Written as an updated version of our 2018 paper, this paper outlines our process for assembling our batteries and explains the reasoning behind our improvements. We also cover topics such as disassembly, disposal, safety, and reliable connections to robot components.
https://team900.org/blog/2023-ZebraPower-2.0/
Direct PDF Link: Battery Upgrade 2.0.pdf - Google Drive
Loved this paper when 1.0 came out and this is full of good stuff for round 2. Great work.
@JamesCH95 any plans to make the “grasshopper nuts” less “priceless” and more available?
We currently run 4awg into SB50. I think we could be getting too much flash with our current crimp tool & process, but when I went a size up it looked “worse” subjectively - with a simple bandsaw cut, it didn’t look like I was getting cold welding of the strands with the larger of the two dies.
Anyone know a good way to section & polish or etch with high-school-safe tools? We don’t have access to a fume hood.
The design is open source and I’m working with a vendor to get them out there in retail form. But the life of an frc vendor is challenging. More noise on CD about wanting them will help motivation though!
For a good section view that is HS safe: cut and polish. Work down through finer and finer grits of sandpaper or files. Usually 220-320 is fine enough for most purposes.
So you’re doing 2 awg on the battery and the one lead going to the breaker, and 4 awg for the other two to accommodate the PDH?
I’m curious if @R.C in combination with Fabworks (who incidentally is a sponsor and has sheet metal and pem nut services) could help with this. It’s been on my mind to ask them so I’ll send an email when I can (if they don’t see this).
In Step 11: Cover the Terminal Connection
We don’t recommend the use of electrical tape as it leaves residue and makes cleanup harder.
Instead, we use heat shrink over our connectors. We like to use double wall, adhesive lined
(glue-like) heat shrink but we also believe silicone repair tape (self-vulcanizing) is an equally
effective alternative that minimizes residue.
I’ve heard that you shouldn’t use adhesive-lined heat shrink because it leaves behind a horrible mess to clean up, so how did you guys do it so that it wasn’t a mess worse than electrical tape?
Ours definitely isn’t messy. The disassembly process is covered and we explain how to remove it and you can even see it doesn’t leave residue.
Edit: I imagine there are different varieties out there but the stuff we have seems to peel up like hot glue and works about the same.
Yep.
So it just kinda comes off with the heat? Are most like that?
Hah, just edited that post. Most of the ones I’ve played with have been. It’s not special stuff really. Several links in the paper for stuff we have used.
Wonderful work!
I’m envious of your test load 
I had to make my own…
May I offer a couple of suggestions:
One abrupt battery failure mode is a dead or damaged single cell. A battery with this failure WILL test as “ready to use” on a Battery Beak! BUT, in competition or heavy load, the output Voltage will collapse quickly, leading to brownouts. If you identify a battery connected to a mysterious brownout, TAG IT before recharging.
The cheapest and fastest way to identify a bad cell vs some other power failure is to charge it, and then hit it with an automotive battery tester; just cut the alligator clips off of one and install your favorite Anderson. They will give you a 100 Amp load for 10 seconds. A bad cell battery will drop in Voltage/current in just a couple of seconds. A good battery will stay pretty steady over the 10 second test. Easy enough to do at a competition!
Here’s what the test looks like: https://www.youtube.com/watch?v=q2sYL4yIWxg
In between competitions, its worth considering a round of energy testing. Knowing the actual Amp-hours on your batteries will allow you to grade them and to identify failed ones. I have a long thread developing a cheap way to do this test: Battery testing data - #79 by Weldingrod1 I don’t love the low current testers; they take longer than a meeting to get data, and they don’t load them the way we actually USE the batteries. This post lists all the various commercial solutions FRC folks use.
We had to leave something for the trilogy.
In all seriousness, testing is useful but given the yearly turnover and usage patterns, it’s a lot easier to strap these batteries on robots than constantly test them. I’m not saying “don’t test” but rather, just as you said, keep an eye out for anomalous behaviors in practice and rotate through them.
Personally, what we’ve seen, is that these style of batteries are all remarkably similar and the state of charge could likely be inferred in software to a reasonable degree with some assumptions. It would be interesting to see that modeled and applied. This would help all teams and reduce the need for special test equipment.
That was my opinion too, until we had THREE new-for-the-season batteries fail with dead cells in one year. Two of them right after SteamPunk came in second on the planet using OUR batteries… Very weird.
That’s what pushed me to develop the cheap screening test (automobile battery tester) and the high current capacity tester! I agree, the high current tester is not something a rookie team needs to be thinking about.
The document never actually says what to do with the zip-tie clips! They just show up in the “completed battery” image on page 19.
FWIW, 3494 started using them on our batteries last season, inspired by this post. We used the exact part suggested by @Jim_Zondag: Panduit MCMS12-P-C from DigiKey, which appears similar to the alternative “all-metal” clip recommended in the document.
The only difficulty we had was that the clips need to be pushed very hard, with sort of a twisting motion to open them up, to get them on the battery. Once on, they’re never coming off, and that’s a good thing!
Yeah. Apologies if we missed that. We figured the picture on page 19 was doing the heavy lifting.
The all metal clips are great. We found a couple (not overly hard) whacks with a rubber mallet could get them on without too much fuss.
The metal/plastic ones are kind of wimpy… wpuld not recommend
I noticed in this updated paper, y’all use lugs with a #10 hole, whereas in the 2018 paper, y’all used lugs with a 1/4" hole. I imagine the change was made for the greater contact area, but why didn’t y’all do that in 2018? Does that tiny bit of extra contact area matter for the average team running 6 AWG?
It’s something we changed to better fit the hardware we were using. We spent more time on this and the minutiae.
But does it really matter any way? Like, does it change how effective the nord-locks are or anything? Or does this really not matter? I ask because having two different types of lugs is just another part to stock and keep track of, and the less complexity, the better.
It is perfectly valid to value having a reduced inventory over doing every possible conductivity optimization. Determining a “notion of best”, tailored to match your (and your stakeholders’) values, is a fundamental aspect of engineering.
