Questions and Suggestions for New Build. Advice Greatly Appreciated

Joined
Aug 24, 2021
Messages
63
Hey Guys,

I'm somewhat new to EV and building my first custom battery for my electric scooter, Vsett 10+. I am looking for any suggestions and advice regarding a few things. I would really super appreciate any input and I am more than open to suggestions/criticisms.
I'll start by saying I have read through many of the discussions on ES regarding techniques, materials, power flow, etc... I've done tons of research and educating myself, so I do this right and safe. This is my first build and want to do it right the first time. I tend to be an overkill type of guy, so if any of my methods/materials seem outlandish, it's because I want to be safe in all aspects.
New battery will be a 20s9p 21700 Samsung 48X cells that arrived to me naked (unwrapped) and each cell was perfectly balanced at 3.58v. All cell imprints reflect that they are from the same run lot or batch. These cells have a continuous rating of 10a but tests (by Mooch on ECF) confirm the temps are safe and stable up to around 16-17a in moderate bursts. 48X cells have an estimated 3000 cycle life according to Samsung, so this battery may end up being used in a different frame in the future. So I want to build it with no bottlenecks or limits due to insufficient connections or wires.
The 1400w dual motors are rated at 60v (16s) and have injected each with 6ml of statorade from Grin Tech. I estimate at 72v I should be not be pulling more than 120-130 battery amps (13-14a per cell), even at peak. Dual Nucular 12F controllers can handle this with ease.
The scooter deck (home for battery) is a cast aluminum block and has zero flex and limited vibrations since it will be tucked in very tightly. Usable deck width is 172mm wide and 455mm long. The current wooden jig have the wrapped cells at 167.5mm wide and currently they're at 445mm long. My problem and question comes here... Due to the "W" shape layout of the cells, as I narrow the pack or add fish paper between the series groups, the length increases exponentially and I'm already cutting it close.

Would it be safe in my situation (no deck flex and limited vibrations) to wrap each cell with 1 or 2 layers of kapton tape instead of using fish paper to insulate the series groups? I plan on lightly sanding the cell wrappings so the hot glue sticks well. I have extremely good hot glue, the kind meant for auto dent removal and on my test cells it holds them together very nicely and secure. I'm confident with this method. Cells spacers are not an option in my case. Final pack will be completely wrapped in kapton and fish and heat shrink as a final wrap. Also using .3 thick epoxy resin board on the sidewalls. Is wrapping each cell in 1 or 2 layers of kapton sufficient? As I add the .25mm thick fish paper between each of the 20 groups, my length is now thrown off (.25mm fish paper x 20 pieces equals roughly 5-6mm of additional length), so trying to avoid using the fish paper if possible.

I bought the Malectrics V4 spot welder along with a strong Lipo (battery terminals purchased as a backup as well) and multiple types of material to do test welds before I settle on a particular material. I currently have 50mm wide .1 and .15 copper sheet (as shown in the picture), 50mm wide .2 nickel sheet, .15 nickel plated copper but only 10mm wide (not wide enough for 45mm wide "plates" so would need to use strips), .1 nickel plated steel and .15 nickel tabs with precut slots (see picture). The nickel plated steel and slotted nickel tabs are to use as welding tabs only and would go on top of the copper sheet to aid with the welds/heat, since copper isn't easily weldable on it's own.

Does anyone have a recommendation for which material to use as the parallel/series "plates" keeping the 15a(ish) cell discharge max in mind? Would the .2 nickel suffice in my case? Or should I try the copper? The plates and cross section is quite large, so not sure which way to go. 48 hour salt water test shows the .2 nickel sheets are as pure as it gets. Hoot hoot LoL.
The copper plate pictured was my first test template. I do plan on trimming the plates down to more accurate size to reduce the amount of possible shorts. I will also be adding another barley paper o ring to each positive battery end to protect from shoulder shorts.

I'm still brainstorming about how to make my terminal power wire connections to main negative and main positive. BMS will be sitting on top of the battery (made a custom 24mm deck spacer to raise the deck height to allow for this) and is 150a discharge continuous with 1a active cell balancing. Charging through Nucular controller with Eltek Flatpack2. BMS has common wires for discharge and charge, so no separate C- charge wire. Should work perfectly with the Nucular charge function. Discharge/Charge/Active Balancing modes are toggled within the app. BMS P- and B- are both dual 7awg wire, so planning on spacing those out on the terminal bus ends. Anti spark 150a connectors to be used.

I'm on hold until I can figure out the best way to insulate the series groups without affecting my overall length too much. I won't be gluing the P groups together until I settle on a comfortable solution. I'm not in a rush as it's winter time here and I still have my stock 16s battery as a backup for the time being.

As I progress, I'm certain I will have more questions and looking for suggestions on certain aspects. I figured this would be a good place to start since I'm unsure regarding my questions/concerns as stated above.

Any suggestions, concerns, criticism, comments, etc.. are absolutely welcome.

Very Much Appreciated Guys!

Adam
 

Attachments

  • 20220116_214103.jpg
    20220116_214103.jpg
    2 MB · Views: 397
  • 20220116_213700.jpg
    20220116_213700.jpg
    1.5 MB · Views: 397
  • 20220109_023525.jpg
    20220109_023525.jpg
    2.1 MB · Views: 397
  • 20220110_213454.jpg
    20220110_213454.jpg
    1.8 MB · Views: 397
  • 20220116_214252.jpg
    20220116_214252.jpg
    1.9 MB · Views: 397
Dont do hot glue if you plan on using the pack for 3000 cycles in a vehicle. Simple as that.

You can use nickel just fine og you have 1 series connection Per cell. If not, you can consider copper or stacking the nickel. Per cell this build is not very high amp, but the total amps are somewhat high so take care when mounting the negative and positive leads.
 
Vbruun said:
Dont do hot glue if you plan on using the pack for 3000 cycles in a vehicle. Simple as that.

You can use nickel just fine og you have 1 series connection Per cell. If not, you can consider copper or stacking the nickel. Per cell this build is not very high amp, but the total amps are somewhat high so take care when mounting the negative and positive leads.

Thank you for the suggestions. Unfortunately to keep the 20s9p layout I simply don't have room to use cell holders. A buddy with the same scooter built a 20s battery and was only able to fit 6p with the cell holders. I do plan on sanding the wrappings and using really strong hot glue. On my test cells I was barely able to break 2 cells back apart. In addition to the hot glue, I will be wrapping each p group (as a whole) with kapton tape to further hold them together securely. The hot glue I have is not the typical clear hobby glue. It is the black nanoparticle stuff used in automotive dent removal and it is very strong, especially with the wrappings sanded lightly. Do you have any other suggestions on a type of adhesive to use other than hot glue? As I mentioned, with the design, cell holders simply aren't an option. Any suggestions would be appreciated
 
I see what you want to do. Imo the beat way would be to pot the pack with some compound ala what GRIN has done, but I think potting is a bit controversial. I have not seen any comvincing arguments against a slightly flexible compound so thats what I would go with. I think @liveforphysics suggested urethane as a good kind of compound for potting, so i would look into that.
 
Does anyone have a recommendation for which material to use as the parallel/series "plates" keeping the 15a(ish) cell discharge max in mind?

The pure copper sheet is best for the material touching the cell ends. The nickel-plated steel strips are best for the welding caps on top.
 
A layer of kapton tape around each p group should give some protection against chafing and vibration. Are you going to use stick on gaskets on the top of each cell?

Do you plan to replace a cell or p group should one fail?
 
spinningmagnets said:
Does anyone have a recommendation for which material to use as the parallel/series "plates" keeping the 15a(ish) cell discharge max in mind?

The pure copper sheet is best for the material touching the cell ends. The nickel-plated steel strips are best for the welding caps on top.

Thank you Sir! All of my potential ideas have come from information provided on this forum. Copper sandwich, brass welding and common mistakes discussions were all extremely helpful in many ways. Once Lipo arrives I will be testing combinations of various weld tabs and plate materials/thicknesses. I bought the nickel plated steel strip and nickel slotted tabs for this purpose exclusively. For my tests I will probably be slotting the nickel plated steel or using 2 separated parallel strips.
I think it's beyond interesting that the copper serves many purposes. Not only is it 4x less resistive and has 4x the current carrying capacity but it also acts as a heatsink during the welding process (less heat to cell) and also as a heat sink during operation. As we all know, the more a metals heats up, the more resistive it becomes, further diminishing it's efficiency. Ideally, will be using the copper plates. I hope the Malectrics V4 is up to the task. That's what I'll be experimenting with when the time comes. Thank You!
 
electric_nz said:
A layer of kapton tape around each p group should give some protection against chafing and vibration. Are you going to use stick on gaskets on the top of each cell?

Do you plan to replace a cell or p group should one fail?

I was thinking the same. Actually, I was planning on wrapping each individual cell with 2 layers of kapton (.08mm X2 would be .16mm of kapton on each cell) essentially giving 4 layers (.32mm of kapton, plus the heat shrink) of protection between each of the adjoining cells. I am definitely using an additional o ring (gasket) on the positive end of each cell. When I wrapped the cells in heat shrink, I used the thin PVC o rings/gaskets, just like Samsung would have if they came pre-wrapped. So, PVC o ring in addition to a barley paper heat resistant gasket on each positive end to protect from shoulder shorts.
As far as replacing a cell/group if it becomes faulty, I would like to have the option. I do not plan on gluing the P groups to eachother, only gluing the cells in each p group. So that way I would be able to remove 1 p group with somewhat ease. I do plan on using a layer of kapton tape to hold each series group to the next (in pairs of 2) to help reduce movement and vibrations between the series groups to a certain degree.
 
The discussion about brass sheet was quite serious until the "copper/nickel sandwich" was discovered.

It was speculated that with enough welding current, maybe the brass could be welded to the cells for the series current path. Brass is about twice as good as pure nickel. Copper was considered unweldable unless you used a very expensive laser-welder.

Copper sheet and nickel-plated steel ribbon perform better than pure nickel, and also cost less.
 
Was looking into some different types of filament tape for wrapping the individual cells, to insulate between series/parallel groups and/or each cell individually. I'm mostly finding fiberglass, PTFE and Teflon filament types. Most of them are 3-5mils or appox .1-.15mm per layer of thickness. Wrapping each cell individually, by default, would give 2 layers of protection (one layer on each cell) between every single cell in the pack. So pretty thin stuff with exceptional abrasion resistance. Very interesting.
On a side note...
I did a very rudimentary abrasion test last night with some kapton tape and a pair of scissors. Mad scientist experiment lol. I was quite impressed with how tough one layer of that stuff actually is. The kapton tape is around 3mils or .08mm thickness. After scratching at the 1 layer of kapton tape with the pointy end of scissors and failing to score it or damage it... It leads me down the path of thinking; how difficult it would be for 2 heat shrunk, glued, taped and spot welded cells to abrade enough to actually wear through the layer on both cells.
However, I am a realist and have my doubts, which is why I'm asking for advice and suggestions because I simply do not know for sure
Tonight I'm going to try sandwiching fish paper between the series groups but genuinely think it's going to add too much bulk/length since the cells won't be able to be tucked together so tightly. Then I would still have the parallel grouped cells without added protection besides the heat shrink, which is less important than the series groups, but still important nonetheless
The quest continues for now.
 

Attachments

  • Screenshot_20220120-174629_Amazon Shopping.jpg
    Screenshot_20220120-174629_Amazon Shopping.jpg
    650.4 KB · Views: 342
  • Screenshot_20220120-174038_Amazon Shopping.jpg
    Screenshot_20220120-174038_Amazon Shopping.jpg
    641.9 KB · Views: 342
Back
Top