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

[SV THIRD DAY]

I have a client running the External Diodes on 160A alternator with a 800AH LiFePO4 bank and Balmar 614. So far the results are disappointing and he isn't seeing enough of a reduction in alternator temps to justify the cost. (His words)

So don't run and jump too fast yet....

There still is not a good solution out there yet for alternator charging of your LiFePO4 bank...that is something that will need to change before the technology goes mainstream or the alternator manufacturers will start excluding warranty to LiFePO4 users.

[senormechanico]

Keel cooled diodes next?

[SV THIRD DAY]

Originally Posted by senormechanico

Keel cooled diodes next?

Oh no...
Not the Air vs Water cooled debate all over again...ha ha ha

[evm1024]

Originally Posted by SV THIRD DAY

I have a client running the External Diodes on 160A alternator with a 800AH LiFePO4 bank and Balmar 614. So far the results are disappointing and he isn't seeing enough of a reduction in alternator temps to justify the cost. (His words)

So don't run and jump too fast yet....

There still is not a good solution out there yet for alternator charging of your LiFePO4 bank...that is something that will need to change before the technology goes mainstream or the alternator manufacturers will start excluding warranty to LiFePO4 users.

SV3d
THats good to know. As I said it will be interesting to see what MainSail gets for results. The limits might be elsewhere.

OSS
No need to invoke magic diodes. I would not expect a factor of 2 difference. I would think one might get 5% or 10%.

[isaacchacon]

Originally Posted by isaacchacon

Hi all. THis is my second post in this great forum, this time i want your help to get feedback about my electrical diagram. I am building a van conversion with 200AH LIFEPO4, since very few resources online, I come here hoping to get responses even though i don't have a boat.

Please help and tear apart my design!!!!!

I decided to buy Elite power solution's 200AH battery + BMS system, and am just planning the main system. I will later buy a 3000 watt victron inverter / charger. but for now i will bear with the 15A charger provided. I have not yet used / tested the battery. The package included a charger, a video screen for monitoring SoC, BMS, the battery, and a shunt. (all of them shown in the diagram below).

Contrary to many opinions that i've seen, I decided to go with a latching relay for switching the main loads (BLue sea 7700, 500 amp) , and will be controlled by relays and the BMS. THis latching relay can be manually switched without power, saving me the need to buy a switch.

My installation is somehow inspired on the mar azul boat blog, and i slightly modified the relays to control the under voltage. There are also very helpful manuals that were provided by elite power solutions, and of course learned some from this forum.

The below diagram is a reduced version that omits some components of the BMS package (FUSES, Ground connections, video screen, and others), hoping to give a good understanding for controlling Under VOltage and Over Voltage Conditions:
Attachment 124491
Attachment 124317

THe BMS has ( Black biggest central component) , among other, these outputs:
- OV : Over voltage, +12V normally, GND if any cell >3.6V
- UV : Under voltage , +12V normally, GND if any cell <2.5V

(note) both OV and UV go to GND for other errors like over temperature or BMS sensing failure

For OV control, it is directly connected to the sense cables of the charger, and i am adding a buzzer ( utilizing a relay on the right side), so I can hear when there's an OV condition. This should make the charger stop charging in an OV event.

For UV control, i decided to connect two relays ( to the left of the BMS system) and then the latching relay: One Normally Closed Relay that, when actuated, it opens the circuit. If there are normal conditions on the BMS, this relay would be OPEN. IF there is an Under Voltage condition, this relay would close the circuit and activate the 2nd relay (a flashing relay), which would send the signal OPEN for a few seconds ( to the orange connection) to the latching relay. THis would OPEN the main connection


All the relays, bms, and all loads are connected after the latching relay, if an Under Voltage event occurs, all the loads get disconnected, including BMS, RElays, and everything. I added to the diagram a couple of override switches to manually connect the bms and the main latching relay for diagnosis / repairs.

Please let me know your feedback, below is the complete diagram ( i ommitted fuses and ground connections for the sake of simplicity), including all the components that would need the bms to be operable.
Attachment 124318
Attachment 124490
I welcome your feedback on every error / issue / pottential issue that you see in my design.

Several thanks in advance for your kind help!!!!!!

Hi all, please bear with this newbie. I'm quite excited about my project and I feel like I'm in the right path- please see & validate my circuit and if you see issues or want to suggest one or two points, please do!!


Thanks in advance ¡!!!

[offpist]

I think the "solution" when it comes to charging LIFEPO4 with alternator is:
-Two alternators.
-Fit as large oversize alternators as you can
-Limit them at 60-80% output.
-Make sure your external regulator senses the voltage directly on the battery, AND senses the temp of the alternator.
-Let the BMS disconnect your regulator if an High Voltage event should happen. Then the alternator is shut down in an safe manner.
-Make sure you have good ventilation of engine room.

I run 1 pcs Balmar AT 200A and 1 pcs Balmar 150A on Serpentine belts.
If I run them at 230A combined the alternators never reach more than 95`C even after many hours.

I use two Mastervolt Alpha Pro MB regulators.

[ColdEH]

Or get an Eco-tech alternator and couple it to the the crank shaft and forget about it .

https://m.youtube.com/watch?v=yVvBCat9qsw

Regards


Sent from my iPhone using Cruisers Sailing Forum

[Anjin]

Originally Posted by lagoon100

Hello for a 1200ah bank i have the following question:

1. Would you make 2 Banks of 600ah each and parallel them each whit its own BMS, or would you do 1 Banks of 1200ah and one bms (each cell is 200ah)?
Some say its better to have a big bank as cells are more stable and is difficult to reconnect the Banks if even just out by a few mili V as the current for equalizations would be to big.
Others say its better to have more Banks in parallel so as it is easier to identify a week set of cells and finally the weak cell, i tend to go that direktion, but not sure?

I'm replacing the 840ah bank on my Lagoon 450 with a single 1440ah LiFePO4 bank. There was a very good discussion earlier on this thread about two banks vs a single bank (I believe it was led by Ocean Planet) but the focus was more on operational advantages rather than technical benefits - on using one as a backup in case of issues as well as charging strategies.

I was tempted to go in that direction for those reasons but ultimately decided against it just because it added complexity to an already intimidating upgrade. And BTW use caution with the word "cell" - the "cell" you purchase will come from the manufacturer with an ah rating but you may parallel those into a 3.2v "cell" that will have the ah rating of your bank, and then connect those in series to form the bank. There may be more precise vocabulary but I haven't picked it up.

Originally Posted by lagoon100

2. Would you fuse whit-in the Bank, in the parallel strings (3P or 6P), some suggest that, i’m unsure as how much to fuse and if would it reale work if one cell ges bad as the current from the bad cell would have to be quite big in the failure … would it reale be big enough to pup a fuse.

Again, I'm an amateur, but I haven't heard of anyone expressing the need for fusing at the 3.2v level within the bank. I think of fusing as protecting against a short that would try to draw too much current from the bank. But by all means, protect your battery terminals so that dropping something on them won't cause a short - lots of energy stored in these batteries.

My best advice is to find a smart local electrician (mine is the local supplier for Victron), buy your stuff from him and use him for advice and maybe pay him for some consulting or some real help in design, assembly, and test. As others have pointed out (especially Maine Sail), this is not really ready yet for DIYers on their own.

Cheers -

- Art

[exMaggieDrum]

Originally Posted by Anjin

And BTW use caution with the word "cell" - the "cell" you purchase will come from the manufacturer with an ah rating but you may parallel those into a 3.2v "cell" that will have the ah rating of your bank, and then connect those in series to form the bank. There may be more precise vocabulary but I haven't picked it up.
- Art

There is a terminology for that. A 1P4S battery is one set of 4 cells in series to make 13.2V at the same Ah as each individual cells.

2P4S is two sets of 13.2V cells making 2x the Ah as one individual set.

3P4S is three sets at 3x Ah.

1P8S is one battery at 24V at same Ah of each cell.

2P8S is two batteries at 24V at 2x Ah, and so on.

[Maine Sail]

Originally Posted by exMaggieDrum

There is a terminology for that. A 1P4S battery is one set of 4 cells in series to make 13.2V at the same Ah as each individual cells.

The "P" stand for parallel the S for series. What you describe is really just a 4S pack (four cells in series). You don't need the 1P in there as there is really no parallel...

[Anjin]

Originally Posted by exMaggieDrum

There is a terminology for that. A 1P4S battery is one set of 4 cells in series to make 13.2V at the same Ah as each individual cells.

2P4S is two sets of 13.2V cells making 2x the Ah as one individual set.

3P4S is three sets at 3x Ah.

1P8S is one battery at 24V at same Ah of each cell.

2P8S is two batteries at 24V at 2x Ah, and so on.

Thanks, I'm familiar with the xPyS terminology but I think there is still some confusion in how people use the terms cell, primarily because cells wired in parallel behave as a higher capacity single cell. Examples that come to mind - the housepower BMS describes their MiniBMS "Cell" Module as "Get one cell module for each cell or for each group of cells connected in parallel if your pack contains groups of smaller cells in series/parallel configuration" which is correct and agrees with the definition you cited - but it's still called a cell module and it's job in the literature is to monitor the cell. And in discussions on this thread, especially around monitoring cells, the word cell is often shorthand for the group of cells wired in parallel - as in "if you wire the cells in parallel first then in series, you will only have to monitor four cells with your BMS."

Certainly confused newbies like me when I first started reading.

[exMaggieDrum]

Originally Posted by Maine Sail

The "P" stand for parallel the S for series. What you describe is really just a 4S pack (four cells in series). You don't need the 1P in there as there is really no parallel...

Correct. I just used it to show how it works. Hope no one got confused.

[exMaggieDrum]

Originally Posted by Anjin

Thanks, I'm familiar with the xPyS terminology but I think there is still some confusion in how people use the terms cell, primarily because cells wired in parallel behave as a higher capacity single cell. Examples that come to mind - the housepower BMS describes their MiniBMS "Cell" Module as "Get one cell module for each cell or for each group of cells connected in parallel if your pack contains groups of smaller cells in series/parallel configuration" which is correct and agrees with the definition you cited - but it's still called a cell module and it's job in the literature is to monitor the cell. And in discussions on this thread, especially around monitoring cells, the word cell is often shorthand for the group of cells wired in parallel - as in "if you wire the cells in parallel first then in series, you will only have to monitor four cells with your BMS."

Certainly confused newbies like me when I first started reading.

I got confused when ordering the House Power cell monitors (miniBMS) and my head board (what is called a mini BMS by some on here and HousePower BMS by House Power) and also the number of cell monitoring boards required. I was corrected by Dimitri at House Power. I only need 4 miniBMS's for 16 100Ah cells to make a 4P4S configuration at 12v.

Dimitri: "Name “MiniBMS” refers to cell boards, which are the same for all applications. What is different in different applications is a head board, there are 2 of them. For house banks like yours, the head board is called HousePower BMS, which is used in all projects you read about on CF and various blogs. For high voltage packs used in Electric Vehicles there is “MiniBMS Head Board”, the name of which could be confusing as it contains “MiniBMS” in the name. I am sorry for this marketing confusion, it’s a historical development as our BMS was first developed for EVs and then HousePower BMS version was developed later."

Joe

[OceanSeaSpray]

There is no issue with charging LFPs with existing alternators as such. There is a control issue in the sense that there is no proper way of limiting their current output.

Limiting field current, whether using regulator settings or adding a resistor in the field circuit works, but it is a patch only, because if you increase the revs, the current increases and it also reduces output at lower revs, which is not completely desirable.

This is why I built a regulator with current sensing, but for internally regulated/externally sensed alternators like the Mitsubishi 115A found on the new Volvos.
Making a version that replaces the regulator altogether and truly limits the alternator current is a thought that has been around, but there is a fair amount of work involved in developing and validating this kind of hardware. It would truly settle the LFP vs alternator question however, because you could then preset the maximum current and really keep any alternator within its capability limits.

Some of the existing regulator design (sensing, CPU etc) could be reused as-is, but changes would be needed to manage the field current instead of synthesising a voltage for the alternator to sense like now and the firmware would need to be adapted as well.

[Singleprop]

This link describes a way to limit alternator current (I think its a Balmar 612 regulator):

Alternator Output Management with External Regulators [Small Engine Mode]