Advice on Lithium Solar setup

[chuck3]

Hello,

I'm about to move into a conversion van full-time, 80% boondocking, and would like your advice about what I'm planning. My electrical requirements won't be much. I don't expect to use more than 400 AC watts or it's DC equivalent at any one time, and normal usage will be less than 200 AC watts or it's DC equivalent for 9 hours a day.

Currently I have 2 x 100w flexible solar panels and a 600w pure sine inverter.

I'm planning on getting a 100ah lithium (LiFePO4) battery ( https://battlebornbatteries.com/shop/12v-lifepo4-deep-cycle-battery/) and a solar MPPT charge controller for it (probably https://www.amazon.com/Victron-SmartSolar-Charge-Controller-Bluetooth/dp/B073ZJ3L13). A big reason that I'd like a lithium battery is because I will occasionally want to bring it with me to a friend's house to charge it from the wall, and 30 pounds in a backpack allows that.

Here are my biggest questions: how will I know when the lithium battery is fully charged, and how will I know how much capacity is used or remaining? With lead batteries it's fairly simple because the volts indicate remaining capacity, but from my understanding, lithium's voltage doesn't change much between it's full and empty states (I could be wrong here).

[Almot]

brulaz wrote:

Battleborn says that while you can charge up their LiFePO4 cells at lower voltages, maybe as low as 3.4V per cell (13.6V for 4), you want to go higher to *balance* the cells, up to their recommended 14.4-14.6V. The balancing is strictly a BMS function and the BMS balancing procedure controls the individual cell voltages.

So what happens if you have only weak solar or alternator V's that never get up high enough to balance the cells? 😞

Every BMS is different. "Normally" it needs voltages corresponding to +90% SOC to get it balanced, some BMS do this at 91% and some at 99%.

This is why it is "sometimes" necessary to charge over 90%, carefully.

This is why you want a charger with easily controlled voltage. Ideally, it should have 2 profiles set - 90% for daily charging and 100% (or whatever is necessary) for balancing once a week or so. There is no need to balance it every day: Balancing

I would also like to choose my charging current and to start at lower current, increasing it later. Something that advanced e-bike chargers do, but automotive 12V chargers don't.

There are many 12V, 36V and 48V boxes packed with 18650 LiFePO4 cells, btw. They don't bother to tell you cells size, only chemistry and approximate voltages.

brulaz wrote:
on another forum there's a guy who said he can still pump 20A into a 300Ah LifeBlue battery after it reaches 100% SOC and his charger's V drops below 14V (EDIT: actually his weak solar charger never gets up that high)

Could be poorly controlled experiment, SOC way below 100%. Hard to tell without knowing the details.

[2oldman]

landyacht318 wrote:
The one pic shown appears to have a bunch of 18650 or some other type of cylindrical size cell.18650 cells are not Lifepo4 but Lithium colbalt or Limaganese.

Battery specs

[landyacht318]

The one pic shown appears to have a bunch of 18650 or some other type of cylindrical size cell.

18650 cells are not Lifepo4 but Lithium colbalt or Limaganese.

Been some blanket statements across the whole lithium spectrum in this thread. LiCo cells are fully charged at 4.2v.

As to the laptops power requirements, I am glad you checked again with full batteries, as 20 to 30 watts is a huge difference from 80 watts continuous. Lowering screen brightness can save a few watts too.

I know many people's solution to powering laptops is an inverter powering the original power brick, but a dc to dc 'car adapter' is much more efficient.

The issue with them is the ciggy plug. They are only good for 60 watts or so continuous. They will eventually fail.

Blue seas makes a much better replacement plug. I opted to use anderson powerpoles instead. My PWR+ dc to dc car adapter uses significantly less wattage powering my laptop, compared to my 400 watt PSW inverter powering the original 120vac to 19.5dc power brick.

It is closing in on 8 years old. The dc output cable has a third wire for my dell which has proven to be fragile, and required replacement more than once though. if the 3rd wire breaks the battery will not charge and the laptop runs slow.

The last time I used a longer but thicker 3 wire bundle soldered from circuit board to barrell connector. It likely receives higher voltage from adapter now.

If possible, save your laptop batteries for use in the morning, so the solar can get the batteries to absorption voltage sooner. Once they reach absorption voltage then not too long afterwards there is solar excess wattage that can be utilized for the recharging of the laptop batteries.

The DC to DC adapter will likely also be more efficient the higher voltage it receives, so keep the circuit feeding it, short and fat despite a 90 watt maximum flow.

12v ciggy plugs and receptacles insult DC electricity. It is a shame they are a worldwide standard.

[Itinerant1]

pianotuna wrote:
Brulaz,

All the reading I've done suggests that LI do not wish to be floated.

I've gleaned that 90% soc is going to give the best results, and that 20% is as low as you wish to go.

The question becomes does the maker "code" that into the battery management system? To me that would make the most sense.

Pianotuna example of my system not drop in batteries but still LiFePo4. CC (bulk charge) is terminated after it reaches (our) 100% SoC which happens at 3.6 volts per cell (Vpc) or 14.4 battery. This leaves ample room below actual 100% SoC so the battery is stressed less on each cycle. It's then held at CV (float) of 3.4 Vpc (13.6). This keeps the battery from cycling. **Per Elite power spec of cell the nominal is 3.2v (12.8). Stress area of the cell is max 4.0 but they recommend not to exceed 3.80V.
So even though our soc is 95% capacity of the actual bank the display has been programmed to show 100%.
With the Magnum PT controller linked to the Magnum inverter/ charger.
The following parameters are for programming the ME-ARC Version 4 advanced control.

Setup 02B LBCO Setting: Volts = 12.0
Setup 03C Battery Type: = Custom
Set Absorb Volts: = 14.2
Set Float Volts: = 13.6
Set EQ Volts: = 14.2
Set EQ Done Time: = 0.1
Setup 03D Absorb Done Time: = 0.1
Setup 03E Max Charge Rate: = 100%
Setup 03G Final Charge Stage: = Float

The BMS has no need to control the Magnum charger because the Magnum has sufficient programability to properly charge the battery. The same applies to the Magnum PV solar charge controller. It follows the inverter/charger programming. If anything happens that would cause conditions beyond programmed limits, our BMS will disconnect the charge source or loads.

I would think the drop in batteries can have basically the same programming if the charger had programmable settings seeing they have a bms to protect the battery

[MEXICOWANDERER]

Trying to outguess battery cells prone to hysteria and paranoia without individual cell monitoring and individual charging is just plain nuts.

I have been playing with a different chemistry lithium battery (3 x 3.7 volts in series) and in order to successfully charge it, I have to gradually increase voltage to 11.2 volts while monitoring individual cell temperature -- and that still isn't the correct way to do it --

It takes programmed software management for both charging and discharging management. An A to D software program and individual cell managers. Automobiles have such sophistication but an aftermarket battery?

Until I see modules with independent individual cell monitoring and management capabilities I intend to keep my distance.

[brulaz]

pianotuna wrote:
Brulaz,

All the reading I've done suggests that LI do not wish to be floated.

I've gleaned that 90% soc is going to give the best results, and that 20% is as low as you wish to go.

The question becomes does the maker "code" that into the battery management system? To me that would make the most sense.

Yes that's the problem. What is the BMS doing?

Battleborn says that while you can charge up their LiFePO4 cells at lower voltages, maybe as low as 3.4V per cell (13.6V for 4), you want to go higher to *balance* the cells, up to their recommended 14.4-14.6V. The balancing is strictly a BMS function and the BMS balancing procedure controls the individual cell voltages.

So what happens if you have only weak solar or alternator V's that never get up high enough to balance the cells? I dunno. But this guy on the other forum may have a BMS (LifeBlue) that's unsuccessfully trying to balance at low voltages. Again, I dunno.

I do like the BattleBorn's BMS though. They seem to have protected the cells against most issues, including low temperature. But still too expensive for me though. 😞

[pianotuna]

Brulaz,

All the reading I've done suggests that LI do not wish to be floated.

I've gleaned that 90% soc is going to give the best results, and that 20% is as low as you wish to go.

The question becomes does the maker "code" that into the battery management system? To me that would make the most sense.

[brulaz]

Almot wrote:

brulaz wrote:

... will get you you to ~95% SOC soon enough and allow cell-to-cell balancing. And just about any 1 or 2 stage charger with an AGM setting, or even an adjustable V Meanwell cc/cv power supply should be able to do that.

... as long as it's able to terminate charging at this point (or, better yet, few % below). These chargers will keep it in Absorb where you may not want to be, and/or drop to fixed Float where it "should" terminate charging, but I recall some reports few pages ago that it doesn't always stop charging in Float. Can anybody provide a link to that discussion, btw?

You want to be able to adjust charging profile. Meanwell looks better in this respect. Perhaps, coupled with $3 kitchen timer.
...

I suppose so. But do you really need to terminate the charge and why?
With lead acid you *do* want to terminate absorb because of out gassing, loss of water and so on.

With LiFePO4, if you fix the CV charge at 3.65V or whatever per cell, the current does eventually drop way down according to all the charts I've seen. So basically a Meanwell could just sit there, applying V but not pumping any (or very little) current into the battery? I dunno ... would that be a problem?

But this may not be what happens, as on another forum there's a guy who said he can still pump 20A into a 300Ah LifeBlue battery after it reaches 100% SOC and his charger's V drops below 14V (EDIT: actually his weak solar charger never gets up that high). A mystery. Where do those 20A go? Feeding the BMS, raising the internal temperature, maybe the SOC estimate is wrong (he doesn't think so)? He ends up manually stopping the charge.

[brulaz]

2oldman wrote:

brulaz wrote:
2Oldman had a couple of Schumacher chargers modified for his 48V LiFePO4 battery bank. Is that 16 cells in series or ?
Curious what V they get up to: 3.65x16=58.4V ?

58.4 is the bulk charge end voltage.

You tell me! I think there's a couple stacks of these cylinders.

Wow, that's a lot of cells.
Since it's 58.4V, I'm guessing some multiple of 16 cells (16s?p) with each cell charged at 3.65V, which seems to be a common standard.

[chuck3]

2oldman wrote:
You mentioned you could us an extension cord with AGM, but you had to take the Li in to a "wall charger"? Is that something permanently attached to a wall?

No, but the extension cord would need to be in the neighborhood of 100ft. Plus, I was thinking it would be better to have it inside so I could watch the charge meter and feel the battery to see if it was getting warm. Not big issues, just preferences.

landyacht318 wrote:

One think I wonder about is the OP's Kill a watt readings.

80 watts for a laptop is fairly high. My laptop is an older dell that requires a 90 watt adapter, but if the battery is already full, it uses about 30 watts to type this.

Turns out you're correct. The readings I took were when the laptop batteries had a nearly empty charge. As the laptop charged, it slowly tapered off on watt usage and once the battery was fully charged, the usage bounced between 32 and 20 watts...much better!

As for your other suggestions, using the alternator is something I'll definitely want to do to supplement the solar. I've seen that the alternator alone can charge house batteries only up to 80%, give or take some, so solar and possibly another source will be needed. I plan on driving every other day for 1-2 hours, and at least 2 days a week it will be 4 hours. As for mounting the panels, I'll use some kind of rack because, as you noted, heat becomes an issue otherwise. Any van should be able to hold at least 400w of panels, and some will be able to hold more. I'm keeping an eye out for extended size. I'll need to build a frame for the flexible panels but that shouldn't be difficult.

[Almot]

brulaz wrote:

... will get you you to ~95% SOC soon enough and allow cell-to-cell balancing. And just about any 1 or 2 stage charger with an AGM setting, or even an adjustable V Meanwell cc/cv power supply should be able to do that.

... as long as it's able to terminate charging at this point (or, better yet, few % below). These chargers will keep it in Absorb where you may not want to be, and/or drop to fixed Float where it "should" terminate charging, but I recall some reports few pages ago that it doesn't always stop charging in Float. Can anybody provide a link to that discussion, btw?

You want to be able to adjust charging profile. Meanwell looks better in this respect. Perhaps, coupled with $3 kitchen timer.

Not really 12V charger, but these are features that I would like to see. Dial page 9 or scroll to page 17 in the booklet.
Cycle Satiator 12-36-48V.

Top voltage - adjustable.
Bulk amps - adjustable.
Slow start at low "trickle amps" until it's safe to switch to full amps - adjustable.

Oddly, I don't see any timer and they are talking about low currents that might flow when the charging is complete.

Graph shows charging nominal 36V pack.

[2oldman]

brulaz wrote:
2Oldman had a couple of Schumacher chargers modified for his 48V LiFePO4 battery bank. Is that 16 cells in series or ?
Curious what V they get up to: 3.65x16=58.4V ?

58.4 is the bulk charge end voltage.

You tell me! I think there's a couple stacks of these cylinders.

[brulaz]

And to clarify some other issues about best charging voltages,
here's some more info. Seems a lot depends upon the battery cell manufacturer:

https://www.powerstream.com/lithium-phosphate-charge-voltage.htm

https://www.powerstream.com/LLLF.htm

[brulaz]

grizzzman wrote:

...

This info if used will put an end to lfp much sooner. Never subject the cell to hit 4.2 volts.

Yes, I'm not recommending charging to 4.2V per cell. That's just what that graph used for the constant voltage phase for some reason.

A 3.65V per cell (or 14.6V for 4 cells) constant voltage phase as recommended by BattleBorn will get you you to ~95% SOC soon enough and allow cell-to-cell balancing. And just about any 1 or 2 stage charger with an AGM setting, or even an adjustable V Meanwell cc/cv power supply should be able to do that.

2Oldman had a couple of Schumacher chargers modified for his 48V LiFePO4 battery bank. Is that 16 cells in series or ?
Curious what V they get up to: 3.65x16=58.4V ?