Jonathan in Hiram
1 kW
- Joined
- Oct 6, 2009
- Messages
- 446
A recent comment of Luke's in a laptop cell thread got me thinking of the possible failure modes of a lot of 18650 cells paralleled and how individual cell fuses could be fitted in a pack assembled from random salvaged laptop cells of unknown provenance and questionable health in order to help them keep from spontaneously combusting due to an internal short in one of the cells shunting the entire parallel group through a single shorted cell.
I started looking around my shop/junk pile for some way to create a pack structure out of random stuff I had laying around when my eye lit on a bundle of bamboo barbecue skewers, a ball of thin nylon string and a bottle of superglue. After making a quick and dirty jig to align the pieces I made two grids of bamboo skewers tied together with nylon string which is then soaked with superglue for a fairly strong joint. This is how I used to fix my model airplanes when I would crash them and that's what made me think of using this technique.
For fuses I used individual strands stripped from stranded copper wire, in the size I used which is .008" or .20 mm in diameter (32 gauge) the fusing point is about 7 amps. Since I'm only expecting around 4 to 5 amps for each cell 7 amps should be high enough to not blow in normal operation but low enough to provide for some internal short protection for the paralleled groups of cells. The fuses go from the cell positive terminal to the 10 ga solid copper bus bars on the top and bottom of the pack, the negative terminals are attached with I think 24 ga wire, I wanted something heavy enough to be reasonably strong but not so big it was tough to solder on the cells. Since the bus is not connected directly to the cells it can be big and heavy without needlessly heating up the cell when soldering. On the Mark II I'll use a similarly sized stranded wire for the negative cell connections, the solid is a bit more stiff than I would like.
The cells are drawn up in groups of four with the same nylon string I used to bind the bamboo joints and then hot melt glued in place to the bamboo grid, I think the support is adequate for a street bike although for offroad use I'd find some more robust means of withstanding the vibration and shock forces involved. The dowels in the Mark II grid are somewhat larger in diameter than the bamboo skewers but I have decreased the space between the cells where there is no dowel to compensate so the entire package is no bigger and may even be a little smaller.
So far I have a about 100 miles on the 6s8p pack in conjunction with the other half of my 96 cells in my previous more "brick style pack", the "fuses" definitely work when a cell is deliberately shorted and so far I haven't blown any from using the throttle. The cells also stay a lot cooler in the open grid arrangement than when they are all massed together touching even without active ventilation. The entire assembled pack is a good bit more stiff than you might think and doesn't have much give to it at all. I have the pack temporarily in a cardboard box that is then inside a heavy cordura camera equipment bag.
Here are pictures of the Mark I shade tree individually fused 6s8p pack and a picture of the Mark II grid for a 5s10p pack made from 3/16 hardwood dowels as well as the Mark II jig for assembling the grid, a Forstner bit in a drill press was used to make the holes to allow access for string wrapping and superglue soaking. When I get the pack design the way I want it I'll cut a box from 1/4" plywood or acrylic or something, drill holes for the grid to stick into and wrap the entire thing up with some cooling air inlet scoops and exits made from plastic spoons on the bottom.
This technique wiring individual cell fuses up came from this video on making an electric '57 VW Bus.
[youtube]raBWFsPlx7w[/youtube]
I started looking around my shop/junk pile for some way to create a pack structure out of random stuff I had laying around when my eye lit on a bundle of bamboo barbecue skewers, a ball of thin nylon string and a bottle of superglue. After making a quick and dirty jig to align the pieces I made two grids of bamboo skewers tied together with nylon string which is then soaked with superglue for a fairly strong joint. This is how I used to fix my model airplanes when I would crash them and that's what made me think of using this technique.
For fuses I used individual strands stripped from stranded copper wire, in the size I used which is .008" or .20 mm in diameter (32 gauge) the fusing point is about 7 amps. Since I'm only expecting around 4 to 5 amps for each cell 7 amps should be high enough to not blow in normal operation but low enough to provide for some internal short protection for the paralleled groups of cells. The fuses go from the cell positive terminal to the 10 ga solid copper bus bars on the top and bottom of the pack, the negative terminals are attached with I think 24 ga wire, I wanted something heavy enough to be reasonably strong but not so big it was tough to solder on the cells. Since the bus is not connected directly to the cells it can be big and heavy without needlessly heating up the cell when soldering. On the Mark II I'll use a similarly sized stranded wire for the negative cell connections, the solid is a bit more stiff than I would like.
The cells are drawn up in groups of four with the same nylon string I used to bind the bamboo joints and then hot melt glued in place to the bamboo grid, I think the support is adequate for a street bike although for offroad use I'd find some more robust means of withstanding the vibration and shock forces involved. The dowels in the Mark II grid are somewhat larger in diameter than the bamboo skewers but I have decreased the space between the cells where there is no dowel to compensate so the entire package is no bigger and may even be a little smaller.
So far I have a about 100 miles on the 6s8p pack in conjunction with the other half of my 96 cells in my previous more "brick style pack", the "fuses" definitely work when a cell is deliberately shorted and so far I haven't blown any from using the throttle. The cells also stay a lot cooler in the open grid arrangement than when they are all massed together touching even without active ventilation. The entire assembled pack is a good bit more stiff than you might think and doesn't have much give to it at all. I have the pack temporarily in a cardboard box that is then inside a heavy cordura camera equipment bag.
Here are pictures of the Mark I shade tree individually fused 6s8p pack and a picture of the Mark II grid for a 5s10p pack made from 3/16 hardwood dowels as well as the Mark II jig for assembling the grid, a Forstner bit in a drill press was used to make the holes to allow access for string wrapping and superglue soaking. When I get the pack design the way I want it I'll cut a box from 1/4" plywood or acrylic or something, drill holes for the grid to stick into and wrap the entire thing up with some cooling air inlet scoops and exits made from plastic spoons on the bottom.
This technique wiring individual cell fuses up came from this video on making an electric '57 VW Bus.
[youtube]raBWFsPlx7w[/youtube]
