Collect All The Samples

Author: Adrian Edwards

Posted on:

Experiment 7

Purpose of test: To have data on a piece that includes most if not all the different pieces of data so far (particularly force curve, accurate distance and time for estimating RPM, and force/impulse data)

Procedure

This experiment redid most of the code for logging from an erg so as to capture data in a while loop as often as possible and continually compare it to the last time the datapoint changed in any way so as to only save the changes to the data. This way it was possible to log the time and distance as often as they changed and use this to estimate RPM

Data:

  • CSV generated while testing the script (this is probably useless but eh whatever)
  • warmup.csv warmup pic-drill with roughly 20 strokes per position (arms, body, quarter-ish, half, threee-quarter, full-ish)
  • 20k.csv the 20k piece with a sprint during the last < 20 strokes or so

screenshot of the liveplot results at the end of the 20k

Analysis/Observations:

  • it seemed like - for some reason - the impule data this time around was a lot better and not dipping at the end of each stroke (not sure why it was doing that before)
  • RPM data seemed to be plotted negative on the graph when the liveplot was running. not sure if this is a bug in the liveplot or the data collection

Also, here is Ben’s Graphs with one overlaying RPM on each stroke and the othershowing the average RPM per stroke with 1 standard deviation marked.

Ben's analysis plot

Future Improvements:

Seems like RPM data is still very rough and not available too freuently. maybe it can be collected directly from what i presume are pulses on each rotation coming from the flywheel cable?