1400 Ah Lithium from cells

[john61ct]

A charger designed for lead is fine, as long as it does what you want it to, and nothing you don't. Behaviour of any charge source needs verifying with trusted instruments anyway.

Even for FLA, no one here thinks automatic equalization is a good idea.

LFP doesn't do "stages".

Best charge source is a simple and robust power supply. Intelligently current limited, amp rate no higher than you consider safe. For me, standard 13.8V is perfect, or get a V adjustable one.

No "smarts", no multiple stages, nothing automated.

Get to your defined Full voltage and shut it down, done.

If you use a automated relay for that last bit, great, but IMO don't fully trust it.

[john61ct]

Originally Posted by CatNewBee

maximum voltage per cell 4.0V (translates to 16.0V), cell full at 3.65V (translates to 14.6V). initialization of new cells to 4.0V as top balancing. Lowest voltage before damage 2.8V (11.2V)

Again, any user following this to guide their day to usage will see greatly reduce bank cycles lifetime, maybe 20-30% of what is possible avoiding the shoulders.

The 100% Full point is where the owner decides. Regularly going over 3.5V, you may say "no damage" but to me, losing thousands of cycles off the back end **for no reason** is just silly. For what? gaining a few percent greater capacity?

Your money, your choice.


> Victron BMV, set the relay to disconnect at 90% SOC, and turn back at 50. This beast will drift, because it never will see SOC 100% to re-sync

First, no BM is accurate enough to control mission critical operations based on SoC.

Second, it is the owners job to regularly and manually reset the BM to Full when their setpoint is reached.

> way to reduce voltage for any charger in 0.7V steps. Just add a diode in the path

The actual voltage reduction will vary based on source level and current.

Such kludgey workarounds have no place managing such an expensive asset. IMO just buy the right infrastructure in the first place, even .05V can make a difference


> if you do not have a programmable charge controller

If you want automated, with an expensive LFP bank why on earth would you have one that wasn't?

[CatNewBee]

Hi john61ct,

the world is not just black and white.

Honestly, you have many systems on board or in a vehicle, that charge the batteries. Buying a LFP charger is a good idea for stationary installations and a single source of charge. Then yes, the investment is some how rational.

In a vessel and in a RV there are legacy systems installed, that some people want to keep or re-use. This happens out there. The main question is is it safe and can it be converted some how to extend the usage.

For a green field approach - yes, buy all LFP capable gear, change all factory settings to a more conservative value and off you go.

As main sail stated correctly - not every product labeled LFP compatible is configured to the best possible settings out of the box. Even Victron Multiplus / Quattro charger, that got a recommendation use 13.5V float, 14.2V absorption for 1h as default LFP profile.

So lets see what we have:

- Alternators and built-in regulators
- Wind generators
- Solar panels with a broad range of controllers
- B2B - Chargers
- Hydro-generators
- DC generators
- AC generators
- shore power charger

and all this from different brands, age and quality.

In a perfect world you would rip out all the stuff and add a suitable controller to each source - in the real world you would try to use what you have and to add what's necessary to make it safe.

At the end of the day, all those provide current at a voltage, that is limited by the capability of the source (you can't have more than it could deliver) and the Battery (the batteries voltage must be lower than the source voltage so a charge current can flow, because the cells resistance is much lower then the resistance of the source, the cell voltage rules).

So what are your options (from low-tech to high-tech):

- cut the connection of the charge current
- turn off a source gracefully by a remote switch
- tell the source to reduce voltage and / or current

Some sources do not like brute-force disconnections under load (alternators, solar controller with automatic voltage sensing of the batteries...) and also relays and switches don't like disconnections under heavy current.

To the BMV. It does not need to be accurate at all, there is still the 2nd and 3rd line of protection by the BMS, the notification/warning leven and the emergency disconnect level. It does not matter, if the BMV really switches charging on at 50%, at 30% or at 60% when set to 50% and when it switches off, as long as it is below the warning level - you will be fine.

If not - charge one time to full and re-sync. This is a semi-automatic approach that would support the main sail suggestions to keep the LFP somewhere in the middle and do not use the full potential / energy range.
Also to not do automatic balancing in regular use.

It is not exact, but as long as you have your BMS watching, the risk is minimal. Anything extreme would raise the alarm and the BMS will take action. It just does not happen regularly. Even if not balancing - the BMS watches at the cell voltages and will raise an alarm when the difference is bigger then set as acceptable, and eventually even emergency disconnects everything if this failure exceeds the limits too much.

That said, in the "safe zone" you can use any device for charging - as long as it can stand the lithium currents.

[john61ct]

I don't disagree, and have long advocated as an implementation detail for Sterling's B2B to be the sole direct charge source.

That allows any old source to be primary, no replacements needed.

But my points above had nothing to do with implementation details.

[bobfnbw]

I'm not sure I understand why someone would spend thousands of dollars on Lifepo4 batteries and not upgrade the charging system. Makes no sense.
Seems to me that the best way to safe guard them is to limit the amount that the BMS sees. So if you have 400 amp hours of battery, set the bms to 360. Set the charge voltage at max for that battery type, but cut off the float, or divert it to the water heater.
I don't think that boats that spend a lot of time hooked up to a marina power supply should have lithium banks. Seems to easy for something to happen. But liveaboards on solar should be fine, as long as the BMS doesn't crap out on you. Still can have other ways of keeping the battery from fully discharging or fully charging.
And while you might not be able to both top and bottom balance when charging, you can bottom balance to start off, and check it from time to time.
I think, if I went with this system, that I would have a victron BMV702 or equivalent in the cockpit to keep track of it all.
My last boat I went with TPPL agms due to the high acceptance capability they had. I used the Oddysey ones marketed by sears. I had high hopes for them, installed a high amp balmar alternator with a serpentine belt, a victron multiplus 3000/100 inverter charger, and was upgrading the solar charging system. Never got to use it, so not sure how it all worked out. But lithium battery house bank seems like the only way to go, in the way I plan to use them, except maybe the firefly batteries. But they take up more room than I want to devote to them, and weight to much.
The not charging while temps are low thing can be gotten around easily by putting the batteries inside the cabin. Just don't charge them when the temp is low. Come back, turn on the diesel heater, get the cabin warm, then charge away!
But I can see how a goof, mistake or just plain stupidity can ruin it all. That would suck. But then lead acid batteries can be easily ruined as well by overcharging, under charging or sulfation.
This is one of the BMS that I have been looking at
https://www.ev-power.eu/Battery-Mana...tooth-4-0.html
the other being ElectroDacus and his solar bms. Anyone have any experience with either on a boat?

[CatNewBee]

User manual 123Smart BMS
123Smart BMS homepage
Hi, I have seen a L420 or 440? with a quite large Winston LFP bank using this BMS offered on Yachtworld last year. I know 2 or 3 RV'er using it in DIY projects with 200Ah cells and seem to be happy with the setup.

However I have not further investigated / evaluated it, because I do not wanted dedicated cell modules and the clutter on the cells, also the balancing current seemed to me quite low with 1A for my cell capacity. On boats we have salty and moist air, I do not know how exposed surface mounted circuit boards will react over time regarding corrosion and how protected the modules are. Another Issue I am a little sensitive about is the connetion of the boards to the cells. Usually you want the cell-bridges be as tight as possible connected for low resistance and high currents, there are minimum torque requirements to make a good connection. I would not squeeze the boards in between the screw and the bridge. There are special screws around with thread inside for a smaller second screw for the boards, so you can have a tight, torqued connection of the bridge and on top a not so tight optimal connection of the balancer board - especially for larger cells.

As far as I understood, there are lots of DIP-switches to pre-set values, and some can be set by software.

The system only provides 2 NO/NC bi-stable relay outputs, so it can only control over voltage protection and under voltage protection when the first cell reaches one of the programmed thresholds. to disconnect either the loads or the charging sources, balancing is done by heat (1A discharge of all cells that go over the pre-set balancing voltage called V-bypass, so technically it is a programmable passive top balancing), the modules also measure the temp and can disconnect on failures.

The BMS comes also with a current sensor and has a coulomb counter in his software, so it can measure charging / discharging and calculate SOC. The UI is realized in a Android or Apple App, connected by Bluetooth.

I guess, for smaller installation it's a good alternative to the pre-programed cell-modules BMS, it is more flexible.

What it does not have, are contacts, that can be used to control the sources before the first cell reaches one of the limits, so it is more a protection system for a LFP than a BMS that not only is a safety-switch, but also capable to control charging and discharging during regular operation without dropping the circuits. On the up-side - it is very affordable and you can use your phone as a display, you can also save on a separate battery computer.

But I would probably install on top a BMV to have a second source of information and more - important - a programmable independent relay.

[CatNewBee]

Originally Posted by john61ct

I don't disagree, and have long advocated as an implementation detail for Sterling's B2B to be the sole direct charge source.

That allows any old source to be primary, no replacements needed.

But my points above had nothing to do with implementation details.

Yes, this could be an alternative to combine all in-capable sources through a B2B charger. I am not familiar with the sterling products, but if they are programmable, that will be OK - with one exception of course.

Alternators and some solar controller need permanently a connected battery, so you need a small buffer-battery upfront on the input side of the B2B charger - it could be the engine start battery.

There are integrated sources - like the Quattro inverter/charger you cannot separate, so they will live in parallel - but in most cases they are capable of programming so not a problem for the LFP bank.

The downside is: you need a quite large device that can handle all sources, and you will loose some energy during the transformation - this devices produce some heat - so the efficiency of the system will be not so good - especially on weak sources like wind generators in light winds or solar panels in cloudy weather.

[john61ct]

Originally Posted by CatNewBee

I am not familiar with the sterling products, but if they are programmable, that will be OK - with one exception of course.

I haven't come across any others that are adjustable. Charles Sterling has told me he's coming out with a 180A unit "RSN".

Yes, having a cheaper lead batt buffering up-front is a non-issue. Efficiency loss is inevitable at many points in the whole system, also not much concern to me.

With the ability to recharge in an hour on demand a few times per week, I don't see much point in solar anymore, unless you can go solar-only for weeks at a time, and don't mind the windage nor how it looks.


> integrated sources - like the Quattro inverter/charger you cannot separate

To me, just not suitable, as keeping the charge buss separate from distribution to consumer loads is a fundamental design principle.

[john61ct]

Originally Posted by bobfnbw

Seems to me that the best way to safe guard them is to limit the amount that the BMS sees. So if you have 400 amp hours of battery, set the bms to 360. Set the charge voltage at max for that battery type, but cut off the float, or divert it to the water heater.

First off we're discussing LFP specifically, which includes Winston; "lithium" is a useless term for specifics.

The problem is the "max for that battery type" is not objective, as the vendors set it way too high. For those who want to get better longevity, the "Full" point voltage needs to drop way down, IMO 3.45Vpc as discussed above.

Secondly, BMS is really a set of protective functions, for a large bank should be the "last ditch" set in multiple redundant layers of mechanisms, not counting those used in day to day operations.

SoC is a guesstimation at best, and no meter is accurate or trustworthy enough to use for those lower-level functions that ensure bank survival.

> I don't think that boats that spend a lot of time hooked up to a marina power supply should have lithium banks. Seems to easy for something to happen.

You are conflating LFP with all the other riskier lithium chemistries.


> But liveaboards on solar should be fine

LFP owners can take advantage of all the charge sources used for lead, and (see above) solar may often be reduced or eliminated in the transition, since there is no longer any need to get to Full.

> I would have a victron BMV702 or equivalent in the cockpit to keep track of it all.

Fine unit, but with only one relay, not sufficient.

[CatNewBee]

Originally Posted by john61ct

> I would have a victron BMV702 or equivalent in the cockpit to keep track of it all.

Fine unit, but with only one relay, not sufficient.

If you use it in combination with your BMS, it can take care on charging when programmed as charging relay based on both SOC and Voltage or even on SOC only. Set Peukert to 1.01, the LFP are very efficient, so you have almost no drift between charged Ah and discharged Ah, the BMV is accurate enough for many cycles.

You can automate the level control of your SOC to your wishes e.g. between 50% and 80% (relay goes on below 50 and turns off over 80) so you do not need to watch after charging all the time, if you want a higher safe capacity, set it to 70 to 80 or any other value, it will just enable your charger only when needed and disable it long before your cells are full.

It's quite useful in this conservative scenario if you want to max-out the lifetime of the cells.

[john61ct]

Yes as I said, great as just one piece of an overall system.

Would be good as a front end to that Sterling B2B, set to output charge at 13.8V

and then just isolate its input side when the bank actually reaches that setpoint.

Recommence charging at say 12.5V for readily available high-amp dino juice sources,
but set lots higher, say 13V when solar or wind is the main source.

Note this is just controlling charging, not to be confused with overall LVD / OVD bank protection.

Which I would not depend on just one layer, the lowest failsafe "BMS" level would isolate the bank completely at below 11.9V or above 14.1V

But load buss protection would be set to alarm at 12.5V, shut down big-amp load circuits at 12.3V, only leaving essential safety/navigation systems active, then shutting even those down at 12.1.

Obviously delays are critical for LVDs, since a heavy startup load like a motor can briefly pull voltage down below a setpoint.

If there's a Reserve bank (mine is Firefly Oasis), then rather than just cutting consumer loads off, a relay can switch them over to that.

Besides Victron, I want to try Bogart's Pentametric, Magnum BMK, Maretron, I'm sure MasterVolt is good, maybe BEP.

For non-critical adjustable voltage relays, apparently these are pretty reliable for the price: https://m.ebay.com/sch/i.html?sid=sb...ons.sales.usa4

An important piece for relays controlling the overall BMS functionality, is to be realistic about ampacity.

Just because a contactor is rated for 400A, doesn't mean you should regularly be opening and closing with currents live at anywhere near that rate.

And that's where your protective infrastructure gets pricey.

[CatNewBee]

That's what I am saying.

If you understand LFP requirements, and add some logic at the right place,
you do not need to replace all components of your circuitry only to have
all with "LFP capable" badges... Suitable for Lithium does not mean mutch either, most controller supplier do not understand the various LFP and LION technologies and safe thresholds.

You need 2 things: a BMS that takes care and pulls the plug if the cells run out of sync or out of bonds as a "circuit breaker", and some independent logic that keeps an eye on the SOC and controls all charging sources while charging and implements your chosen charging regime.

You can funnel it through a configurable a B2B charger - like the Sterling - you can also use suitable relays to disconnect the unregulated sources from the house bank. Having the start battery to be charged in parallel too, you can just drop the current to the house bank - the alternators and the other stuff will follow his curve on the starter AGM and keep her happy.

Yes, you need the right gear for the job, that can connect or disconnect safely the occurring currents. If you go down the road with the B2B charger, you can chose one, that has a remote switch ( remote control contact like the D+ that you can interrupt by a small relay.) So no heavy duty relays necessary if you go this path. You have to be carefull not to drain your start battery by the sterling, set a starting voltage for the B2B high enough.

It is doable, but you must know what you do and why you are doing it.

[CatNewBee]

Just want to give you some updates on my LiFeYPO4 testing.

Meanwhile a lot of stuff arrived and I started to install and test it in my workshop to see how it works.

As I announced earlier, we want to get rid of all propane in the galley and go electric only.

So what's new:

I set up the 1000Ah battery bank I reported earlier, built the interface circuitry, attached the BMS and the monitoring system, added a BMV712 to watch the Amp's / Ah, Voltage etc, also connected a Victron Quattro 12-5000 and programmed it to my findings.

The Galley consist now of a fancy variable induction cook-top (8 zones, that can be combined to 4 Burner or 3 burner or 2 large zones) with maximum draft of 7kW, a mid-size Oven with 3kW and a 2000W Mastervolt MassSine Inverter.

Both inverter are set up in a master-slave configuration where the Victron is the master. Using the wizards the Victron starts the MassSine when the load goes over 3kW and turns it off if the load drops below 1800W, it also sets the current limit to 8A on Input2 and activates PowerAssist to boost the 2kW with its 5kW up to a total of 7kW - and of course sets the charge current to zero when the MassSineInverter is running on AC-Input 2. If shore power is on - (on AC-Input2), the current limiter is pre-set to 25A and charging current to maximum = 220A. This is controlled by an external signal from a relay that recognizes incoming shore power.

This works completely automatic and unattended.

Checked the Induction cook-top, even with power-boost's on both sides the inverter manage to supply enough energy to boil a lot of water during the test, dragging up to 500A peak from the battery.

I also realized, that this load does not stay consistent all the time, the cook-top reduces the power when the temperature is reached, also you do not cook all the time with the power booster - so the load while cooking is much lower.

Also tested the Oven for 270°C with Power Boost and baking for 1h, it drags round about 180Ah during this time, maximum load around 220A, after reaching temperature phases of 110A - and 5A. So baking bread and cooking (not at the same time all-in) is absolutely feasible on board. The batteries would have sufficient power without recharge for 3-5 days on this usage I guess.

Also tested a lot of other stuff, like espresso machine, sailrite, dyson hot air blower, battery chargers, soldering station, hot-knife, kitchen-aid and anything I could find with a 220V plug. I am really impressed by the performance of the Lithium cells.

Also tested my Victron MPPT charger with an alternative power source, final tests here will be done on board after the installation.

Will do some more stress-tests on the system the next weeks and hopefully install all the stuff in April when I start the re-fit on board.

Can't wait to rip out the propane installation and the old AGM's.

[senormechanico]

It sounds like a great system. Even with eye popping draws, I'm confident your boat will be a great success.

Lead will eventually be joining buggy whips in the ash heap of history.

[CatNewBee]

Here the changes to the planned 220V installation, the signals between the components and the interface for the BMS/BMV to the ML RBS I will provide on a separate post.