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

[OceanSeaSpray]

Originally Posted by klaus53123

Thanks for the link. I have already cross read the article. This is the source of my question concerning the alternator. However: this is the first time I am hearing this kind of advice. Would be great if others could comment on this also. I am concerned :-( to spend another few hundred bucks unexpectedly.

Thanks
Klaus

Klaus,

There are a few different "good" ways and many have been used by people here:

1/ Use an alternator with an adjustable regulator and lower the voltage. How low is a question of how long you run your engine etc. You can also add a manual switch to turn it off if you motor for a really long time and it opens options.
I would suggest 13.5V if you run the engine for reasonable periods, based on some of my own tests, and 13.6-14.0V if you use it for charging in a cycling mode depending on engine run times, higher voltage for short runs at higher C-rates.

2/ Use the method proposed by Rolf (see earlier posts with full explanations) and switch the sensing point of the alternator from downstream of a diode isolator to upstream, effectively lowering the voltage. Your alternator needs to be externally sensed, or you need to extract the voltage sensing input and add a bit of gear around.

3/ Forget about switching the sensing point or externally sensed alternators. Just charge through a (diode-type, not fancy MOSFET) battery isolator and lose around 0.6V; most conventional internally regulated alternators should be fine. It is less efficient as method 2 though and a side-effect is that it will probably somewhat limit your alternator current, but that will also keep it cooler. The overall result depends on its internal regulation point (14.2-14.3V is quite common) and how much you lose over the blocking diodes at high current. If it works for you, it is the easiest, cheapest and most reliable and you can still add a on/off switch if you run the engine for a very long time.
Note that some alternators have 2-step built-in regulators. If you own one of those, it will charge up to its target voltage, hold for a little while and then step down, which almost equates to turning it off. This is a very useful feature.

4/ Use an external programmable regulator. When it is already there, it is straightforward, otherwise it is costly and involves hacking the alternator further. You can also lose alternator over-temperature protection and you MUST disable any battery temperature compensation scheme. Set it to bulk charge up to 13.6-14.0V max with near-zero absorption time. Lower C-rate and/or long engine runs call for lower voltage.
Reliability is not always great, the ones we hear about the most are also the ones I have seen blown most often: they are expensive and you don't always buy them once either.

Overheating: problem is worst at fast idle, when the alternator reaches full output and fan speed/air flow is low. You can de-rate any alternator by adding a power resistor into the field circuit (brushes/slip rings, which you need to access anyway for over-voltage protection/disconnect). 1-2 Ohm is usually enough. It will still reach full output, but now at higher RPM with more cooling. The power resistor should be rated for 10W or so and mounted on some kind of heatsink.
I like to use two 1 Ohm resistors and then I can test with one (1 Ohm), two in parallel (0.5 Ohm) or two in series (2 Ohms). You can also select a 1 Ohm and a 2.2 Ohms if overheating is serious.
Limiting field current reduces the torque absorbed and output current and you can get that for a few bucks only.

The first step with alternators is always working out what they do: grab a multimeter, start the engine and monitor the output voltage at the alternator terminal. You want to figure out the regulation voltage (stable maximum) and whether it just holds it or steps down after a while. This can take 20 minutes sometimes, no rule.
The second step is working out whether it is internally or externally sensed. If you have splitting diodes in the circuit and you read an alternator regulation voltage around 14.8V, it means it is being externally sensed at the battery and you will need to reconfigure that. In this case, you should measure the regulation voltage at the battery instead.

[klaus53123]

Originally Posted by OceanSeaSpray

Klaus,

There are a few different "good" ways and many have been used by people here:

1/ Use an alternator with an adjustable regulator and lower the voltage. How low is a question of how long you run your engine etc. You can also add a manual switch to turn it off if you motor for a really long time and it opens options.
I would suggest 13.5V if you run the engine for reasonable periods, based on some of my own tests, and 13.6-14.0V if you use it for charging in a cycling mode depending on engine run times, higher voltage for short runs at higher C-rates.

2/ Use the method proposed by Rolf (see earlier posts with full explanations) and switch the sensing point of the alternator from downstream of a diode isolator to upstream, effectively lowering the voltage. Your alternator needs to be externally sensed, or you need to extract the voltage sensing input and add a bit of gear around.

3/ Forget about switching the sensing point or externally sensed alternators. Just charge through a (diode-type, not fancy MOSFET) battery isolator and lose around 0.6V; most conventional internally regulated alternators should be fine. It is less efficient as method 2 though and a side-effect is that it will probably somewhat limit your alternator current, but that will also keep it cooler. The overall result depends on its internal regulation point (14.2-14.3V is quite common) and how much you lose over the blocking diodes at high current. If it works for you, it is the easiest, cheapest and most reliable and you can still add a on/off switch if you run the engine for a very long time.
Note that some alternators have 2-step built-in regulators. If you own one of those, it will charge up to its target voltage, hold for a little while and then step down, which almost equates to turning it off. This is a very useful feature.

4/ Use an external programmable regulator. When it is already there, it is straightforward, otherwise it is costly and involves hacking the alternator further. You can also lose alternator over-temperature protection and you MUST disable any battery temperature compensation scheme. Set it to bulk charge up to 13.6-14.0V max with near-zero absorption time. Lower C-rate and/or long engine runs call for lower voltage.
Reliability is not always great, the ones we hear about the most are also the ones I have seen blown most often: they are expensive and you don't always buy them once either.

Overheating: problem is worst at fast idle, when the alternator reaches full output and fan speed/air flow is low. You can de-rate any alternator by adding a power resistor into the field circuit (brushes/slip rings, which you need to access anyway for over-voltage protection/disconnect). 1-2 Ohm is usually enough. It will still reach full output, but now at higher RPM with more cooling. The power resistor should be rated for 10W or so and mounted on some kind of heatsink.
I like to use two 1 Ohm resistors and then I can test with one (1 Ohm), two in parallel (0.5 Ohm) or two in series (2 Ohms). You can also select a 1 Ohm and a 2.2 Ohms if overheating is serious.
Limiting field current reduces the torque absorbed and output current and you can get that for a few bucks only.

The first step with alternators is always working out what they do: grab a multimeter, start the engine and monitor the output voltage at the alternator terminal. You want to figure out the regulation voltage (stable maximum) and whether it just holds it or steps down after a while. This can take 20 minutes sometimes, no rule.
The second step is working out whether it is internally or externally sensed. If you have splitting diodes in the circuit and you read an alternator regulation voltage around 14.8V, it means it is being externally sensed at the battery and you will need to reconfigure that. In this case, you should measure the regulation voltage at the battery instead.

Many thanks for this detailed explanation. Just for clarification:

I have a Sterling Digital Pro external regulator installed. I can adjust the max output voltage to 14.1/14.4/14.8V. Floating voltage is than 13.5/13.65/13.65 V. The sterling has an external voltage sensor which i than just connect downstream to the diode which will drop the Voltage by around 0.6 V. Which voltage is best to adjust at the Sterling (14.1, 14.4 or 14.8).

My main mean of charging is solar power. I have MPPT chargers installed suitable for LFP adjusted to 14.2V main charge and 13.5V float.

Same values as solar will be adjusted to my Victron battery charger from land power (but very rarely used :-))

I don't fully understand how Rolf manages to change the voltage sensor from downstream diode to upstream diode. There must be a sensor device installed that creates the signal for the 5$ relais. Am I wrong????

Best regards
Klaus

[roetter]

LiFePo cells like to be charged to a certain voltage and then stop charging completely. Very simple.

I use a programmable voltmeter with 2 relays from LightObject (about $40).
I program the HVC (load bus) to disconnect at 13.8. Reconnect at 13.3.
I program the LVC (charge bus) to disconnect at 12.0 and reconnect at 12.4.

The two small relays in the voltmeter then drive Blue Seas contractors.

Very simple, fancy rework on any of the chargers. I still program what I can (solar, short charger) to a just slightly higher voltage than the cut offs.
The alternators I leave as they are. When the charge bus disconnects the starter batteries absorb the spike. I am on the boat when the alternators are running, and should the HVC ever fail I will get an alarm from my $20 Junsi with a $3 buzzer. The Junsi also records all cell voltages every 2 seconds and I can read that out on my PC every few weeks.


Rolf

[senormechanico]

roetter,

You said,

"I program the HVC (load bus) to disconnect at 13.8. Reconnect at 13.3.
I program the LVC (charge bus) to disconnect at 12.0 and reconnect at 12.4."

I think you meant HVC (charge bus)
and LVC (load bus)

right?

[klaus53123]

Originally Posted by roetter

I use a programmable voltmeter with 2 relays from LightObject (about $40).
I program the HVC (load bus) to disconnect at 13.8. Reconnect at 13.3.
I program the LVC (charge bus) to disconnect at 12.0 and reconnect at 12.4.



Rolf

Rolf, does that mean, you just let the charger run as they are and control cut off point with the voltmeter/relais? What happened to your Diode solution described earlier in the threat? Do you use no BMS?

Other question, probably already discussed earlier here, but 4k posts + is just too much to follow up with:

is 14.2 Volts too much???? My supplier tells me I can charge up to 3.65 V per cell (14,4 Volts) with no problem whatsoever. Even 3.8 V not too problematic!

By the way, I believe I found the programmable Voltmeter. Sells at 39$. Here in Europe more than twice the price.....

BR
Klaus

[roetter]

Originally Posted by senormechanico

roetter,

You said,

"I program the HVC (load bus) to disconnect at 13.8. Reconnect at 13.3.
I program the LVC (charge bus) to disconnect at 12.0 and reconnect at 12.4."

I think you meant HVC (charge bus)
and LVC (load bus)

right?

Had no coffee this morning an hanging around noisy airport. Is shows. You are correct. HVC is Charge Bus, LVC is Load Bus

[roetter]

Originally Posted by klaus53123

Rolf, does that mean, you just let the charger run as they are and control cut off point with the voltmeter/relais? What happened to your Diode solution described earlier in the threat? Do you use no BMS?

Other question, probably already discussed earlier here, but 4k posts + is just too much to follow up with:

is 14.2 Volts too much???? My supplier tells me I can charge up to 3.65 V per cell (14,4 Volts) with no problem whatsoever. Even 3.8 V not too problematic!

By the way, I believe I found the programmable Voltmeter. Sells at 39$. Here in Europe more than twice the price.....

BR
Klaus

I have two boats and plyed with different things. I did the mono first and was still into doing all kinds of complex stuff.
The diode thingy is on my mono. The mono is just for some weekend trips in the PNW.

On the cat we live for 6 months of the year in The Bahamas. Here, I just let the alternators do their thing and disconnect them when reaching the HVC. Shore charger (alsmost never used) programmed to 13.8V with sensing at the battery. Solar programmed to 14V, witth no sensing at the terminal, so there is a drop on the cables of about 0.2V at 45A. I have two solar banks which provide 40+A peak every day, some days 50+A.

The safe charge voltage depend very much on how many amps you are pushing in.

If you are charging with 1C - i.e. 600A on a 600Ah bank, then high voltages are fine, as you are spending very little time in the high voltage electrode plating field. Probably just 2-3 minutes. However, is you are charging a 600Ah battery at 30A then you would be spending over an hour in the high voltage electrode plating area.

So, you can not recommend a charge voltage for lithium without the C-rate that you are chrging with. As soon as you start hitting the upper knee (i.e. 13.4V+) at low charging rates your cells are 90+% full. If you switched all of a sudden to 1C charge rate the voltage would immediately jump to 15+V, as the internal resitance of the cells is going up at this SOC.

[OceanSeaSpray]

Originally Posted by klaus53123

Many thanks for this detailed explanation. Just for clarification:

I have a Sterling Digital Pro external regulator installed. I can adjust the max output voltage to 14.1/14.4/14.8V. Floating voltage is than 13.5/13.65/13.65 V. The sterling has an external voltage sensor which i than just connect downstream to the diode which will drop the Voltage by around 0.6 V. Which voltage is best to adjust at the Sterling (14.1, 14.4 or 14.8).

My main mean of charging is solar power. I have MPPT chargers installed suitable for LFP adjusted to 14.2V main charge and 13.5V float.

Same values as solar will be adjusted to my Victron battery charger from land power (but very rarely used :-))

I don't fully understand how Rolf manages to change the voltage sensor from downstream diode to upstream diode. There must be a sensor device installed that creates the signal for the 5$ relais. Am I wrong????

Best regards
Klaus

Klaus,

The alternator regulator should be reasonably fine on its lowest setting, especially since your main source is solar. Several manufacturers have now stopped making crazy charging voltage claims, but remember there is a big difference between charging hard to a high peak voltage and then discharging right away, or charging at low C-rates and keeping reserve capacity. It is not just a question of voltage, but also time spent at voltages of 13.6V and over.

The solar regulators would worry me much more in terms of battery life. They are likely to do some absorption at 14.2V before stepping down and will push your bank to 100% all the time. Unless the regulator can charge and cut off at about 13.8V, I consider that 13.50-13.55V is the maximum that can be sustained with CV charging. For me, 13.55V leads to ~90% SOC by the end of the day and if you don't use any power, it is likely to keep creeping up very slowly the next day and so on.
It really needs to cut off, either completely or by stepping down to a voltage matching what the battery would supply on discharge at low current (13.32V).

This problem keeps coming back and I have now hacked several regulators to make them control at lower voltages and help people out. I can't keep doing that.
I have sketched a few things around building a decent LFP solar controller for house banks, but one thing at a time. Right now I am finishing up a dedicated 4-cell battery protection module. We have installed a few so far driving bi-stable charge and load power disconnectors and it is all coming together, but it needs proper documentation etc.

I like having the protection system for protection and have regulators that regulate as they should, so I treat disconnecting mechanically as a last resort, but blending the two together can make life easier when building a new system.
I have been thinking about maybe programming a mode of operation that would allow that with a no-alarm HV disconnect to first get things going in some cases when regulators are inadequate.
The question is whether the system should protect against events that would destroy the battery or protect against more general situations where battery life would be reduced over time. Maybe the end-user should be able to decide.

Regards,

Eric

[klaus53123]

I have no caught up reading till page 234 and things become a bit brighter. There is light at the end of the tunnel :-)

thanks Rolf and Eric for your comments. They are greatly appreciated.

I have about 600Wp Solar panels on my boat so on a nice day they provide me max 35A. My battery bank will be 400Ah so with Solar I am miles away from 0.5C charge. My two alternators have design capacity of around 60A each so I will get closer there.

Just checked the "curves" for LFP charge of my solar chargers:
"Main Charge" = 14,2 V 0,5-3hrs
"Float" = 13,5 V permanent

I am not too much concerned about getting the batteries really full full. 90% will be enough I now have 300Ah LA installed means less than 50% of the available capacity compared to my future 400Ah LFPs.

Cutting off the Solar and switching it back on as described by Roetter is pretty straight forward. What causes me some headake is the alternator. Can I just cut off the +charge coming from the alternator or will this blow off my diodes???

[transmitterdan]

Originally Posted by klaus53123

Cutting off the Solar and switching it back on as described by Roetter is pretty straight forward. What causes me some headake is the alternator. Can I just cut off the +charge coming from the alternator or will this blow off my diodes???

Turning off the field voltage to the alternator is the safest way to stop charging. It won't blow the diodes and does not require a high current relay. Most field currents are less than 20A. Some are less than 10A.

Opening the main DC output from the alternator to the battery bank without also turning off the field current first will definitely blow up the diodes.

[garrobito]

I starting to design my own battery pack lithium-based.
All information on this threat help me a lot..
As engineer I can understand perfectly the slightly differences between chargers, BMs and the chemical involve on these systems.
Thanks you! For everyone looking for basic/advance information Basic to Advanced Battery Information from Battery University is an excellent reference... People building electric cars mostly time build they're own lithium pack too.
Unless somebody have a very good connection with Tesla on California and can get a Tesla (Panasonic) battery pack..

[klaus53123]

Originally Posted by transmitterdan

Turning off the field voltage to the alternator is the safest way to stop charging. It won't blow the diodes and does not require a high current relay. Most field currents are less than 20A. Some are less than 10A.

Opening the main DC output from the alternator to the battery bank without also turning off the field current first will definitely blow up the diodes.

****, I knew this will be the answer. Why can not at least something in life be easy

My alternator has the following connections:

B+
B-
W
D+

I assume D+ is the Field. Correct????

Next life I will definitely become an electrician or marry a very rich women to be able to pay for electricians :-)

[OceanSeaSpray]

Originally Posted by klaus53123

****, I knew this will be the answer. Why can not at least something in life be easy

My alternator has the following connections:

B+
B-
W
D+

I assume D+ is the Field. Correct????

Next life I will definitely become an electrician or marry a very rich women to be able to pay for electricians :-)

B+ is your output, don't disconnect that ever. B- is the isolated negative (nice), don't touch that either. W is an AC output for a tacho (RPM).
D+ is the connection for the alternator light. You need power on D+ to get the alternator to kick in, initially, and then you find the same voltage on it as B+ while it is running. Disconnecting D+ on a running self-excited alternator does nothing.

The field is at the brushes/slip rings. Disconnect one of these wires (either) and run the circuit through a small relay. Opening the field circuit disables the alternator 100% guaranteed no matter what else might have failed. There is only a few amps in that circuit.

However, you were saying you have an external regulator. In this case, someone must have tapped into the field connections already and you can break that circuit.

By the way, not charging to 100% is precisely the objective for battery life. The alternator side will come right quite easily, but watch out the solar panels.

Eric

[OceanSeaSpray]

Originally Posted by klaus53123

...
Just checked the "curves" for LFP charge of my solar chargers:
"Main Charge" = 14,2 V 0,5-3hrs
"Float" = 13,5 V permanent

I am not too much concerned about getting the batteries really full full.

That sounds like a Votronic regulator. It will recharge to full and then the 13.5V setting is going to hold you there until the sun goes down. It is not very good news at all on the longer run.

[roetter]

Some people seem to wonder how I have done the "before/after" voltage sensing for the alternator.

Requirements:
a - alternator that allows remote sensing. On the Volvo on my mono I have a Mitsubishi that has remote sensing. On my cat the Yanmars have Hitachi without remote sensing, so another solution was chosen.
b - You need true old fashioned diodes. These produce approximately a 0.7V drop. They cost efficiency, but that is what my mono had and I did not want to change it. I try to keep things as simple and cost effective as possible.

Here a schematics. I hope I made no mistakes. I am jet lagged at an airport in Europe now.

Attachment 95559

The Automotive relay has two positions - Normally open and normally closed.
Attachment 95560