LFP FLA hybrid

[Qayaq]

I’ve looked at LFP for a long time.
FLA heartily sick of them. Good for starting engines but degrade quickly if you actually deep discharge them. And I’ve had enough of diesel use and engine wear trying to keep them from sulfating.

I have an electric bike and I’m thoroughly impressed with that battery.

LFP stumbling block is the expense and it’s a bit of an experiment. FLA one thing good about them they are reliable and cheap…..ish.

So why not marry them together, FLAs are great at short big bursts of power, don’t like being discharged or charged but are super reliable. So just keep them charged up & ready for back up when the LFP goes flat.

LFPs Are great for discharge, re charge really efficiently.
So all the stuff that normally drags down my FLA goes on that circuit.

So being that I want to keep my cash for cruising vouchers and that my boat has some good FLA charging gear. Here's my thoughts, drawing attached

Charging sources
Alt max 50amp, dialed back on Balmar 614. (single belt)
4 x 100W solar
Wind generator
Pro mariner 30-amp 240v charger

The idea is to lower the voltage from charge Bus to LFP by fitting an old style splitter diode pack which also stops feedback from LFP to FLA, may be able to pick one that has just right voltage drop. And they are cheap.

If it turns out that the FLA tries to charge the LFP through the diode splitter then maybe a Schottky diode could be put on that line

LVC. Because my LFP load is light approx. total worse case about 25amps, the LVC contactor could even be a 40amp NC/NO relay for testing, would prefer to find a better low current unit, also need something that would switch it at specified voltages, perhaps a Raspberry? Hopefully something easier..

HVC, Poss max on HVC could be as high as 90/100amps on a windy sunny day just after starting the engine so would need to look into a suitable unit for that. Again need to find something to switch at specified voltages. The reason I would want the HVC is to prevent float on the LFP.

So in the event of an LVC event which I would expect to happen occasionally, it simply switches the switch board back from being split LFP/FLA back to all FLA. Win Win.

OK the radar and auto pilot etc would reset but that’s not a major.

Can you see holes in this idea?

Would appreciate your input

Drawing attached.

Cheers
Simon

[44'cruisingcat]

There's a bit of a discussion here that might be helpful:

http://www.cruisersforum.com/forums/...ghlight=Heaven

[CatNewBee]

Originally Posted by Qayaq

I’ve looked at LFP for a long time.
FLA heartily sick of them. Good for starting engines but degrade quickly if you actually deep discharge them. And I’ve had enough of diesel use and engine wear trying to keep them from sulfating.

I have an electric bike and I’m thoroughly impressed with that battery.

LFP stumbling block is the expense and it’s a bit of an experiment. FLA one thing good about them they are reliable and cheap…..ish.

So why not marry them together, FLAs are great at short big bursts of power, don’t like being discharged or charged but are super reliable. So just keep them charged up & ready for back up when the LFP goes flat.

LFPs Are great for discharge, re charge really efficiently.
So all the stuff that normally drags down my FLA goes on that circuit.

So being that I want to keep my cash for cruising vouchers and that my boat has some good FLA charging gear. Here's my thoughts, drawing attached

Charging sources
Alt max 50amp, dialed back on Balmar 614. (single belt)
4 x 100W solar
Wind generator
Pro mariner 30-amp 240v charger

The idea is to lower the voltage from charge Bus to LFP by fitting an old style splitter diode pack which also stops feedback from LFP to FLA, may be able to pick one that has just right voltage drop. And they are cheap.

If it turns out that the FLA tries to charge the LFP through the diode splitter then maybe a Schottky diode could be put on that line

LVC. Because my LFP load is light approx. total worse case about 25amps, the LVC contactor could even be a 40amp NC/NO relay for testing, would prefer to find a better low current unit, also need something that would switch it at specified voltages, perhaps a Raspberry? Hopefully something easier..

HVC, Poss max on HVC could be as high as 90/100amps on a windy sunny day just after starting the engine so would need to look into a suitable unit for that. Again need to find something to switch at specified voltages. The reason I would want the HVC is to prevent float on the LFP.

So in the event of an LVC event which I would expect to happen occasionally, it simply switches the switch board back from being split LFP/FLA back to all FLA. Win Win.

OK the radar and auto pilot etc would reset but that’s not a major.

Can you see holes in this idea?

Would appreciate your input

Drawing attached.

Cheers
Simon

A hybrid solution is possible, however the devil is in the details.

Some just parallel LFP and FLA, What happens is, empty LFP gets most of the Amps in the first charging time until it reaches about 13.5.. 13.6V (80%), FLA gets then more amps and bulk charges, while LFP still draws a significant amount of amps, While the voltage rises over 14.2V, LFP is full, and FLA enters Absorption. It is not very healthy to keep LFP at this voltage, if you disconnect it will slowly drop to 13.5.. 13.6V.

Charging from LFP to FLA for absorption means using a DC/DC converter. Doable with some circuitry around.

When discharging and having both batteries in parallel, LFP would stay at 13.4V, what is float for FLA, so all current drawn will come from the lfp as long as the voltage stays above 12.8V, so you first will discharge the lfp down to 30..40% depending on chosen capacity before your FLA will contribute any significant current.

Again doable, but why bother with FLA, when you will run most of the time on LFP power anyway, why spent on electronics to switch/charge the FLA, when you can use this money for a larger LFP bank and a less sophisticated control circuitry?

[Qayaq]

Hi Catnewbee,
Thanks for your post.
I'm concerned about LFP failure and I like FLA reliability. If you check the drawing you should see the idea is minimal expense on control electronics, splitting my loads to be suitable for each type of battery while retaining reliability and increasing efficiency.
The idea is not to parallel but to split the switchboard loads.at no point should the battery be joined and I'm thinking that a diode splitter with the volt drop it naturally has could achieve the ability to keep all my old FLA charging while dropping the voltage suitable for LFP.
Cheers Simon

[CatNewBee]

LFP cells are usualy at least as reliable as FLA batteries, what can fail is the BMS to keep the cells balanced and could eventually switch the charging or discharging off.

More error-prone are discrete cell modules, little copper circuit boards directly screwed to the cell poles, like those of 123BMS, and others. They are daysy-chained to each other or have bus wires connected to talk to each other. There comes the problem with this setup.

First is vibrations in a vehicle or boat., the pretty thin wires and the copper plane secured by a single screw may produce a circuit break or a plane cut over time, the soldering may detorate if not properly coated, the contacts can corrode - we are talking about 20 years life expectancy....

There are encapsulated BMS with single sense wires to the cells, much better dessign and easier to fix at the cells, however then its all or nothing. If it breakes, you replace the whole BMS and not a single circuit board.

When the BMS kicks in, what never should happen in normal operation, it does not mean, the battery is dead, it means the bms has found an issue it cannot handle himself. You always can just bypass the solenoid and use the cells as a regular battery if you have external BMS and solenoids.

And this is the main drawback of batteries with integrated BMS, there is no way to override unless you open the battery and lose warranty. Another one is those batteries allow much less Amps for charge and discharge than the used cells would allow, just to keep the circuitry cheep and small. The useability of those packs is much lower than of batteries from single cells and external BMS, and this batteries are way more expensive too.

Anyway, it is a push-button thing to override the error and continue to use the cells - just be careful to not discharge too low and recharge as soon as you can. Fix the issue and re-enable the BMS.

Some even just use tha bare cells without protection, balance them manually eventually and accept a shorter life expectancy by human errors. This battery will newer disconnect, except it is really empty discharged way below the 2.8V per cell.

[Dockhead]

Originally Posted by Qayaq

Hi Catnewbee,
Thanks for your post.
I'm concerned about LFP failure and I like FLA reliability. If you check the drawing you should see the idea is minimal expense on control electronics, splitting my loads to be suitable for each type of battery while retaining reliability and increasing efficiency.
The idea is not to parallel but to split the switchboard loads.at no point should the battery be joined and I'm thinking that a diode splitter with the volt drop it naturally has could achieve the ability to keep all my old FLA charging while dropping the voltage suitable for LFP.
Cheers Simon

My own opinion, for whatever little it may be worth, is that you will not be able to get the correct charging profile for both battery types, charging them off the same bus, diode splitter or no. What they need is very different and just taking one charge profile and reducing by a fixed amount of voltage is not going to produce anything satisfactory.


I am contemplating a system like this myself, and although it will cost a bit of complexity, I would definitely provide each battery bank with its entirely separate charging profile. How?


* Separate AC chargers. That one is easy, and makes it easier to expand charging capacity to match the huge acceptance rate of lithium.


* Connect solar and alternator only to ONE of the two banks, using appropriate controllers to give that bank exactly the charge profile it needs.


* Then, charge the other bank using a battery-to-battery charger, which will allow a completely separate charge profile to be created.




I would keep all my loads connected to the lead bank so that cutout of the lithium back will not shut down the system. Then simply charge the lead from solar -- when the lead is not getting its own charge via separate AC charger -- with the B2B charger.


The B2B charger also acts as a power supply, so when the lead is charged, power will be passed straight through from the lithium to the loads, or from the alternator or solar.




This is somewhat more complex, but seems to me to be the minimum to make the system work right.




It has a huge advantage in that the lead bank will be given ideal conditions, with a full and total finish charge each cycle.

[CatNewBee]

It comes down to amps and voltage.

No problem to find a profile that fits both, by dropping one earlier. A GEL profile is pretty close to the LFP profile, except the long absorption time. Also you need to disable temp compensation, because it varies the absoption voltage too much for a LFP.

Another option is to disconnect based on SOC (measured by a coulomb counter / battery monitor), the problem is the drift over time when you never reach the reset point of 100%.

[Qayaq]

Hi Dockhead.
Thanks for your post,
I was hoping to avoid the B2B. Mainly based on cost.
My thought being that the LFP would take any charge that my system is capable of putting in to it...Not that much...then hit HVC /charge cut out which doesn't cut back in until there is a reasonable voltage drop, and being that my loads on the LFP are low but constant IE fridge etc, the load wouldn't re trigger the HVC or incidentally LVC re trigger at low SOC.

Catnewbee, do you know of a controller could trigger or how you could trigger a HVC on SOC . That would be interesting.

I'm not that far away from carrying out a battery system refit, the PO fitted the batteries in the keel, so they are getting moved along with cabling switches etc and at the same time in goes a small LFP bank, so a bit of R&D is on the cards. I'm hoping to keep it cheap and re use as much original gear as possible but it's likely to end up with some unplanned purchases...
Should be fun.
Cheers Simon

[CatNewBee]

Originally Posted by Qayaq

Hi Dockhead.
Thanks for your post,
I was hoping to avoid the B2B. Mainly based on cost.
My thought being that the LFP would take any charge that my system is capable of putting in to it...Not that much...then hit HVC /charge cut out which doesn't cut back in until there is a reasonable voltage drop, and being that my loads on the LFP are low but constant IE fridge etc, the load wouldn't re trigger the HVC or incidentally LVC re trigger at low SOC.

Catnewbee, do you know of a controller could trigger or how you could trigger a HVC on SOC . That would be interesting.

I'm not that far away from carrying out a battery system refit, the PO fitted the batteries in the keel, so they are getting moved along with cabling switches etc and at the same time in goes a small LFP bank, so a bit of R&D is on the cards. I'm hoping to keep it cheap and re use as much original gear as possible but it's likely to end up with some unplanned purchases...
Should be fun.
Cheers Simon

Victron BMV 700/ 712 have programmable relay contacts, you can configure them to anything...

[john61ct]

A fully setpoint adjustable B2B DCDC charger will convert any primary source feeding your FLA bank to exactly what you want your LFP bank to receive.

If anyone knows one besides Sterling's let me know.

Much better solution than doubling up on separate mains / solar / alt+VR setups.

[CatNewBee]

Originally Posted by john61ct

A fully setpoint adjustable B2B DCDC charger will convert any primary source feeding your FLA bank to exactly what you want your LFP bank to receive.

If anyone knows one besides Sterling's let me know.

Much better solution than doubling up on separate mains / solar / alt+VR setups.

Votronic in the EU...
https://www.votronic.de/index.php/de...nnungs-wandler

also Mastervolt, Victron etc...

[john61ct]

They give a set of canned choices.

I mean user custom, like if I decide I want 3.42Vpc rather than 3.47

Also, Votronic refuses to support the NA market.

I'd love to see Victron come out with such a line, MV is also pretty poorly supported here and IMO overpriced.

[Dockhead]

Originally Posted by four winds

Thanks Dockhead. I hear you and understand your input. Obviously when you mentioned the hybrid banks idea it got me thinking. Though I missed the diagram you posted. It did occur to me I might be able to extrapolate your notes down from "mega yacht" to "garbage scow" status for my needs. A little hyperbole, not much. The other thing is you are gonna do it and I'm just thinking for fun.

Keeping the lead bank topped off, as a starting/light loads/reserve bank appeals. Lasts long time and there when you need it. Like when the finicky Li bank goes down.

Letting the Li bank handle the heavier loads and being less concerned with the SOC being low for a while appeals.

And in both cases promote longevity of the banks.

About the B2B, I have a knowledge vacuum there. Sounds like it's smarter than I thought, more than a charge combiner relay. I considered it to be operated manually, turned on to charge the Li bank (in my case) when needed and chosen by the operator.

Why the alt going to the lead bank in my case? To ensure the lead (reserve) bank is almost always full. And because I don't fully trust my experimental, finicky Li bank, bless its heart. And maybe no LVC induced alt failure, too, or the need for more hardware.

So the lead bank has priority. The B2B stays off til I say so. And I keep the banks and alt charge wire on the 1-2 switch as well. So I could direct alt charge to the Li bank as a back up.

Add a few of those cheap volt/amp meters here and there, and be my own BMS. I'm always on the boat. I'd just have to be vigilant when charging the Li bank at a time of my choosing. That time would be when the lead bank is "full" and the alt output goes to the Li bank via the B2B.

If I'm away from the boat for a few days the lead bank will take care of bilge and anchor light. And the Li bank would be off.


About the B2B again. Let's say the alt goes to the Li bank and B2B to the lead bank. When the Li bank is "full" does the B2B stop all charging to the Li bank and send everything to the lead bank?

In my "backwards" scenario I assumed the B2B would send most of the charge to the Li bank while maintaining absorption or float on the "full" lead bank.

I'm bringing this discussion out of the "Depth of Discharge Myth?" thread and into here.

[Dockhead]

So, in response to Four Winds:


I think that something like this might actually make a great deal of sense even on a small boat and on a limited budget.


I think the idea of a hybrid lithium-lead bank is very powerful and might be applicable to a wide variety of use cases.


First of all, for you to advance your thinking on this, you need to read carefully the manual for the Sterling B2B charger and get a thorough understanding of how it works. Here: https://sterling-power.com/products/...rip-proof-ip21



It is most certainly not a combiner. It is a full fledged battery charger which creates its own charging profile. It is merely powered by DC; the charging profile has nothing to do with what is going on on the supply side. This is really cool and key to an application like this, because it means you can regulate the alternator (or other chargers, too) to suit the primary bank exactly, and then charge the secondary bank with its own completely different profile.





I would definitely put the alternator on the lithium bank, for the reasons I mentioned in the last post. If the lithium bank goes down, your loads will continue to be powered by the lead bank, and you can manually override the B2B charger to force it to work with no primary battery attached (Sterling calls the primary battery the "output battery"), passing alternator power on to it.


The simplest way to do this on a small boat would be to create a small lithium bank which is not connected to any load except the B2B charger and which is used exclusively for topping off the lead bank and passing power through the lead system.


If the lithium is quite small in relation to the lead bank size, I think you could then go back to connecting the alternator to the lead bank, but then you would need TWO B2B chargers (£££), and you would need to construct a way to switch them over so they're not connected at the same time. But, you would gain a great deal of simplicity and simplicity in operation. You would just need to manually switch back and forth between charging the lithium and using power from the lithium.

[Dockhead]

Highly schematic idea of how it would work on my boat:






So this would use a great deal of the existing configuration and equipment.



I would add a Victron Skylla charger to the lead bank set at 0.2C so that each bank has its own mains power charger. I actually already have a 30 amp Newmar charger which I could use, but I think I will sell that and buy a new Skylla which will easier to integrate into the new system. They are not that expensive.



Alternator charging would be shared via a Sterling B2B charger.


That's about all the major equipment which needs to be added.



I would add Victron battery monitors and shunts for each bank so that I could easily keep up with the total current flows, but I would continue to monitor the lead bank with a SmartGauge.


As far as control goes -- I would need to be able to manually override the B2B automatic controls to (1) charge the lead from the lithium in the absence of alternator power; and (2) charge the lead from the alternator in the absence of an "output battery" in the system, in case the lithium bank goes down.




A good question here is whether one wants the inverter on the lead or on the lithium. I think lithium, but if someone setting up a hybrid bank were to choose a small lithium bank, with most of the system's real capacity in the lead, then you might prefer to connect it to the lead. This will be somewhat simpler since you wouldn't need the cascade of LVC and HVC which is needed in my system to effectively separate loads from charging sources -- the charger/inverter can be BOTH a load AND a charging source; hence the need for the cascade.