My First 18650 Battery Pack - KFF

mauimart

100 W
Joined
Nov 10, 2009
Messages
216
Location
maui
This is my first attempt at a 18650 pack build. The pack will be fitted to a Specialized Big Hit running a first generation MXUS in a 24" wheel. I still have to build an enclosure for the pack.

Specs are as follows:

- 20S8P using Sanyo NCR18650GA cells (from tumich)
- Custom designed and 3D printed battery holder (solvent welded ABS)
- 0.2mm x 8 mm pure nickel strip
- >2600 welds at 8ms pulse using custom Arduino controlled spot welder (based on Riba design)
 

Attachments

  • 18650 battery spacer _ 20x8 ASM-page-001 (Small).jpg
    18650 battery spacer _ 20x8 ASM-page-001 (Small).jpg
    42.9 KB · Views: 5,135
  • IMG_4110 (Small).JPG
    IMG_4110 (Small).JPG
    79.9 KB · Views: 5,135
  • IMG_4140 (Small).JPG
    IMG_4140 (Small).JPG
    109.6 KB · Views: 5,135
  • IMG_4304 (Small).JPG
    IMG_4304 (Small).JPG
    82.2 KB · Views: 5,136
  • IMG_0161 (Small).JPG
    IMG_0161 (Small).JPG
    77.9 KB · Views: 5,136
  • IMG_0141 (Small).JPG
    IMG_0141 (Small).JPG
    73 KB · Views: 5,134
  • IMG_9700 (Small).JPG
    IMG_9700 (Small).JPG
    89.1 KB · Views: 5,134
  • schematic.jpg
    schematic.jpg
    136.8 KB · Views: 5,135
  • 20160209_171707 (Small).jpg
    20160209_171707 (Small).jpg
    65.5 KB · Views: 5,135
I made a small amount of progress on the pack, just prior to a significant setback. I soldered the end connections to a strip of nickel plate and then spot welded the plate to the terminals of the pack. I was quite pleased with the connections and confident that it would hold up to the expected 80A max current during operation.

IMG_0296 (Large).JPG

The next step was to route the wires and terminate with a XT90 connector. In the process of routing the 10AWG cable through some zip ties, near the cells, I inadvertently allowed the negative cable to make contact with the can of a cell near the positive terminal (18th in the series so ~60V) and you can guess the rest. A giant ball of plasma and heat materialized from the short. The pain was instant yet I could not immediately see the carbon buildup on my fingers or the damage to the pack as my pupils were constricted from the flash. The episode was quite spectacular. My pain was quickly overshadowed by discontent when I saw that the wall of the cell had vaporized. What a PITA it is going to be to replace the cell(s).

View attachment 3

IMG_0298 (Large).JPG

Some lessons learned:

The heat shrink insulation on 18650's is quite thin and cannot be relied upon to actually provide insulation. Yes, it is true that in hindsight I should not have used pliers to pull the wire through the zip tie near the can of a cell. Before proceeding with the routing I was well aware of the danger and that if the end of the cable (which was flush cut) were to compromise the battery wrapper there would be trouble. Knowing this I continued the routing anyway, with the precision of a surgeon.

Don't route the power cables near cells.

Submerse the charred parts of your flesh in cold water as soon as possible. I made sure to do this prior to snapping the photos.

IMG_0291 (Large).JPG

IMG_0294 (Large).JPG

Next update should hopefully show the repaired pack. I should probably have a camera running while I do the repair...
 
That's pretty bad. :shock:

Interesting that while it vaporized parts of the cell itself, the ziptie millimeters away appears undamaged (though your skin is not). Two other zipties on either side appear to be broken (or were they cut afterward?


It is very fortunate that the cell damage did not cause an internal short leading to a fire....
 
Ouch thats something most people would overlook at first glance, I had to double take and think what shorted the positive lead?...then realized the cell can!!! DOH

Wow
 
Painful and potentially expensive lesson learned...

Heal up quickly!
 
Lurkin said:
Your cell holders are perfect/sweet.

Did you print them or order them from someone else?

I designed the holders using Onshape, which BTW rocks, (search '18650 battery spacer') and printed in ABS on a Prusa.

amberwolf said:
That's pretty bad. :shock:

Interesting that while it vaporized parts of the cell itself, the ziptie millimeters away appears undamaged (though your skin is not). Two other zipties on either side appear to be broken (or were they cut afterward?


It is very fortunate that the cell damage did not cause an internal short leading to a fire....

All the zip ties survived the initial blast. I later cut them to get the cable away from the cells as soon as possible. I am lucky that short did not cause a fire and/or propogate. Voltage measurments indicate that only that cell is bad. I also moved the pack out doors as soon as I could.

skeetab5780 said:
Ouch thats something most people would overlook at first glance, I had to double take and think what shorted the positive lead?...then realized the cell can!!! DOH

Wow

The part that is most frustrating is that I was clearly aware of the danger prior to routing. I just thought that with my "delicate touch" and the insulating wrapper the risk would be minimal. DOH!
 
KFF happens.
 
Glad you're OK, and thanks for being honest and posting the carnage. That is the most valuable way to show newbies about why the experienced builders do things a certain way.

I'll post a pic later, but I was an idiot (what else is new?) and rode without a helmet recently. Fortunately I was wearing safety goggles (that look like sunglasses), and took a spill. My pride was the part that suffered the most hurt, but I also now have on extra scar with which to impress the ladies at the rest home for the elderly. If I hadn't been wearing safety glasses, I might be typing this through one eye.

I'm impressed by the plastic cell-holders in a honeycomb layout. That is the way I would do it. This is ES, we need to be the ones to lead the way in providing links and resources so a builder can get on a website and easily draw two plastic side-plates for holding 18650's, and then paypal a separate website to 3D print them out and ship.

edit: I added you to the KFF index in the battery section stickey:

https://endless-sphere.com/forums/viewtopic.php?f=1&t=46713&p=1254427#p1254427
 
Thanks for sharing.
This makes me scared of my current build now. I was somewhat relying on the cell insulation to shield my cells from each other as I'm just gluing them directly to each other. I think I'll start adding a layer of un-shrunk heat shrink between each group now.
DSC_0008.jpg


Cheers
 
It was looking really good before the accident.

In the pack building process, it's important to plan the steps to avoid having the potential for accidents as much as possible (easy to say with 20/20 hindsight). Things usually go pretty smoothly until you get to attaching the main wires and balance wires. Attaching the BMS is the really scary part. Strategically located connectors help a lot.
 
So I was having a conversation with a friend Bobc about modifying his loose cells in soilpipe design to up the cells in series and reducing parallel (more volts, less capacity, same length. The new design ends up having cells/cans running parallel which are quite a bit of potential difference apart. This KFF incident has further made me think about the necessity for printed spacers so that we aren't just reliant on the battery cell insulation for insulation. Thanks for posting, hope you heal well. Btw its still a great looking pack.
 
That's a good point. If the cell insulation gets a little hot, it may become soft or melt completely if really hot. A space between groups would be a good idea. Something like Nomex paper or fish paper between groups would also be good (won't melt or puncture easily).

Also consider what might happen if the bike crashes or just tips over. The cells need to be securely mounted to prevent things from shorting from impact.
 
Or vibration, etc.

I had a wiring fire on the SB Cruiser last summer because of that; a wire got pinched in the space between box and frame (repeatedly, from flex), and becuase *another* wire further down the cable at the rear lights had apparently worn thru, this let the lighting battery pour full current thru the cables, and started a fire under the seat and inside the lighting battery compartment. I was able to rip the wires out by hand and yank the battery itself out before it was damaged or ignited by the incident, but I was burned in the process, and if it had been the traction pack I doubt it would have worked out that well. :(
 
Some progress in pictures...
 

Attachments

  • 18650 battery V2-page-001.jpg
    18650 battery V2-page-001.jpg
    167.7 KB · Views: 4,566
  • IMG_0661 (Large).JPG
    IMG_0661 (Large).JPG
    199.2 KB · Views: 4,565
  • IMG_0665 (Large).JPG
    IMG_0665 (Large).JPG
    183.4 KB · Views: 4,566
  • IMG_4659 (Large).JPG
    IMG_4659 (Large).JPG
    264.4 KB · Views: 4,566
  • IMG_0668 (Large).JPG
    IMG_0668 (Large).JPG
    223.4 KB · Views: 4,566
  • a01.jpg
    a01.jpg
    255 KB · Views: 4,566
  • IMG_0682 (Large).JPG
    IMG_0682 (Large).JPG
    233.9 KB · Views: 4,566
Cool battery box, but doesn't that Bomber fork have enough travel for the wheel to hit the battery enclosure?
 
fechter said:
Looking good.

What is the housing material?

It's 6061 aluminum, 5mm top plate, 3mm all other plates. Fasteners are M3.

John in CR said:
Cool battery box, but doesn't that Bomber fork have enough travel for the wheel to hit the battery enclosure?

It does look like the tire will hit under compression yet bouncing up and down on the handlebars with all of my weight I can't get it to touch. I think I'm compressing it close to full travel. If the pack does end up making contact with the tire I can always slide the pack further down the down tube.
 
More progress. I added a bracket for the motor controller and the circuit breaker. Also installed a precharge switch and controller enable switch. Current shunt is also installed via a 3D printed holder.
 

Attachments

  • IMG_02.JPG
    IMG_02.JPG
    256 KB · Views: 4,418
  • IMG_03.JPG
    IMG_03.JPG
    209.2 KB · Views: 4,418
  • IMG_04.JPG
    IMG_04.JPG
    135.1 KB · Views: 2,747
  • IMG_05.JPG
    IMG_05.JPG
    220 KB · Views: 4,418
Back
Top