Hey guys, what setting do you prefer for the power averaging setting on the display. I've been using 2 sec for a long time, but I've been thinking I might like a little slower setting.
PT averaging ?
- Thread starter flapsupcleanup
- Start date
I use 3 sec. I've always kinda thought 5 seconds might be the smoothest, since that would allow pretty close to 4 complete reports from the hub (4 x 1.26 = 5.04). If my understanding is correct, that would mostly eliminate the problem of fractional reporting creating jumpiness on the display.
Thanks, I went out today for 2x20's and changed the setting to 3 sec. It seemed a bit better, but not a huge difference. I'll try 5 next time out.
I have understood that the fluctuation is caused by the differences in the amount of work done during the 1.26 s measurement period. Like, e.g. riding with 60 rpm, some 1.26 s periods would include two push downs, some would include three, so 50 % differences would be reported. Averaging four measurement periods could still have some fluctuations, like between 9-11 push downs during 5 seconds.frenchyge said:
I agree with what you said, and certainly the recorded data (as seen via software) would still have large fluctuations at 60 rpm no matter what averaging interval is used. I'm honestly not sure whether the display updates every X seconds (as the manual mentions), or every X samples (as rmur17 alludes to above), and maybe any potential mismatch between the reporting and display update is already taken care of in the computer averaging. I do know that I see a regular precession in the display of ~10% using the 3 second display averaging. The precession seems to be 3 low reports, and then 1 high report, so I assumed it had something to do with the 1.26 sec reporting from the hub.sidewind said:
I went back and looked at the data on my sets of intervals. They were done outside and it was a bit windy so it was a challenge to keep the power constant. The set where I had 2 second averaging was quite a bit smoother than the set with 3 second averaging. Evidently with quicker feedback I was keeping my output more constant. At the same time I noticed that with the slower setting I was paying more attention to my body and less to the display, only checking it every handful of seconds rather than constantly.
When you view the data on a PC you are seeing the raw data from the hub. Any averaging applied to the PT display does not affect that. Cadence can affect the smoothness of the data as viewed on a PC (by synching the downstrokes with the 1.26 sec recording interval, or a multiple thereof), so if your cadences were different on each rep then that's a possible explanation also.
I know what you mean, (aliasing, isnt it called? I was an engr in a previous life) but in this case I suspect that it's more psychological. I'm pretty sure my actual power output was not as smooth because the averaging hid little highs and lows from me seeing them and compensating.frenchyge said:
Outside, it's nearly impossible for me to keep a good constant power output. Indoors on the trainer, I can draw a straight line.
The SRM however averages around a complete crank cycle and reports that data -- thereby eliminating this 'apparent' variability or varability where little actually exists.flapsupcleanup said:
it's another reason I like to ride 'blind' past the first 2-3 minutes of a TT - why watch a bouncing display when one's breathing rate and leg sensations tell the story? Maybe I'm just old
rmur
