Wolf's PowIRwall

[Wolf]

It is time!
The journey is about to begin its second phase and I will record the progress here.
If you are an avid reader of this board you already know that I have been collecting sorting and testing 18650 cells for about 8 month now.
I already have a functional Pb based system up and running and it is working very well for what it was designed to do. Supplement my electrical usage from the grid by about half. That has been successful.

This is the next step in the evolution of the design. To take a throwaway product extract the best cells out of them and reuse these cells to store and recover energy produced by a totally green energy source. "The Sun".

I would have started this thread sooner but I had a small setback with my cells as I was initially going to go by pre IR measurement of the cell as the criteria of acceptance into the PowIRwall build, but changed my mind to go with post 30 day voltage and IR checks for the acceptable IR range. That decision set me back about 162 cells from my goal of 2856 cells. That deficiency has been made up and knowing what I know about IR the cells that are waiting for their 30 day check are more than likely going to make the grade. So here I am to start sharing my journey to the 492Ah, 27.5kWh PowIRwall.

In this journey I have discovered and learned to use many new tools among them Excel, Ubuntu, Influxdb, Grafana, Microprocessors, Arduino IDE, Sensors and many more things. Notwithstanding the collective minds of this board and their guidance and influence. I could not have made it this far without you. Thank you!

So without further adieu the prototype of soon to be 14 packs of 204 cells each.
This is just a prototype to experiment with for bus bars, fuses and construction.



So stay tuned and experience the journey with me.

Pictures for the Project page.






Wolf

 

[ben teker]

It is time!



So stay tuned and experience the journey with me.

I.R. ... yeah.Succes! I will follow your steps.
Great timing. Today started a first small project marked 'dasPil #1'

4s24p -96 cells - avg 2000 mAh -> learning attemptwith weakest kept cells
4x 4.1 V -> 0.78 kWh
3x 2.5 mm copper twisted busbar - 2A fuse/positive pole

 

[OffGridInTheCity]

Congratulations! Look forward to following your progress.

 

[Oz18650]

I have enjoyed your contributions so far and I am looking forward to the next part of your project.
Good luck with it.

 

[100kwh-hunter]

Yes...Wolfman, finally! congrats!

Again i will keep a close eye on you.

Best and thanks for your effort.

 

[Generic]

Nice going Wolf! Congrats on making it to Phase 2! I love the cell arrangement, your attention to detail is second-to-none.

 

[Wolf]

Hello All,

After a large setback in time as my dog Houdini got very sick ( the vet said he had 3 paws in the grave and 1 on a banana peel ) and after considerable expense I'm back to the building project. Houdini is doing well but has to have insulin twice a day.

So after careful consideration on what and how to build I have decided to go with 14s80p as the initial pack.
I believe a smaller pack size (I was considering a 204 pack) is far more serviceable than the monstrosity I was planning to build.
Easier to handle and easier to diagnose if something goes bad.

As you all know I have over 6000 cells documented in my spreadsheet so I did some filtering with the results and decided to
narrow the field. I decided to go with ≈+ 85% SOH, between 40mΩ to 55mΩ IR, 2250mAh to 2750mAh capacity, and 4.1V after 60 day test and up.
I have many cells with a great SOH ≥ 80% with IR ≤40mΩ or ≥55mΩ with which I will build separate packs. The idea is to keep the IR of each pack within a reasonable proximity.

As you all know I did an extensive testing scenario with how cells work in parallel and decided to keep the mΩ range and mAh range as close together as possible. These parameters resulted in 1556 cells meeting the criteria. I will be building 14 +1 as to have a spare just in case.

I am using Nemo's Repackr and am measuring each cell as I check it off and the mΩ has changed very little if any on these cells. mostly to the + by 1 or 2 mΩ. The cells that have a V below 4.1 or a substantial (>3mΩ) increase will be set aside.

The plan is to use the common 18650 cell holders in a 5X4 X 4 configuration and the Battery Hookup cell level fuses on both the + and - sides.
I will be using a KWeld which I really like as I seem to have it dialed in just right. (picture is a 4S6P build for my covered utility trailer)







Once I get all the cells checked out and the packs populated pictures of the process will be posted.

Till later

Wolf

 

[Wolf]

Hi all. Progress report.

OK I have sorted all the cells of which 1207 made the grade.
The cells are in Brake rotor boxes that worked well for storage.




I have entered them into Nemos RePackr with these results.




I'm pretty happy with that! All cells are >2250mAh and <2750mAh. Also IR is >40mΩ and <55mΩ. Voltage of course is ≈4.1V.

Layed out the 15 holders




And started to populate them. Rechecking each cell for proper IR and V.




That's it for now.

Wolf

 

[Wolf]

Update

861 to go.




Wolf

 

[Wolf]

Update

So I now have all the holders populated.
I was almost there with this picture.




Then came the question to re-wrap or not.
Since I had devoted a year to these cells I decided what's another couple of days to re-wrap.
I already had a bunch of wraps from 18650ed so I decided to try to go with the closest color of the cells original wrap.
I kinda like the color difference.
Turns out it was a good idea. As I removed the wrap of one cell I took the insulator off and lo and behold underneath was a bunch of what looked like rust. Now maybe it was coffee from a spill but it looked suspicious so I replaced the cell with one of the same value.
Better safe than sorry.

Plus it gave me one last chance to test V and IR before totally committing the cells to the pack.

Lined up ready for the wrap.




Now lined up for the shrink.




And the completed pack. only 14 more to go.




Wolf

 

[Wolf]

A mentioned above here is the picture of the cell that had the rust hidden under the insulator.
I am now re-wrapping all my cells.








Wolf

 

[Wolf]

Update on 2 things.
# 1 I found another rusted cell that looked perfect on the outside had good readings as far as mAh,V and IR.

Sorry the picture is a little fuzzy but was trying to get as good as a closeup as I could.




#2 After watching HBpowerwalls youtube video

on rewrapping cells I gave it a shot.
Just a couple of changes.
I put the magnet on the cell first and slide it to the center then I put the cells on the back of a baking tin







Push the wrap down till it touches the tin then very carefully use the heat gun on low fan, high heat setting and point it straight down onto the top of the cell. slowly getting closer till the wrap closes over the insulator. You may still get an occasional insulator fly off but much less than if you attack them from the side. Once the top of the cell is done you are in the clear.
The rest is easy and you will have perfect results every time. Takes the guesswork of the alignment of the wrap. Thanks for a brilliant method hbpowerwall!!






10 more packs to go

Wolf

 

[ajw22]

Those bastard insulators always time it just right, flying off as the wrapper starts to shrink!
Perhaps some sort of weight, or shrink wrapped magnet can hold it in place?

 

[Wolf]

Update:

So after Spot Welding the 5p Fuse sheets from Battery Hookup (I really hope they get some more of theses sheets soon)I encountered a dilemma.
How was I going to get the power from the batteries to a connector safely and mechanically sound.
I couldn't just solder a big wire to the sheet somewhere and call it good.
Besides taking a mΩ reading of the sheet which is 12.5 inches from end to end is ~ 5mΩ. enough to cause a small voltage drop.
So onto busbar design and planning





I spent a bunch of time looking at all kinds of busbar designs and sketched a bunch of them out and finally decided to go with my design #5.
The sides will be covered with Lexan to protect the cells and the wiring.




To bring it to reality I spun up some AWG #8 solid copper wire with a drill to straighten it did some bending attached a 1/16 X 1/2 copper bus bar with solder and this is what the prototype looks like. It will be soldered to the fuse sheet every 3 or 4 inches. From the busbar I will use a marine tinned copper wire lugs AWG #6 1/4 Ring complimented with marine AWG #6 tinned cable, terminated with Anderson PP75 Connectors.









Now to build a jig so I can mass produce 29 more of them for now.

Wolf

 

[OffGridInTheCity]

Those bastard insulators always time it just right, flying off as the wrapper starts to shrink!
Perhaps some sort of weight, or shrink wrapped magnet can hold it in place?

Yes sir - you've got to blow upwards 'just right' till the edge traps the insulator - burning your fingers sometimes

 

[Wolf]

I have found with the magnets and the baking pan trick if you hold the heat gun straight above the cell at about 6 inches with high heat (setting 4 on theWagner Furno 500)and low fan (default setting) you can eliminate the insulator fly off almost to 100%. Patience is the key and slow movement of the heat gun. Additional benefit fingers and hands need not be used and no burns




Again once the top closes you are in the clear.
This batch was done without a single fly off









Wolf

 

[Wolf]

Update

So how to replicate the prototype bus bars with a jig.

Ah Nice piece of wood cut 4 slots into it with a skill saw, 2 passes of the blade worked perfectly for a AWG 8 wire. Depth adjusted to the bend so the buss bars can all be at the same depth.




Then turn it upside down and place the copper plate under it with a concrete backer and solder. I used a Weller 100 Watt iron that worked perfectly.




OK so the jig is set up now and works great! Now how to repeat this process and duplicate the bus bars bends the same every time.
Spin up the copper wire with a drill to straighten it, cut 14" lengths and bend the end with a pair of lineman's pliers.





End result



A little tweaking on some but that can be done on the second bend

Now how to reliably replicate the 2nd bend. Crescent wrench in a vice and a pair of slip joint pliers.





End result




After soldering the 4 bus bars to the copper plate The hole can be drilled while still in the jig to avoid any bending of the bars.



Nice consistent results.
Now on to soldering the bars to the batteries. Since I was using the cell level fuse sheets by battery hookup that was pretty easy.
Still using the Weller 100W iron it took mere seconds for each solder joint to attach to the sheet.






For the finished product.






Now only 11 more to go unfortunately I ran out of #8 wire but I should have some more today.

Wolf

 

[NotSoAlien]

Update

So how to replicate the prototype bus bars with a jig.

Ah Nice piece of wood cut 4 slots into it with a skill saw, 2 passes of the blade worked perfectly for a AWG 8 wire. Depthadjusted to the bend so the buss bars can all be at the same depth.




Then turn it upside down and place the copper plate under it with a concrete backer and solder. I used a Weller 100 Watt iron that worked perfectly.




OK so the jig is set up now and works great!Nowhow to repeat this process and duplicate the bus bars bends the same every time.
Spin up the copper wire with a drill to straighten it, cut 14" lengths and bend the end with a pair of lineman's pliers.





End result



A little tweaking on some but that can be done on the second bend

Now how to reliably replicate the 2nd bend. Crescent wrench in a vice and a pair of slip joint pliers.





End result




After soldering the 4 bus bars to the copper plate The hole can be drilled while still in the jig to avoid any bending of the bars.



Nice consistent results.
Now on to soldering the bars to the batteries. Since I was using the cell level fuse sheets by battery hookup that was pretty easy.
Still using the Weller 100W iron it took mere seconds for each solder joint to attach to the sheet.






For the finished product.






Now only 11 more to go unfortunately I ran out of #8 wire but I should have some more today.

Wolf

This looks really nice, good job. What happens if a cell goes bad? I don't really see an easy way to replace/remove it.

Thanks for posting all the pictures and letting us follow your progress.

 

[Wolf]

This looks really nice, good job. What happens if a cell goes bad? I don't really see an easy way to replace/remove it.

Thanks for posting all the pictures and letting us follow your progress.

Thanks.


If a cell goes bad.............

Yea I was thinking that too. But as I have checked these cells 3 times I do not foresee a problem in the near future.
All these cells have passed an extensive V and IR check at least 3 times if not more.
That being said I am also building a 15th pack in case I have a pack in trouble.
Now to replace a cell yea well you would have to pull one side of the 5p strip off and find the offending cell and replace it.
You would then have to solder a fuse to the bottom cell and replace the sheet on the top.
In my case I would disassemble the whole pack check all the cells and replace as necessary. That's why I decided to go with an 80p
as my original 204p was going to be impossible to work with in this configuration. My ideal solution is to have a couple or three spares for easy replacement and giving me the time to rebuild the offending pack.
I'm kinda sticking my neck out as I am a huge proponent of IR and the importance of it indicating the health and longevity of a cell.
I look at all the packs that have been put together without any regard of the IR of the cells and they are still running so with me and my IR "phobia"
I expect to be in good shape.
Only time will tell.

Wolf

 

[Wolf]

Production has begun on the packs.
All bus bars have been completed and 9 of these packs have been built.
Every one (720) have been rechecked for IR and V and re-wrapped.







I must say I do like the Kweld a lot once you get it dialed in it gives consistent results.
Only 5 more to go and 1 spare.
Testing the Ah of the packs right now and will post results as soon as all of them have been tested.
Testing parameters are:
Using Icharger x6 initial testing was done with a external resistor to allow 20A discharge. Once 3 packs where initially tested
those were hooked up in series to provide a charging and regenerative discharge holding tank capable of 30A either in or out.
Sneak peek at the results so far? OK





Wolf