Sunday, November 22, 2009

PT Calibration Curve

It's raining, I'm in a rest week, and I needed a break from house chores and work; so instead of a ride today I did a full range of Power Tap static calibration tests, as suggested by Dr. Chung.

I found that my PT is pretty accurate. It appears to read an average of about 2.3% low overall, which is within the advertised range, I think.

The errors are slightly larger than that (up to 4-5%) when I'm in my 25-tooth cog or in my 12-tooth cog (climbing or sprinting, just when I'd like to see big numbers the most!). The errors are slightly lower than that (0-2%) when I'm in my middle cogs (16-19 tooth range). It makes sense that the PT strain gauges might have more trouble measuring torque when it's applied to the very ends of the measuring cylinder (hub body).

Errors that small won't make bit of difference to riders when using their PTs for training. But I can use the information when Chung testing to get more accurate results. I'll just keep the chain on the 17 cog and scale my power numbers back by about 1.5% before doing the calcs.

And I'll run a few tests every 3 months or so to give me confidence that my PM calibration isn't drifting.

4 Comments:

Trey Gavin said...

How long did this take? May be worth a trip to casa de DataHo to see what I got across the range...

Anonymous said...
This comment has been removed by a blog administrator.
Chris said...

Where can I find details on how do to this type of calibration test?

Robert Jordan said...

I included a brief test protocol that I found somewhere else in one of my recent posts. You can find it at http://jasperga.blogspot.com/2009/11/power-tap-static-torque-test.html