HV battery health check

How did you generate those plots? Anything exotic or just Excel?

 

@2009Prius, thanks so much for the interesting visualizations throughout this sub-forum (and for all of the work on PriiDash)! I'm not using PriiDash yet (mainly because I'm hesitant to carry another device around), but I'm intrigued and love seeing how others are using their OBD2 log data!

I thought this was a really interesting way to visualize these data related to the HV battery, so I thought I'd try recreating the visualizations using Tableau and data that I collected using the EngineLink app (iOS). I know that there may be some differences in sampling frequency, precision, etc., between EngineLink and PriiDash.


2013 Prius two
Data collected May 2016 (~3 years old, ~35,000 miles)
Data collected using an OBD2 wireless dongle + EngineLink app for iOS on an iPhone 5

I've logged over 230 trips since I got my new-to-me 2013 Prius in Feb 2016. My normal commute is mostly highway, so I don't see a lot of extremely high/low current draws. The data below are for five trips in May 2016 that had observed draws of +/- 100 A, although there still very few observations in that range. Most of my results are consistent with the earlier analyses in this thread, although I know that my results may be suspect due to not observing enough stress on the battery.




If I correctly understand the charts, they are comparing each individual block's reading at timestamp t to the average of the 14 blocks at timestamp t. Using that method...


Scatterplots:
In general, no blocks appear far from the average. Consistent with the results from previous analyses above, the "outside" blocks (e.g. 1,14) tend to run higher, and the "inside" blocks (e.g. 7) tend to run lower.



Histograms:
I was surprised to see that the mode for block voltage was pretty far from zero for a few blocks (e.g. 1, 14). However, each bar represents only 10 mV, so it's really not that far off!



On the block resistances, I was again surprised to see the lack of symmetry around zero, particularly in blocks 1 and 14. These results are again pretty consistent with those above. The measured precision is to the nearest milliohm, so some of the spikiness is probably due to rounding.



Here's the link to this dashboard on Tableau Public:
| Tableau Public

More Prius visualizations on Tableau Public:
| Tableau Public

 

My pleasure - thanks for the idea. We'll see how it looks in a couple of years...

Also, I recently discovered your treasure trove of visualization links here - awesome!