Lithium Battery Confusion

[Scalf77]

Great stuff here guys. Is leakage or just the fact the charge path disconnected, and discharge path not. Also, did your previous solar controller have any high voltage disconnect settings? Does this BMS give any external control signal when it disconnects.

To the original poster, when looking for a lithium replacement, I would tell them the equipment you have and see what data they have. The reputable ones will try to help you. I would also look at somebody that batteries is their core business. A big difference in a somebody like Battleborn and Renogy when it comes to lithium batteries.

-greg

[posplayr]

Originally Posted by cotrailhead

also if you connect all your 12 volt devices to the load output they would be protected. But the majority of users connect there devices directly to the battery.


This is the part of the post that confuses me. The bms should live between the battery and all discharge/charging devices. If the bms shuts down all discharging, but there is a load connected to the battery directly, your battery is still in danger of over discharge.

I assume he is contrasting a protected load connection at the solar controller vs a direct connection to an integrated battery system where the bms is part of that integrated battery. So nothing is “going around” the bms

[posplayr]

Originally Posted by Scalf77

Great stuff here guys. Is leakage or just the fact the charge path disconnected, and discharge path not. Also, did your previous solar controller have any high voltage disconnect settings? Does this BMS give any external control signal when it

-greg

If the are in fact charging and discharging moSFETs and a over voltage charging disconnect happens then I don’t know why you should see anything but something close to the cell voltage as measured through the discharge mosfet.

This would not appear to be what Jonyjoe is describing; he seems to have a complete disconnect.

[jonyjoe101]

This would not appear to be what Jonyjoe is describing; he seems to have a complete disconnect.

More of an incomplete disconnect, current won't flow through the bms, but the bms still shows a voltage reading of 11.7 volts. The controller see's this voltage and tries to forcibly charge the battery, resulting in voltage surges when the battery won't accept the amps.


Batteries like battleborn and similar will encounter these situations if the bms is activated on those batteries, thats one reason they recommend a bulk charge of 14.4 volts, even though lifepo4 can handle 14.6 volts.


One thing about charging lithium with solar, is voltage drop, all controllers I owned have at least a .5 volt difference. If the controller reads 14.4 volts, the battery terminals read 13.9 volts. In this situation the controller thinking the battery is full reduce the charge amps to about 2 amps, so battery takes a long time to fully charge. By raising the bulk voltage it increases the charge amps, but also has a better chance of activating the bms since you will be fast charging the battery. I learned about this voltage drop when I switched from lead acid to lithium and my lithiums werent getting fully charge. Increasing the bulk voltage got my lithiums charged faster but it also brought occasional voltage surges.

[posplayr]

Originally Posted by jonyjoe101

More of an incomplete disconnect, current won't flow through the bms, but the bms still shows a voltage reading of 11.7 volts. The controller see's this voltage and tries to forcibly charge the battery, resulting in voltage surges when the battery won't accept the amps.

jonyjoe, you describe the 11.7V as a MOSFET leakage voltage, but under a charging disconnect, you should still see the battery voltage to support discharge. If the BMS is still allowing discharge you are going to essentially see full battery voltage less a body diode forward drop (see more on this below). I can't reconcile this 11.7V except for a badly (over-voltage) imbalanced cell based on the following analysis.

That said, I can certainly see how, with a charge disconnected BMS, an overzealous charger (solar, DC-DC or 120VAC-DC) can push up the terminal voltage well above 15V with a constant current charging supply.


Analysis:
So I finally put the effort in this morning to try and figure out what basic topology a BMS is using to control MOSFET charge and discharge. If anybody has any expertise in power electronics maybe they can chime in. (This is not my area of expertise by a long shot). This is not a comprehensive tutorial, but rather hits the main high points of what is going to be likely BMS behavior.

First what we know should be common for all LiFePO4 BMS:

1. Charge and discharge imply that there is a reversal of currents at the terminal of a BMS.
2. We will assume Power MOSFET control.

This link provides a variety of approaches for using transistors in reverse (i.e. Assumption #1 above)

https://www.reddit.com/r/electronics...ors_backwards/https://www.reddit.com/r/electronics...ors_backwards/

Power MOSFETS (i.e. see assumption #2) have what is known as a "body diode" that are intrinsic to the device and are anti-parallel diodes pointed in the opposite direction to the normal Drain-Source current flow. This means that even when the gate of OFF, if the device is reverse biased, current will flow backward through the body diode.

The link provides an example of Two enhancement MOSFETs in anti-series

The schematic explicitly shows the intrinsic "back to back" body diodes of the anti-series MOSFET pair. When both MOSFET gates are OFF the body diodes block flow in either direction. The MOSFETS can be controlled ON individually which effectively short out their respective body diodes allowing current to flow in their normal respective direction by going through the reversed Body diode of the complementary MOSFET.

So to corroborate jonyjoe's observation about 0.5V drops, this would be the nominal voltage drop across one of the body diodes in the anti-series pair.

You will see that this author also shows a 1M ohm resistor to the ground between the two MOSFETS. This will tend to pull the voltage to the ground if the center tap is not being driven from either side. So in the case of a high voltage over charging disconnect, M1 (on the left) would be turned OFF (leaving the M1 body diode to block incoming (from charger) current.

Since M2 (allowing discharge) and is still ON (shorting out body diode D2), the voltage observed at the charger input terminal will be whatever nominal voltage is at the battery. This is because even though there is a body diode D1 if the load current is very low then that drop across the diode will be essentially zero.

If it there were so significant load current then the D1 voltage drop would be closer to 0.5V. So if we assume the charger is a Solar controller, if under the disconnect the solar controller is still providing battery current to 12V DC loads, then there will be a voltage drop corresponding to the IV curve of D1.

jonyjoe

Regardless the charger should see V_batt - V_D1 at the terminal to the integrated battery. This is where I can not reconcile your measurements, which you report as always seeing 11.7 volts under a BMS over charge disconnect.

The only possibility(i can see) is if there is An unbalanced condition inside the battery and it is hitting an overcharge disconnect at the same point (on a particular cell) where the overall battery is well below fully charged. If I add 0.5V to the 11.7 you are at 12.2V which would indicate a severe imbalance in one cell to cause the disconnect. I'm sure you will object to that theory, but it is the only one I see to explain your observations.

More on MOSFETs

https://www.electronics-tutorials.ws...or/tran_6.html

[campIV]

Lots of very good information here. It does however sum up why I replaced my lead acid house battery with an AGM rather than Lithium. Same solar controller, same alternator, same battery separator, same wiring, and a much less expensive battery. Works perfectly for my needs. I’m a big fan of the KISS school of thought.

Curious to follow and see how this all shakes out tho.

[posplayr]

Originally Posted by campIV

Lots of very good information here. It does however sum up why I replaced my lead acid house battery with an AGM rather than Lithium. Same solar controller, same alternator, same battery separator, same wiring, and a much less expensive battery. Works perfectly for my needs. I’m a big fan of the KISS school of thought.

Curious to follow and see how this all shakes out tho.

There are definitely potential problems by treating all components as commodity items and trying to mix and match all the manufacturer's products including a bevy of (less than well-documented) Chinese products.

However, another part of the problem is trying to shoehorn LiFePO4 in as a lead-acid replacement. It is much more than that as the energy densities and costs improve. A case in point, from what I am seeing you should not have to change alternators, but if you want high-performance charging (40A+) you are going to have to manage the alternator thermal issues.

For example, Renogy's DC-DC charger suggests that if you are delivering 40A to the battery, then you need to size your start battery wiring for 60A. This if due to the effect of stepping up the voltage from saying 12.8V to 14.5V and the 80% efficiency of the DC-DC.

You need a higher input current to deliver this higher voltage output current. In order to create the 60A*12.8V=770 watts, the alternator has to dissipate another 770 watts due to a 50% efficiency. So this cascading series of factors is being caused by the LiFePO4 batteries' increased capability, but that is what you have to pay for with increasing performance.

Let's say you had 200Amp-Hr of lead-acid, then with C/10 charging you would not need more than 20 amps of charging and this will not stress your alternator in almost any condition. Do you want to charge at 40A for LiFePO4? OK now you have to use a little more finesse because you will now need to supply 60Amp in low RPM situations.

For an example of the paradigm change, I am seeing more and more weekender solutions that are non-solar. With a 40A DC-DC charger you charge a DIY 320A-Hr battery bank on the way to your camping destination and you have enough power for a 4-day weekend. You don't need solar. No rooftop panel and no solar controller. An oversized cheap battery gives you this. With retail battery prices are dropping to the $500/100Amp Hr range this is turning into a viable off-the-shelf solution.
Here is a review of the new SOK 100 Amp-hr which you can get delivered for $570. This is apparently a high-quality battery and much cheaper than the "top of the line" Battleborn.

https://www.youtube.com/watch?v=RjpkI8quyzQ

My point is that part of what we are seeing is an incompatibility of treating LiFePO4 as a plug and play for lead-acid because LiFePO4 is really a game-changer and if you are starting fresh you can re think the whole system design.

I don't know about its quality but breaking the lead-acid drop-in replacement paradigm is a 200Amp-Hr $800 LiFePO4 battery. With Ampere Time 200 Amp-Hr battery below, why constrain yourself to lead Acid drop-in replacements?

https://www.amazon.com/Lithium-LiFeP...NsaWNrPXRydWU=

We are not too far away from 24V/48V LiFePO4 battery packs that can be charged pre-trip or en route and deliver enough power for microwave and induction overs systems without needing 0/2 wire connections of 12V. Once the marketplace catches up there will be more battery chargers at higher voltage levels. This is the only one I found on a quick google search. This has the equivalent power but cost almost 3 times the Renogy 40A@12V units.

https://redarcelectronics.com/produc...attery-charger

With all the Electric vehicle tech, the RV market will benefit from increased performance battery systems and Solar already is "optional".

Your decision to replace AGM rather than go LiFepo4 is really one about postponing an electrical system upgrade. There is nothing wrong with that tact. The next generation of products will be more integrated and provide for higher voltage systems such as this Renogy. This would be a potentially better product if they dropped the solar charging and made it 12VDC to 48VDC charging.

https://www.renogy.com/48v-3500w-sol...BoC1W0QAvD_BwE

We have probably way past the point where you should install 100W 12V solar pannels on your RV LOL. Higher voltage means lower current levels.

[Scalf77]

Part of the problem with entry into LiFePO4 batteries in the beginning was cost, the cost of the battery itself, and the cost of equipment that said it was supported LiFePO4. In the beginning there were few LiFePO4 suppliers, but soon you started to see newcomers who often stated they were drop in replacements. While I wouldn't really consider them drop in replacements, many of embarked down that path.

The first thing was the standard float voltage, once they could show that their battery wouldn't be hurt by float, they had knocked down a big hurdle. The introduction of the Battery to Battery charger also was a big driving issue, As this gave them a way to safely charge via the vans alternator.

The next big thing was the charge profile, at this time a
LiFePO4 charge profile was pretty much defined as a two stage profile, pretty loose on any other requirements.

Now they played with existing profiles seen on most standard 3 stage chargers, they new they had to error on the side of not overcharging, such as the 14.4 volt limit, and a hard top level float voltage. They are a little more vague on their cycle lifespan numbers, but I would suspect that they lost some of that by supporting different profiles, probably a little capacity also. But hey, people were able to move into LiFePO4.

I have seen devices that support lithium pretty much have just have a two stage profile with an absorption voltage. I also have a solar charger that was pre- lithium that I can set to a very precise lithium profile. The other big difference in charging profiles is absorption time, LiFePO4 generally needs a much shorter absorption time, there are very few devices that let you program in absorption time. This again is one of the reasons for the 14.4 cut off.

Care needs to be taken when trying to drop in large LiFePO4 replacements, as they are many times still limited current wise by the BMS. This is important if you are a large inverter user, as the current requirements may exceed the BMS limit.

I was this close to going to 24 or 48 volts for my current rig. the only additional big cost would be a DC to DC (24 to 12, 48 to 12) converter to run the 12 volt equipment in the rig. It would be my personal preference to stick with a company, that batteries is their main product offering. You tend to find a bigger support group, and more data on installations. You may find the exact equipment that you have, has been installed and is working with their battery.

note: There are a few companies that I would also purchase their branded battery, only because I know who makes it, or their reputation precedes them in the industry.(Victron)

-greg

[posplayr]

Originally Posted by Scalf77

Part of the problem with entry into LiFePO4 batteries in the beginning was cost

-greg

Times they are a-changing...............