Low cost LiFeP04 "learner bank"

[PNjunction]

Not willing to commit to a small GBS, Winston, CALB etc nominal 3.2v LFP bank just to test the waters for some hands-on experience with your own gear? We're going to go even smaller and relatively affordable where access to each cell is desired.

The major reason I bring this up is to counter some of the totally outlandish and unsafe amateur wiring projects I've come across where people are soldering directly to cells, using speaker wire and duct-tape. STOP.

Basically what we are talking about is testing 3.2v LFP "energy cells", in a sturdy slide-in holder/sled and not some other benchtop hack using a different voltage and chemistry.

I built a small 12v / 4S test system using two of these well made Keystone holders in 26650 size format. This is not the only place you can get them.

http://www.keyelco.com/category.cfm/...olders/id/1244

For batteries, I chose "energy cells" as opposed to "power cells", since these more or less emulate the specs of my much larger GBS prismatics. While a "power cell" like an A123 is a fantastic cell, (provided you can actually get real ones and not rejects / fakes) the capabilities are far more than I need for solar-storage use.

I chose Tenergy 3200mah LiFeP04 26650 flat-top cells. Their specs and usage are very close to my GBS cells, albeit scaled down about 10 times! There are other similar 3.2v 26650 LFP cells out there. Just get them from a *reputable* dealer. READ the specification sheets before use.

Another good alternative may be these although I have not personally tested them:

http://www.batteryspace.com/lifepo4-...ssed-ndgr.aspx

For 12v, you'll need four of these. My recommendation is to get 5 just in case one is not as healthy as it should be, or you want to match up the best 4 out of 5. These cells are not individually matched - close, but not exact.

Lurkers: THESE ARE NOT 3.7V CELLS of a non-lfp chemistry!! I have no interest in those, only to say if you want to go into more details, see Candlepower forums.

The absolute simplest setup would be to use a 3.2v LiFeP04 charger, such as this little one to charge each cell individually:

http://www.batteryspace.com/Smart-Ch...00-240VAC.aspx

In practical use, don't let any cell drop more than 3.0v under load before recharging. Treat them like you would with a larger bank - that is use relatively low current (say no more than 0.1, 0.2C) as if you were planning for a larger solar housebank. This way you can get a better idea of what you'll be facing with the real thing.

This info is not meant to build out a complete project - merely offering a way to build a small "test bed" that you can learn from without the usual haphazard and unsafe practices you see in joke videos. The whole point is to get hands-on inexpensively, before investing in a larger system made from larger prismatics.

With this, you can explore all the aspects such as bottom balancing, top balancing, bms / no bms, and of course a hobby charger like an iCharger makes short work of this. Still, just a single-cell charger, voltmeter(s), and a modicum of common sense and safety should prevail.

These are batteries, and like all batteries need to be attended to by adults who know their own limits for comfort and safety, and not handled by children.

[dc/ac]

I dreamed I was killing my agms (35ha and 155ah) for my starter systems, no kidding, so I decided to pickup LFPs instead.

Your dedication to furthering LFP knowledge is remarkable and much appreciated.

I picked up a system from a _vender who won't be named_ and, after taking it apart (the outer 'shell') it appears to be much as you describe - including a PCB (or bms?). I've had it running a 12v led at night and have not given it a second thought - even though it is woefully under powered.

I also picked up a 12 pack of 18650 A123s to mess around with and try to make my own pack. Not sure if they are 'real' or not - they were bought on ebay from china - 25 bucks - probably not lol but they have been working ok and they were remarkably consistent when I originally checked their charge - all of them except one were at 3.33 - the other at 3.3. I've been wondering about a harness or rig that I can set up and automatically test/thrash these.

Lots of neat stuff happening in this arena. Thanks for sharing your experiences.

[Living Large]

As a lurker, this is great information if I were interested in gaining experience in using and managing an LFP bank. At the moment, I am not. But it would be economical for anyone who wanted to experiment with LFP charge/discharge and balancing to dive in and learn things that can be scaled up. Thanks for posting.

[PNjunction]

I dreamed I was killing my agms (35ha and 155ah) for my starter systems, no kidding, so I decided to pickup LFPs instead.

AGM's can be managed quite successfully in a solar setup - one has to be a little more aggressive in their approach to it (not outright abuse of course), and acknowledging that they need 8-12 hours of float after a so-called full charge no matter what you do. Since that is not possible in a daily cyclic operation, workarounds or some regular preventative maintenance is put into the schedule. But this is about lfp....

Your dedication to furthering LFP knowledge is remarkable and much appreciated.

I'm only standing on the shoulders of giants who have come before!

I also picked up a 12 pack of 18650 A123s to mess around with and try to make my own pack. Not sure if they are 'real' or not - they were bought on ebay from china - 25 bucks - probably not lol but they have been working ok and they were remarkably consistent when I originally checked their charge - all of them except one were at 3.33 - the other at 3.3. I've been wondering about a harness or rig that I can set up and automatically test/thrash these.

Yes, those are counterfeits. Discharge and recycle them, optionally with a mental ceremony of shame. Do not even use them as that supports the criminal industry. There ARE legitimate businesses out there, so lets support them and not take the food out of their kid's mouths.

The point to remember is that the testing done here is to emulate our actual application - low currents and long drain. Energy cells, vs power-cells - designed mostly for motive-power.

[PNjunction]

But it would be economical for anyone who wanted to experiment with LFP charge/discharge and balancing to dive in and learn things that can be scaled up. Thanks for posting.

Only from an upfront-cost. These little cells are actually more expensive capacity-wise than doing an experiment with a small 40ah 4S GBS, Winston, or CALB.

NOTE: Interestingly enough, you see these large prismatics used for EV conversions - but many on the low end mimic the "energy cell" characteristics rather than the power-cell characteristics! That is, it makes no sense to use a high-end model of a CALB for a solar housebank, when their lower-end of the line is just fine for us. It is also one reason why high-performance EV wheelchairs don't use large prismatics either - for them they need "power cells", and the low end prismatic doesn't really cut it for them and they use cylindricals instead. But for us - large prismatics on the low end are no problem. (re: the WheelchairDriver forum)

Normally I don't recommend using the 20ah enclosed prismatic versions you may also see as they sometimes don't represent the latest in improvements over the years. AND I don't recommend using flat prismatics without housings either for the beginner - not only because it results in unsafe homebrew-hacks for housings, but in the end does not emulate what you would do for a solar housebank of any appreciable size.

If one wanted to use something a little more in-between, then a 4S setup of 10ah Headways could be used along with their holders and cell links. (Don't over-torque the screws please!) Even though in reality they are power-cells, the upfront cost may be a bit more tolerable if one wanted something a bit more useful for small projects. In fact, I would rather use these rather than cobble up some multi-paralleled pack out of smaller cells. Again, this would physically emulate what you would do if you transition to a larger prismatic like Winston, GBS or Calb.

The emphasis here again is to keep your cell-count down for simplicity and safety for the DIY'er just starting out. If you know what you are doing, that's one thing, but we want to avoid the rat's nest of exposed unsafe connections that the amateur typically comes up with.

http://www.batteryspace.com/lifepo43...h100arate.aspx

I'm only using this link as a convenient reference - you can get these elsewhere too - but as seen from above, use a *reputable* dealer, not some fly-by night online.

[dc/ac]

Yes, those are counterfeits. Discharge and recycle them, optionally with a mental ceremony of shame. Do not even use them as that supports the criminal industry. There ARE legitimate businesses out there, so lets support them and not take the food out of their kid's mouths.

But, but ... <shameface> you are right. Will do.

[PNjunction]

The shame lays on the counterfeiters and distributors - not the end user who simply didn't know any different.

The project, if you can call it that, is really just to provide a test bed to ease fear, instill confidence and respect, and even possibly the demonstration that KISS can work without a rat's nest of wires with some common components found in any serious battery bank, like an lvd.

An inline analyzer like the watts-up, powerwerx or similar is extremely handy for those who don't use a hard-core hobby charger to keep an eye on things. Those who wish to use say a flashlight-type of charger with LFP capabilities may want to charge the cells in only ONE slot and one at a time to make sure that the charge is the same before installing into the sleds - some slots may vary the charge in those.

There is always the variable benchtop supply too as long as one is careful.

[PNjunction]

Picked up another source of 26650 LFP's "energy cells" from Amazon - KingSolar 3300mah. Not sure who made them, but they seem well matched in capacity and internal resistance. Nice little boxes for each. Doing well.

I run them in the simplest way possible. Each was sanity-charged using a single-cell charger to full. (well, my bench-supply actually from 900ma bulk to 3.6v down to about 100mah end current)

From there, my Xantrex C12 pwm solar controller w/30W panel was adjusted for a 13.6 (thirteen point six) bulk (hvd pot) and also 13.6v float (HVR pot just underneath). Used Fluke multimeter and test points, rather than go by the visual arrow pointers on the pots. So now, it will never really fully charge or get into the upper knee. I removed the optional temp-compensation.

So yeah, I don't actually "float" them, and pull the batts when not actually charging. But if I do forget, this value is low enough not to do any harm in the long/short term.

I try not to let it fall much below 12.4v under load as a pack, but I have taken them down to 2.7v/cell (10.8v) under load if I really, really have to.

Every so often, I'll charge them up to full individually again and repeat the psoc operations at 13.6v max afterwards again. I haven't had to recharge individually again yet. I'll let you know in a year or so if I do.

Easy peasy. Sub-C 12v application, and basically LFP's that drive solar garden lights, be they Tenergy or Kingsolar 26650's, or big GBS, Winstons, or low-end Calbs would do fine. No need for "power" cells like A123, higher-end Calbs, or Headways in this application.