diy solar

diy solar

Bought some BIG (280AH) cells on Alibaba

Maast

Compulsive Tinkerer
Joined
Oct 31, 2019
Messages
773
Location
Washington State
I've been buying BIG lifepo4 cells from 5 different vendors on Alibaba in the 260-280AH size, each one weighs 12 pounds. I've taken a nasty hit on shipping by buying just one cell from each vendor but since I intend on buying another 48 of them I thought it worth the investment to get and exhaustively test them before going on with a large order. Each cell has averaged about $125 and shipping has been around $85.

Initially I intend on replacing my aging existing 350AH 48V AGM whole house backup bank and then I might get into the battery business. I've got some thoughts on producing 12V modules with integrated thermal heating/cooling via onboard peltier modules. Or maybe just resell these large cells.

At any rate - to test the cells I've rigged up a 500 amp load bank called a liquid rheostat, it's a ring of sheet metal inside another ring of metal separated by about 3/4 of an inch inside a 30 gallon tank of conductive sodium carbonate (washing soda) solution. The inner ring can be raised or lowered to vary the resistance and control how many amps I want to draw.

I'm still working on automating the test rig, I can manually run it through the 300A shunt w/ AH counter and that works pretty well but I'm assembling a automatic low voltage disconnect from a adjustable lipo low voltage alarm, a relay board and a 400amp normally closed contactor from electriccarparts. That way I can at least start a test cycle and step away from it for a few mins an not have to worry about over discharging.

I want to automate charge/discharge cycle testing w/ data logging using an arduino but thats proving to be a heavy slog.

Thought somebody else might find it interesting.
 

Attachments

  • cells.jpg
    cells.jpg
    58.3 KB · Views: 95
  • liquidrheostat.JPG
    liquidrheostat.JPG
    52.7 KB · Views: 93
Last edited:
It can be either, it's just a huge variable liquid resistor. It turns out making a load bank that'll pull 200+ amps out of a nominal 3.2V source and still be adjustable is a little difficult and this was only thing I could build that wouldnt be expensive and still work so well.

I'll do another test run in a day or so and post it to youtube, it's kind of neat.
 
Keep us informed to know which cells check out the best.
Will do, I'm chugging through a LOT of data collection though and it looks like the answer isnt as simple as it looks. For example I have one cell that produces the most AHs when drawn at .1C but loses to another when drawn down at .5C or 1C
 
I'm very interested - I have just posted about the capacity of some new cells.
Your cells (12 pounds in weight) is 5.4 KG. A 260 amp hour winston brand cell is 8.9 Kg. Winstons can be discharged at 3C.
This is why I am asking when this quantum leap in weight performance happened? Its a huge and sudden jump. Yes please keep us informed.
If we did come up with a standard way of testing capacity.2C .5C 1C etc that we used across this forum then we could churn out some good stats on various batteries.
 
Last edited:
Have you bought a deligreen also? I am looking into buying at them, so any info would be welcome
 
Have you bought a deligreen also? I am looking into buying at them, so any info would be welcome

I bought some Calb branded batteries from Deligreen. I haven't got the chance to test them yet but everything seems legit.
 
Super interesting thread @Maast !! I look forward to seeing your results. Any idea when you will be in a position to share the data and experience?
 
Sorry for the delay, now that it's winter and cold and I had the shop door closed I got concerned at the sheer amount of outgassing from the liquid rheostat so I tore it apart and am replacing the load with loops of .032 stainless steel safety wire in the 30 gallon drum. FYI at 3.4 volts one 16 inch strand is a 5 amp load.
I'm also rebuilding the tester to automate life cycle testing. At first I was going to use a arduino but that was a heavy slog so now I'm using time delay relays and I think I like those a lot better. I've got my worst battery under test now w/ a 60 amp load (.2C ish) so I can take a picture of the readout. Should be done here in a few hours and I'll post it.
 
Here is my first result, thats my worst battery and even that one put out 279AH going from 3.65V all the way down to 2.5V at .2C (ish)
I dont like that cell mainly because its voltage droops .075V under even a mild 20 amp load. Also a pic of my semi torn-apart tester my cc power supply and my mickey-mouse load bank.

.P1030908.JPGP1030909.JPGP1030910.JPGP1030911.JPGP1030912.JPG
 
Nick work. Engineering tests are never pretty! I'm an electrical engineer and I can tell you most of the time they're borderline safe too! Ha.

What liquid is in the barrel now? Water?
 
Yeah just water, with a little bit of copper sulfate to keep it from growing nasties but not enough to make it conductive.
 
Of the 5 batteries I got, one is a ETC, three turned out to be the same GFB battery from three different vendors, one is a ETC battery, and one is a real turkey thats supposed to be a 280AH "Lithtech". What happened is I put out a RFQ for 270-300AH batteries without specifying brand names and this is what I got.

Of the three types, initial capacity testing at .25C-ish gave me two 286AH and one 288AH for the GFB batteries, 279AH for the ETC and 273 for the turkey. I have to say I like the GFB batts quite a lot and those are probably what I'll be going with. I've now got 5 cycles on one of the 286AH GFBs and each one comes back at exactly 286AH, those cycles are from 3.65 all the way down to 2.8, call it a 95% DOD.

I'm limited on my charging so I'm only getting two cycles per day, at this rate its going to take weeks to accumulate enough cycles to form a conclusion, then I have to do it on the ETC batt.

I've got a few RFQs out for a 120A meanwell hrp-600-3.3power supply, that should speed things up.
 
Nice project!

"Of the 5 batteries I got, one is a ETC, three turned out to be the same GFB battery from three different vendors, one is a ETC battery, and one is a real turkey thats supposed to be a 280AH "Lithtech"".

Wondering how well do Batteries from different vendors compare?
What variances do similar rated but differently - shaped/ batched/ quality pose?
Another meter of validating life cycle efficiency of these could be comparing their compliance with Charging/ Discharging rate graphs published by the OEM(s)- however this would need some script that could pick voltage readings throughout the charge / discharge cycle with variable C(s) and temp. for plotting.

All the best and please do keep posted.
This would be a good study for people around the forum!!
TC
 
Last edited:
They're all about the same size and shape, I only weighed one of them and it came in at almost exactly 12 pounds. They all feel to be about the same weight.

I dont know who makes the GFB277 batteries but it's what on the QAR code block on top of the batteries. It's probably something like Guanzou Fabriqe Battery. Or maybe its just a model number, either way I couldnt google it either.

I had to drill and tap each one myself for mounting a threaded stud that I made from a 6mm bolt - except for the turkey as it already had a threaded stud in it. I did find that the aluminum pad is quite thick and I could drill a 1/2" deep hole, I also figured out that if I drilled just the right size hole I could screw in the bolt without tapping it and it'd be plenty secure.

A couple of the vendors have assured me they can add them with no problem. One of them even sent me a video of studs being lazer welded to the batts.

Edit:
Just went through my message traffic and Seplos (Shenzhen Seplos Technology Co ) said that he'd sold two other batteries to other companies and he thought it was probably for me. I dont know if its true or they just all talk to each other, but the videos he sent of the lazer welding and the battery pack assembling all had "Seplos" on the battery. However his quote was the highest per battery at $140 per. Shipping was $85.

Edit to the edit:
Cycle 7 still at 286AH
 
Last edited:
Great!
Did it occur to include Sinopoly cells in the tests?
Found Winston and CALB, reputable suppliers of these as mentioned on one of the sites.
Then there are samsung NCM cells (Nickel-cobalt-manganese) at a slight premium.
Since these are standard suppliers, may be its worth considering especially if eventually a lot of money is on the stake.

I'm myself looking into a similar project, but a portable one.
 
Back
Top