LiFePO4 - The future for off-grid battery banks?

[smily03]

Well, it was worth a hope anyhow Maybe by the time I'm ready to actually buy some batteries, one will be available If nothing else, I'll re-re-read the awesome info that you've all posted out here

Out of curiosity, would there be any way to charge cells in parallel, but discharge them in series? That way you wouldn't have to worry about them getting out of balance, right? Or would the added complexity not be worth it? (Well, that and the amperages would be so much higher too; a miniscule 200w solar array would work down to roughly 60 amps at 3v5...)

I know that most (all?) off-the-shelf charge controllers wouldn't be able to output voltages that low, but just being a curious noob

[PNjunction]

If you know what you are doing, you can "fudge" the lfp from the canned presets, BUT the inability to disable the temp-comp is a deal breaker.

In the same manner, if you use the MSview software, you can customize the algo's, but I haven't seen the software, and still this hinges on being able to DISable temp comp for LFP. Maybe that option is in the software? I dunno'.

From a Pb perspective, I like what I see, including the mild "boost" for agm (like an EQ, but much milder), which one WILL need, since to do agm's right in a cyclic mode, that means 8 HOURS of 13.6v float daily - no matter how fast you finish absorb! - something one will never have a chance to do in a daily cyclic operation, and the higher-voltage boost is a way to deal with that.

[Sunking]

[QUOTE=smily03;n297273]Unfortunately it is just marketing hype. Once you dig into the details, it is not compatible wiht LFP batteries. Any charger for Kithium has to bee able to communicate with a BMS and regulate charge current. This uni tis not capable of doing that.

[PNjunction]

Note to lukers - NOT ALL Morningstar pwm controllers will still work if you rip out the temp-comp sensor for lfp use. My Prostar 15 does. Heck, even my el cheapo EPsolar controller on my folding panel emulates it by allowing the op to set a temp comp of 0.00v.

I like Morningstar gear, but this apparent oversight is a bummer. Hopefully that will get fixed.

[PNjunction]

Just saw a webinar on the new Morningstar ProStar MPPT solar charge controller; they said that it supports lithium battery packs out of the box, as long as it has a compatible BMS.

Do you mean this one:

http://www.morningstarcorp.com/products/prostar-mppt/

I REALLY like it but just from perusing the operators manual. I see they seem to have paid attention to the max-smoke model of charging, or at least a smarter/tamer way of doing it if one wants to stick to the canned presets. This is great for Pb, but I have a problem with LFP!

Unfortunately, I see NO WAY of disabling temperature-compensation for use with LiFePo4, or any lithium batteries for that matter. You don't use temp-comp with li-ion (other than monitoring for total thermal failure). Unlike my inexpensive Morningstar pwm's that will still work with the temp comp sensor ripped out for use with LFP, I'm not sure if trying to do so with one of these would outright kill it.

So as of now, NO it is NOT compatible with li-ion directly. Note that other products also claim compatability with LFP, but put the onus of responsibility on a bms that just shunts away ignorant abusive charging - which by itself can be a major point of failure.

I really like this for Pb operations. But for a guy like me who runs with NO top-balancing bleeder POF's, and sets his stuff up right, I need to disable temp comp for LFP. Maybe they could do a firmware update to enable that so I don't have to go cutting up internal traces and tricking it to do so.

[Iron Bran]

[smily03]

Just saw a webinar on the new Morningstar ProStar MPPT solar charge controller; they said that it supports lithium battery packs out of the box, as long as it has a compatible BMS.

And by the way, I've loved following this thread! So much knowledge in here...

[Sunking]

Despite that, the 20ah cells were not as evenly matched for internal resistance like the larger 40ah cells. So, in the end, the cell with the highest internal resistance set the pace for the rest even though when tested, each cell had nearly identical capacity.

The 40ah cells were much more evenly matched in internal resistance.

Not sure I follow you PN. Ri has no effect on capacity when the cells are in series. Just limits the maximum charge/discharge C-Rate. 1 AH is the exact same through any value of series resistance.

That is one of the reasons, not the main reason, when you use lithium cells in parallel is done in a LADDER configuration as it nullifies any Ri imbalance in parallel configs. Of course the main reason is to eliminate a fuster cluck of Vampire boards and wiring.

The most important thing to balance is the AH capacity of the cells. If within 1% makes no difference how you initially balance them as long as you pick either top or bottom. EV manufactures have that luxury and use high quality cells matched from the factory. Chi-Coms cells are 10% tolerance which makes Top Balance self destructive and requires automation which in itself is prone to failures.

Yeah I know same ole tired argument, but some just do not get it.

[PNjunction]

You mention the smaller 20ah cells are a little all over the place. Is this all the while staying out of the knees? I found mine always were problematic after the 3.45/cell, and the farther north of that was a nightmare.

Yes - although you must go at least *once* to 3.6v / .05C absorb on each. Despite that, the 20ah cells were not as evenly matched for internal resistance like the larger 40ah cells. So, in the end, the cell with the highest internal resistance set the pace for the rest even though when tested, each cell had nearly identical capacity.

The 40ah cells were much more evenly matched in internal resistance.

In the end, this wasn't too big of a problem with the 20ah cells since I run conservatively anyway. But, these 20ah cells are just plopped into a plastic rectangular box to look like the normal 20ah agm type of battery - unlike the 40's, which were banded, strapped, and have the 4-screw terminals which represent the Gen2+ product.

That's why anyone doing an evaluation of GBS lifepo4 should start with at least the 40's, unless they are prepared for lower-end consumer non-critical non-matching IR's. I certainly wouldn't base a large bank around the 20's. I'm not sure if the 20's will ever be upgraded to Gen2.

The 20ah cells were a proof-of-performance test for me, even though not absolutely ideal, and that's when I felt comfortable upgrading a little bit.

Part of my testing was to make sure I could use these things with a KISS method, and just a little bit of care. I had no interest in hanging a rat's nest over them, and if I found that they absolutely had to have that kind of nannying, then they weren't for me. One of my major concerns about top-balancing boards and whatnot was that I was suspicious that these would be band-aids hiding poor cell quality - which fortunately didn't come true for me so far with GBS, although the 20's were a tad interesting.

[Living Large]

Living Large,
You are also off-grid?

~CrazyJerry

I am just a lurker who designed an off-grid system with an LFP bank and then decided not to build it. Derek (Sunking) gave me advice on using and balancing LFP, and others here contributed a lot of advice overall. After 6 months of chipping away at the design, Derek gave me a slap up side of the head which made me think twice about going solar to power my cabin. I simultaneously found a way to get POCO to deliver power to my property - which is a work in progress now delayed by winter. But I remain interested in off-grid, particularly the experience and findings of people who are using LFP.

[reed cundiff]

LFP in colder weather

As is generally accepted, charging LFP below 0 C is believed to cause degradation of the cells. Temperature was down to 24 F last night. The battery suite is in front compartment which is well insulated and the inverter etc are in same compartment so probably produce heat. Have a remote thermometer ($18 or so at Lowe's/Target/Home Depot/etc) and it only got down to 40 F in compartment. Have a 30 W inspection lamp which will place in compartment if it gets below 35 F.

LFP has degradation at high temperatures but the compartment is shaded by 5th wheel overhang. There are three panels that open which gives good ventilation. Have a number of 12 V fans (1 amp at 12 V = 12 W) which will use for increased ventilation if compartment gets above 90 F or so. The inverter will not be friend then when running air conditioning from battery suite/solar panels. We will be heading to Yucatan in two weeks but temperatures in Mexico should be temperate

Reed and Elaine

[Sunking]

Well Crazy I will tell you this. When I first decided to use LFP in an EV, I bought the Orion Jr with the mindset of using Top Balance. I bought the Orion and Cells about a year before I actually built the damn thing. In that year I did an about face from Top Balance to Bottom. I then re-purposed the Orion as a Monitor, Alarm, and a control to turn off the charge. After a couple of months building trust with the method, I no longer used the Orion to shut off the charger. My new EV only has a wiring harness with a Amphenol 50 pair connector in the event I need a cell monitor. No plans to really ever use it. BB is so effective, it needs no other backup. My EV has a LVD output built into the motor controller and triggers on pack voltage. Any Inverter has the same thing. I set my PFC 1500 charger to roughly 80 to 90% SOC, and it terminates charge when current tapers down to 4 amps. No cell ever comes close to 100% SOC or 10% SOC. The batteries are always in a PSOC which is exactly what you want for maximum battery life.

[CrazyJerry]

All sounds good. Like me, you have two different battery packs. To hammer the point home about balance (either way!), take one of your Gen1 cells, and replace a Gen3 cell with it and marry it into your larger Gen3 pack.

Can you top balance it? Yes. Can you bottom balance it? Yes. Ridiculous, and of course the triggering for capacity will be on that Gen1 cell. And, for top-balance, we can see that voltage alone does not really mean balance in capacity. Of course the frankenstein battery will be limited by that Gen1 cell in use.

I actually built this frankenstein just for kicks and it drove home the point that voltage means squat - other than just staying in a safe, comfortable specification for the cell, and trying to spend hours lining up the voltages to .0001v of each other when top-balancing can be done, but in the end, balance really depends on the manufacturer's cell capacity and IR, and small variations can be tolerated. Ie, .01v is about the largest spread among cells before I get concerned.

Anyway, that's my story. Being conservative, keeps it simple, and prevents me from spending too much time top-balancing at high voltages for nothing - since my cells are pretty well matched in capacity and IR. (on the 40ah batt that is - the smaller 20ah is a little all over the place.)

PNjunction,
Now at some point I really want to assemble a "franken-battery" - lol! Not because it would have any real value, just for the photo-op!
You mention the smaller 20ah cells are a little all over the place. Is this all the while staying out of the knees? I found mine always were problematic after the 3.45/cell, and the farther north of that was a nightmare. There was a span of time when I had these wired in several different 12 volt configurations (remember I have 32 cells to work with). As an experiment with 4s8p they were stable as long as they were below the dreaded upper knee.

~CrazyJerry

[CrazyJerry]

So are you saying you use them as voltage monitors? That I understand but there are better ways to do that without risking the boards from ever turning on.

I used an Orion Jr in the beginning with my GBS cells. The EV I uilt motor controller was crude and did not have a Coulomb Counter aka Gas Gauge. The Orion had that bell and whistles the controller did not have and can be programmed to do all sorts of task. It can be used to work with a 16S Lithium pack on solar easily. It does have bypass capability that can be programmed to operate at any voltage. It also has the capability to never use it.

Today I am running 31 Leaf cells and my new controller can do what the Orion Jr does except Balance. From the start I intended to Top Balance. Being an engineer I took my peers word for it until I heard of Bottom Balance and many documented cases of Top Balance system failures directly related to Balance Boards. I digress, I learned something. I am BB all the way an djust about 1/2 off the EV guys have now converted. They learned the hard way and destroyed cells with Top Balance.

As I stated earlier you have enough knowledge to use BB. Stay with it, your cells will last longer, and almost impossible to ever over discharge them to destruction.

SunKing,
Someday I'd like to possess an Orion Jr.. The Elite CPU is no Orion Jr for sure, but, it does appear they are leaning in that direction. A few of the more important elements do exist with their CPU. Output taps for HVD, LVD, do exist on the CPU which also reads individual cell voltage and temps. It can take appropriate action depending on what is hooked to those outputs (in the case of the AC supplied charger it can pause or even terminate the charging). Again, it's no Orion Jr but it does perform a few important functions. It would be nice if it had the optional canbus or modbus. The unit will support 16s. Again, not an Orion Jr but not bad if someone wants a packaged system with almost no fiddling around. The MX60's in my system offer additional ways to play although it's getting kind of redundant.

Looking forward to future reports about your leaf cells!

~CrazyJerry

[CrazyJerry]

Hi Jerry,

From what I can see from the manual, the controller is preprogrammed with the battery size and hopefully "battery leakage" data. As WillyT has already stated it resets the 100% voltage point whenever the charging voltage goes above 3.52*4=14.08 volts. If you are only charging to 13.6 volts the Capacity Meter will never be reset so its accuracy will steadily get worse. If Elite Power are not taking into account "battery leakage", accuracy will decrease fairly rapidly. If their software is really smart they could calculate the "battery leakage" from the change in accuracy. Might be worth contacting them to find out how often they think you should take the charge voltage above 14.08. As WillyT has done you could get your Solar Controller to do an Equalise cycle every now and again at say 14.2 volts to reset the Capacity Meter.

Simon

Hi Simon,
Did get a response from Elite and their recommendation is to charge to 14.08 to reset the counter:
1) Once a week
or
2) If the user notice the capacity becoming inaccurate

----

So, there are five easy possible options with my charging setup to address this:
1) The old MX60 can be programmed to equalize every "X" days at ~14.1v @ minimum 1 hour, or, I can engage that function manually.

2) On a particular day, manually set MX60 controller bulk charge to 14.1v, absorb timer to 0:00 (and/or end amps to a value such as 20amps). When battery hits 14.1v charge is over unless end amps was set to a value and charge controller reduces current to 20 (or whatever one sets it at) then the charge terminates - done! After that session, lower bulk charge setting to the conservative setting it was previously at.

3) Magnum MS2812 allows for user set value - so using it in its charge mode from generator (or shore power, etc) will accomplish the charge. It also has an equalization function that could be set to 14.1v and then manually turned on whenever one would wish to use the inverter/charger to quickly charge to that setpoint, then manually terminate the charge (one-button hold for 4 seconds, event stops).

4) Using the Magnum inverter in charger mode, fed by solar charged old pack number one, will allow using the vendor's supplied AC charger that interfaces with new battery pack's CPU (that's receiving data from its vampire boards). Because the charger interfaces with the CPU the entire charge function for this particular event would be fully automated.

5) Using generator or shore power and vendor supplied AC charger will accomplish the charge to reset the display counter.

Some of the system components are getting old however because they originally came with the ability to be user programmed they are useful with the new LFP batteries. Not perfect - but certainly doable in a variety of ways.

~CrazyJerry