CRUD's Powerwall Project

crud

Member
Joined
Apr 5, 2017
Messages
77
So i'll start off with my projects documentation here.

Short background:

I am planning an off-grid powerwall for the garden house.

Goal is to have enough power to run LED lights, Laptop and maybe the espresso machine (if i get one with watts below 1000 ;-) ).

I'd prefere to go for a 7s 64p system at the time being.

With at least 115 AmpHours i hope to get a 2,5 to 3 kWh pack.

Right now i finished the first two packs and extracted half of the third pack.


image_rdiglg.jpg


Needs resorting and maybe some retesting, but i am happy with the little progress so far.

After each round of getting the cells from the Laptop batteries, i do a charge test, internal resistance and voltage check using these two battery chargers:


image_gzkrwh.jpg


I already tested some fusewires by setting up a little test environment to get the lengthand stuff right.


image_rtccrz.jpg


I use a 50W 1R Wirewound Resistor to validate the correct amperage flow and prevent shortening out warning of the meter or power supply.

So, on we go, i hope i'll get this finished by the end of june to get the chance tocatch some rays with the solar panels :)


Initally, i wanted to build a solderless pack to maintain broken cells easily.

So i came up with this design:


image_anypwj.jpg


But by now, i realized that it might be no good compromise between maintainability and cost ... maybe.


image_muxwdb.jpg


Have to redesign for 4 rows (as shown in the first real picture of the finished pack), because initially i thought of a smaller
setup using only 32p.

Decided to go for asoldered version after all, like everyone else did successfully.

I'd like to keep the busbar design, though.

I want to change the design so the the top of the cell is exposed through an opening in the top plate so the fuse can go from thetop of cell
to the busbar.


Any thoughts and critique highly appreciated :)
 
Neat Design. I like it. Cannot wait to see more
 
So finished my third pack and new cells for the last 4 packs are on their way :)

Today i had some time to think about the basic procedures concerning going solderless.

I found an very interesting thread over at endless-sphere : https://endless-sphere.com/forums/viewtopic.php?f=14&t=85912

I really fancy nuxland's idea concerning the coppersheet layout.

As i still struggle with the idea of a solderless design, i thought i could probably go for a hybrid solution and use pressed copper on the bottom (removable)
and do the best practice copper-busbars with fuses on the top.

On the bottom i am thinking of removing the little edges of the batteryholders permanently to make the cells slip through at the bottom (As shown in
).

I'll check and do a redesign of that, might be worth a try.
 
Found a new source for cells this week:

Used ebike batteries. Lots of defect packs around for sale.

I'll receive 4 of them today (hopefully). The packs contain 60 18650 Cells each. All of them have broken electronics, so
i am looking forward to a good harvest :)
 
They are a good source when they are not too old, older sometime have crappy lipo who inflated. 18650 based are usually like your's, just faulty bms and good cells. :)
 
Disassembled my first used ebike battery today. Got myself a pair of electrician safety gloves first.

Caution with such a pack.

So much fun. So less glue (absolutely easy to remove)! Fingers don't hurt. Got 60 cells from this pack :)

Sure, have to test them first. No guarantee they are still ok. But it took only 1 hour for 60 cells :)
 
So tested cells from the ebike pack. About 25% is dead on the first pack :-(. Expected that, somehow.
 
What do you call dead ? 0v ? <1000mAh ? shorted/leaking ?
 
coolfx said:
What do you call dead ? 0v ? <1000mAh ? shorted/leaking ?

Cells look good, no leakage, but theyare not recognized by the charger. So i supposethey are shorted or 0v.
But now that you say it, i should checkwith the multimeter. Just to make sure.
 
If they are 0v or close, most chargers don't recognize them. just put in parallel in the charge, for few seconds another cell full charged (for that I let some nickel strip on some cells so it's easy to slide between charger terminal and the battery to "jump start").

After you jumpstarted a cell, the charger should start it's process, remove the other cell when voltage is above 1-1.5v, if the cell drop to 0v and charger stop working throw it, it's probably vented.

If cell become hot quicly, throw it, it's probably shorted.

The cell you use as starter can be hot after starting few cells, it's better if you have high discharge cells to do that, or you just let it cool down and use another one. I still use the sames ones for month, and they are junk <1000mAh.

A lot of theses 0v cells went low because of bad BMS or another cell who drained the entire pack and are still good, but keep and eye when slow charging them and throw the ones who became hot. After the charge/test/charge cycle or only charge, let them a few days and check voltage to seek for self-discharging cells.

I salvaged a lot of 0v cells who are in better conditions than some initialy at 4v.
 
Note that if they dont take charge at all the CID is most likely popped and recomendation is to throw them. (The CID popped due to a reason)
 
Very interesting! Thanks a lot for the details.

Indeed, the BMS did some damage (at least what the seller told me).

I'll try to jumpstart in a save environment and doublecheck the cells.

I actually had some cells below one volt. Some did refresh, some stayed at 6-20 mAh after charging.

I currently use some of the cells with the refresh program of my chargers. I'll check on that when i get back from work this evening.

Best
 
So still testing cells :) Got a lot of good (andbad) cells out of the pile:


image_urlaym.jpg



... another 200 to 300Cells resting inside laptop and e-bike batteries waiting to be tested ...
 
Today i received a piece of 0,5 mm copper sheet for developing my cell connectors. Just a cheap piece of copper. For the final busbar i'll use CuETP which is the
kind of copper used in electronics and has the lowest resistance. A little bit more expensive, though.

I recently got back to my idea about creating a solderless batterypack. As mentioned earlier, i am going to combine both solderless (for the bottom/negative connection) and soldered, using the common way of having a busbar connecting to the positive terminal of each cell using fuse wire.

So today i just made one connector - very roughly and by hand - just to get a feeling for the material and thinking about a good process to make 80x7 of them without spending too much time.

So the idea ist to make a connector that will have some kind of material acting as a spring beneath it (the copper is very soft and won't have to much mechanical resistance). Over a endless-sphere, people use e.g. rubber, and more often silicone sheets for compressing the packs and leave some room for expansion.

As seen in the picture, i applied some material (just got the end of a balloon - nothing else around ...). Just for demonstration.

Its all bumpy - just did that quick - but i try to get the idea ...

I'll need some 10mm tool to deepen/cut the material. Still, 0,5mm copper is not that easy to cut with a hand tool.

So i guess i'll have to find a way to eigther cnc mill or lasercut the material.

I also thought of applying e.g. 6 bumps like so: http://bit.ly/2reoBat

... to allow for more robust contact.


image_vllzeb.jpg

image_faomuk.jpg

image_nhabxy.jpg

image_nxzydo.jpg

image_movqux.jpg
 
After some time off, got myself another charger (OPUS 3100 - Actually its exactly the same as the IPC-3 Voltcraft Chargers).

Opened one of the new ebike battery pack arrivals. Good catch for today :) 60 Cells still at 2300mAh ! each :)

I also found a good source around my place for waterjet cutting.

I'll just have to finish the construction for the bottom plate. Then, hopefully, i can let it cut and start the final assembly of the packs.

Missed the chance to get the Watchmon, Shunt and other stuff from Batrium. They are currently out of stock - waiting for a reply when the items will be available again.
 
Finally ordered WatchMon, LongMons and other stuff from Batrium (Thanks !). Try to get on with the construction of the packs in the meantime.
 
Welcome to the Batrium world :)
 
I finally get to the busbar section :)

Just got some 25 meter 3x2,5 cable to try things out.

Extracting the copper was rather easy and i immediately tried out the drilling technique to achieve twisted busbars.


image_sywump.jpg

image_npecyn.jpg


I got away from the copper sheet idea cause i realized that it would cost me a lot more to get this done.
So i am heading in another direction.
 
So! Finally,Batrium Watchmon and other stuff arrived!


image_yhdfcr.jpg


Had to pick it up from customs yesterday. Delivery was very fast, thanks Jaron!

So, currently i have almost everything ready. Beside the packs ...
 
  • Like
Reactions: wim
Back
Top