I’ve been doing a LOT of thinking and testing on battery energy capacity, and I’m wondering if anyone can post some real data on Amp Hours used in matches? I’d like to see how that lines up with my 12 Amp and 50 Amp energy testing data. Ideally I’d like to have a quantitative way to say “this battery is only good for powering the robot cart” vs “ready to rock in a match!”
2 Likes
This is in k joules, usage per PDH port for a match when we encountered a brownout.
0: 1.69
1: 24.01
2: 18.91
3: 1.76
4: 0.19
5: 10.30
6: 6.47
7: 3.38
8: 0.00
9: 0.55
10: 0.21
11: 1.91
12: 0.97
13: 0.20
14: 0.19
15: 9.80
16: 2.35
17: 21.91
18: 1.85
19: 22.20
20: 0.77
21: 0.12
22: 0.79
23: 11.48
Could this be kJ?
yes. It is. That is what the drivers station log viewer shows
1 Like
6328 has some season stats here. They totaled 8187.9 watt hours over the course of the season, with some per-event breakdowns too.
4 Likes
Here’s some data from champs from both Turing and Einstein on amp-hour usage per match. I’ve included a divider to indicate the difference between each division. I also had to remove a few matches from the beginning of Turing as our PDH wasn’t logging any meaningful data. Let me know if you want more thorough data that represents amp-hour usage throughout a match.
3 Likes
That seems shockingly low relative to the rated 18 Amp hour capacity of the batteries! 10 Amps for 3 minutes?
I know we were browning out with a pair of terrible batteries in the off season: they were 1.5 Amp hour capacity at a 12 Amp rate.
Those batteries are rated at 18 amp hours at a 20 hour discharge rate. The capacity drops significantly at higher discharge rates.
I guess if you sum up my kj you end up with 142.01 kj or 142010 joules.
at 3600 j per Wh you end up with 39.45 Wh
Ah = Wh / V so if we go with 12v (which we didn’t have the whole time but works) we end up with 3.29 Ah Does that fit into your expectations @Weldingrod1
1 Like
Yes, that sounds more like what I would expect; a significant fraction of a battery used in the match. Two matches on a battery = brown out likely. Which is exactly what we have seen when the kids forget to change the battery…
@Orion.DeYoe You are absolutely correct! When I test at 12 Amp rate I get easily twice the Amp Hours that I get at a 50 Amp rate! The 50 Amp test really feels like a better fit for FRC load testing, but OMG is it a lot of heat!
1 Like
Looking back, I’m pretty sure I calculated this wrong. The way I calculated it was using the outputted current draw from each of the PDH ports and adding them together as they’re all parallel to each other. Then I multiplied this by the time it was being read for (we log the amount of time our log cycles take in milliseconds) represented in hours. I did this for all my timestamps and then summed them to get the total Amp-Hour usage in a match but I’m not sure if it adds like that. I’m pretty sure the last step is where I went wrong but the dimensional analysis seems to check out if I calculate the Amp-hour usage at each time stamp and add them up together but not sure. I’d love to post the correct data just not sure how I should calculate it correctly 
1 Like
Batteries / Energy is something we take very seriously.
So after worlds we got some data together to analyze what kind of power we were pulling from batteries this year vs our baseline.
Below is a graph of energy use/ match, you can see we’re averaging much more than our 2020 robot at around 100-110Kj / match.
I’m very interesting in continuing to track this data over time to see if we can extract more consistently out of these batteries.
6 Likes
I find the general trend of this chart very interesting, if not very surprising.
It looks like energy usage goes up over the season (I assume driver improvement allowing them to push the robot harder later in the season). It also looks like usage goes up throughout each even and then resets a bit for the next event. Very cool.
3 Likes
That is interesting. For the data I posted above I manually added up the PDP values from the log viewer. I assume you have a more automated way to scrape the logs. Care to share the tool? OR maybe a good definition of the file format for the logs.
You can grab the data in robot code using the PowerDistribution class.
You can also use DSLOG Reader 2 to export log data to csv:
3 Likes
I would concur with this assessment - fits the pattern very very well at both the whole-season and per-event level.
We used the DSLOG Reader 2 tool that @Ryan_Blue linked above to export this information into google sheets for the actual visualization.
2 Likes
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.