based on my current configuration, and my goals, what would you suggest that I do? This is a long and detailed post, but I hope that it shows that I did some homework myself instead of asking you to do it for me.

Current system, as purchased new-to-me a ~year ago:

• All 12V system, no desire to change
• 400AH AGM House about to be changed to 2x 230AH LiFePO4 drop-ins, each with 200A BMS
• No warning signal prior to drop-in LFP BMSs shutting down
• Balmar 170A alternator, serpentine belt, Balmar MC-614 external regulator, regulator has alternator and battery temperature sensors present
• 650W Solar, 150/60 MPPT charging House bank
• Separate AGM Start and AGM windlass batteries
• Old Balmar DigitalDuo Chargers, one from House to Start, one from House to Windlass. These DDCs are like 30A ACRs, but when load exceeds 30A, it completely shuts down, not limits to 30A. Occasionally fires back up to see if load <30A. Not a multi-stage charge, but like an ACR.
• Old Victron Centaur Shore Charger, up to 100A, 3 outputs
• 5kW diesel genset that can also power shore charger
• Victron BMV-712 Smart Battery Monitor
• Windlass is 1200W model, but with ~30' run (1/0 or 2/0, don't remember), as battery is under companionway steps. Will not relocate battery, but that's good as it's close to the house and starter batteries.
• For the sake of this discussion, let's say that the boat is equal parts dock queen, ICW-motoring, and at anchorage. In other words, no priority given to any.

Goals/requirements:

1. Lowest additional cost
2. Reasonably low added complexity
3. Charge LiFePO4 at highest possible rate (priority to LFP). 30A DC-to-DC charger to LFP not acceptable / highly undesirable.
4. Protect alternator from voltage spikes due to LFP disconnect
5. Ability to charge all 3 batteries from all charging sources (alternator, solar, shore, genset). Optimal charging for AGMs not a requirement / using LFP profile acceptable.
6. Bonus: Alternator helping windlass
7. Bonus: LFP helping windlass

A recent call to Balmar suggests that the DigitalDuo Charge may not be optimal for charging the windlass battery because it would shut down if current exceeds 30A. So it likely shuts down if the windlass is used (>30A) and/or if the windlass battery has been deeply discharged. I believe this means that the high-amperage alternator is doing nothing for windlass use and/or charging, but due to the relatively low output of solar (I've seen a max of 580W output, to charge the discharged house AGM), it's unlikely to see >30A into the windlass AGM. (Would need the house bank completely charged, the windlass battery discharged, and full sun. But the windlass battery would get SOME current, likely less than 30A, before the house bank is fully charged or as the sun first comes up.) So what I'm saying is that the DDC potentially does nothing during alternator charging, but is likely adequate for solar charging.

Potential solutions:

1. While there's a Renogy 60A DC-DC charger (about $200 on Amazon), I'm not terribly keen on Renogy. 60A charging to Lithium seems minimal, considering 170A alternator. Really dislike the thought of a Victron 30A model ($250), and really don't like the idea of two of them ($500). Violates goal #1 and/or #3.
2. DC-to-DC charger from Lithium to windlass AGM. This does not meet goal #4, protecting alternator from load dump. Ruled out.
3. Add AroFET 200 from alternator to LFP and windlass. At ~$150, it seems to meet goals 1-6.
4. Add contactor or BlueSea ACR between windlass and LFP. These 120A ACRs are about $90-95 on Amazon.
   • Due to LFP's resting voltage of ~13.3v, per BlueSea's suggestion, would need to add tiny relay in ground signal that would close when ignition is on.
   • Bonus: could also use BMV-712 relay in series to stop LFP charging when SoC >90% or voltage >14.2V (setpoints for example).
   • I believe this may satisfy goals 1-7, with #2 slightly more complex due to the additional relay logic but has added functionality/features. (Goal #7 is met only if LFP isn't >90% charged and/or needs ignition on.)
   • Would need to change wiring such that windlass battery has the direct connection to the alternator, and would need to limit the MC-614 belt manager to the 120A limit of the ACR. An alternative would be to get a more expensive ignition-protected contactor (Blue Sea 150A $133, 250A $175). I think I'm okay with the 120A limit. Far better than 30A DC-DC.

So, what do you think? Solution #3, #4, modifications to #3 or 4, or something completely different?

Bonus question: My AGM windlass battery has died after 7 years. Currently a 4D size. What would you replace it with, especially based on your suggested solution? Going with the lowest additional cost, an AGM Group 27 or 31 would be great instead of 4D. Apparently the alternator nor the 30A DDC has helped the windlass while in use...it was all battery.

[Tin Tin]

I assume your alternator is LifePo4 friendly, if not get a regulator that is. Alternator just charges house bank. House bank charges the AGM batteries with a DC DC 18A charger each - set the input cutoff for 13.5V - and cut in at 14V your AGMs will be charged to float in a short space of time every time the LIFepO4 bank recieves 14+ volts from any source. Generator and shore power charges house bank only with say a 100A Centaur charger (if it has LifePO4 profile), if not replace. Alternator charges the house bank only with 170A. Solar with MPPT to House bank only.


So it would seem worst case to buy then would be 2 * 18A victron DC DC chargers, a LifePO4 friendly regulator for the Balmar and a Lifepo4 charger for Gen and shore power.

[WE9V]

Originally Posted by Tin Tin

I assume your alternator is LifePo4 friendly, if not get a regulator that is.

Balmar MC-614 external regulator can be programmed for LFP and is already in system components listed above.

Originally Posted by Tin Tin

Alternator just charges house bank. House bank charges the AGM batteries with a DC DC 18A charger each ... So it would seem worst case to buy then would be 2 * 18A victron DC DC chargers

To my understanding, this does not satisfy goal #4, protect alternator from voltage spikes due to LFP disconnect. The DC-DC converter cannot absorb much of a spike, and actually may be damaged by it. This would also only help goals #6 and #7 by 18 amps.

[Tin Tin]

Oh well a specific alternator protection module may help, its just more money https://shop.pkys.com/apm-12 $80

[wholybee]

You have a 170Amp alternator with Balmar controller. Just connect that directly to the LFP bank. Disconnect the battery temperature sensor, as that is not needed or compatible with LFP. Use the Duocharge(s) to charge the Lead Acid Batteries. Add an Alternator Protection device to protect from unexpected BMS disconnects.

The only thing you should need to buy from what you have is the APD. The key is that it is all connected correctly and programmed correctly. That is really the make or break of your system. Connected or Programmed wrong you will not be satisfied, and might blow stuff up. Connected and Programmed correctly, and you will forget you even have a battery bank, it will just keep ticking and working for a couple decades.

[Pete7]

Originally Posted by WE9V

To my understanding, this does not satisfy goal #4, protect alternator from voltage spikes due to LFP disconnect. The DC-DC converter cannot absorb much of a spike, and actually may be damaged by it. This would also only help goals #6 and #7 by 18 amps.

Problem isn't it. The OP could always put a lead-acid battery in parallel with the LFP to absorb the spike and cover an emergencies if the LFP shut down.

The hybrid option and long wire method of controlling the alternator is definitely an alternative solution.

[kavinsky]

One alternative to deal with the Alternator cut off issues is to set the alternator to stop charging at 14v.

14v (3.5v/cell) is ~96% of full capacity for LFP anyway, and you can trickle to a higher voltage on the bank using the solar array.

Use the existing duo chargers to charge the start / windlass batteries, kick them in when the LFP is at or above 13.8v (I.E. it is receiving a charging current - fully charged LFP will settle to about 13.4v at 100% charge when not actively receiving current).

For longevity of the batteries, set the Solar and Shore chargers to stop charging at 14.2v - you don't actually want to try and charge the batteries to the "100%" cell charge voltage of 3.65v/cell (14.6v), as some cells will be over, and that's the danger zone for LFP.

You might lose 1% capacity with a slightly lower charge end voltage, but you'll gain many cycles of extra battery life (and should avoid any BMS cut outs).

[emilecantin]

Originally Posted by kavinsky

One alternative to deal with the Alternator cut off issues is to set the alternator to stop charging at 14v.

The BMS can still disconnect the battery for a an individual cell voltage issue, you'll need to be VERY careful about balancing with such a setup. Also, over-temperature protection could still cut you off.


In my opinion he has 3 options:
1. Ditch the "drop-ins" and get proper batteries with a BMS that can tell the alternator to stop charging.
2. Ditch his "fastest charging" requirement and get a proper DC-DC that will give 50A continuously like the Victron Orion XS
3. Install an alternator protection device like the one Sterling makes, but these are still not ideal, I'd suggest more research on them before going that route.

[WE9V]

Originally Posted by emilecantin

In my opinion he has 3 options:
1. Ditch the "drop-ins" and get proper batteries with a BMS that can tell the alternator to stop charging.
2. Ditch his "fastest charging" requirement and get a proper DC-DC that will give 50A continuously like the Victron Orion XS
3. Install an alternator protection device like the one Sterling makes, but these are still not ideal, I'd suggest more research on them before going that route.

And why aren't solutions #3 or #4 valid, or not an option in your opinion? How do they not satisfy the goals?

[WE9V]

Originally Posted by kavinsky

One alternative to deal with the Alternator cut off issues is to set the alternator to stop charging at 14v.

Yes, this is a good idea. The Balmar regulator (as well as the Morningstar solar MPPT) are completely programmable. There is still a possibility of overcharging if solar topped off LFP earlier in the day, and then I fire up the engine. Yes, 14.0 is well below 4x3.65v, but these cells don't seem to be greatly balanced in this drop-in, so still a valid concern. But a bit of a corner case.

Originally Posted by kavinsky

Use the existing duo chargers to charge the start / windlass batteries, kick them in when the LFP is at or above 13.8v

Another good idea. With all AGMs, they were programmed to kick in at 13.0V. Cranking that up above the LFP 13.3-13.4 resting voltage would be good in order to not be charging AGMs without a real charging source working (shore/solar/alternator). So maybe 13.4-13.5v cut-in.

Originally Posted by kavinsky

For longevity of the batteries, set the Solar and Shore chargers to stop charging at 14.2v - you don't actually want to try and charge the batteries to the "100%" cell charge voltage of 3.65v/cell (14.6v)

The only thing not fully programmable is the older Victron Centaur charger. My choices there are Gel (14.2/13.5) or AGM (14.35/13.3). Since typically the shore charger is connected once and left on for days, it goes through one absorption stage and sits at float for days, assuming no power cuts. I was leaning towards the AGM setting for the lower 13.3v float and tolerating the one time only 14.35.

Balmar suggests 14.3 for their Lithium setting on the MC-614, but since it's fully programmable, 14.1 or 14.0 is doable.

The solar MPPT I have programmed to 14.1/13.6, but am considering maybe making that 14.1/13.4, with the float right at the LFP resting voltage.

[emilecantin]

Originally Posted by WE9V

And why aren't solutions #3 or #4 valid, or not an option in your opinion? How do they not satisfy the goals?

They both involve parallelling the batteries either through a transistor or a relay, which is functionnally the same thing. Putting LFP and Lead batteries in parallel is a terrible idea and a good recipe for a fire. If the voltages are even slightly different between the 2, the lithium battery can pour huge amounts of current into the lead battery, causing it to overheat and maybe even explode. It can maybe be done in a non-explodey manner if you have good, reliable fuses that can interrupt these huge currents (pretty much only Class T fuses, MEGA aren't enough). Maybe, just maybe the LFP's internal BMS will trip in such an event, but looking at Will Prowse's channel tells me that a lot of drop-in batteries don't actually have over-current protection (and even then, the FETs they use to disconnect the battery have a tendency to fail closed-circuit, which doesn't help us at all in this case).



Most manufacturers recommend against mixing batteries of different ages, do you really think mixing chemistries is magically okay?


No, a Lead bank needs to be kept well-separated from a LFP bank. The solution is simple, just use DC-DCs between banks if you need to charge one from another. Yes, that's added expense. But it's not much more expensive than proper fusing.

[WE9V]

Originally Posted by emilecantin

They both involve parallelling the batteries either through a transistor or a relay, which is functionnally the same thing.

Not exactly. The FET in the ArgoFET only allows current to flow in one direction, and therefore current cannot flow from the LFP to lead, or vice-versa.

[kavinsky]

Originally Posted by emilecantin

The BMS can still disconnect the battery for a an individual cell voltage issue, you'll need to be VERY careful about balancing with such a setup. Also, over-temperature protection could still cut you off.

I mean, you'd hope that there had been some sort of balancing of the battery. Better quality ones broken down tend to appear to have had a proper top balance, at which point, charging to 3.5v/cell makes it extremely unlikely, unless the battery goes bad.

I do agree that building your own is vastly better (and you can then do a yearly top balance easily enough), but he should be fine as is.

If you're really worried about a BMS disconnect frying the alternator, set the alternator to 13.8v, and move on with life. It'll get you to 93-95% charge that way (the voltage curve is very flat with LFP, you don't lose much by cutting charge from 13.6v and up).

[kavinsky]

Originally Posted by WE9V

Yes, this is a good idea. The Balmar regulator (as well as the Morningstar solar MPPT) are completely programmable. There is still a possibility of overcharging if solar topped off LFP earlier in the day, and then I fire up the engine. Yes, 14.0 is well below 4x3.65v, but these cells don't seem to be greatly balanced in this drop-in, so still a valid concern. But a bit of a corner case.


Another good idea. With all AGMs, they were programmed to kick in at 13.0V. Cranking that up above the LFP 13.3-13.4 resting voltage would be good in order to not be charging AGMs without a real charging source working (shore/solar/alternator). So maybe 13.4-13.5v cut-in.


The only thing not fully programmable is the older Victron Centaur charger. My choices there are Gel (14.2/13.5) or AGM (14.35/13.3). Since typically the shore charger is connected once and left on for days, it goes through one absorption stage and sits at float for days, assuming no power cuts. I was leaning towards the AGM setting for the lower 13.3v float and tolerating the one time only 14.35.

Balmar suggests 14.3 for their Lithium setting on the MC-614, but since it's fully programmable, 14.1 or 14.0 is doable.

The solar MPPT I have programmed to 14.1/13.6, but am considering maybe making that 14.1/13.4, with the float right at the LFP resting voltage.

Gel is the better setting.

Lower peak voltage is far more valuable for longterm lifespan than worrying about a 13.5v float LFP can float at 3.375v/cell for many years with no damage. When I was doing an initial charge on my most recent set of 4 cells, 14.35v was enough to push one of the cells close to 3.65v, prior to doing a proper top balance. 14.2v did not come close to pushing even the unbalanced cells that far out of alignment.

The better usage recommendation range for LFP is 10%-90% anyway, not 20%-100% (same usable AH, not pushing the voltage too high on charge).

[jcapo]

I Had 600AH of AGM batteries that I was going to replace with 400AH of LFP. I installed 600AH of LFP. I am very happy that I went with more capacity.

I went the ArgoFET route. If the new Victron 50A XS chargers had been available, I might have gone that route.