Downloaded data and PT display don't jive.



tonyzackery

Well-Known Member
Dec 23, 2006
3,517
80
0
57
Did a search for similar issues and found nothing. My display unit will show 'X' for max wattage and the downloaded max wattage in Power Agent will be lower than 'X'...Anybody know what's up with this??
 
tonyzackery said:
Did a search for similar issues and found nothing. My display unit will show 'X' for max wattage and the downloaded max wattage in Power Agent will be lower than 'X'...Anybody know what's up with this??
Any chance your PT recording interval is set to something longer than the minimum?

-Dave
 
daveryanwyoming said:
Any chance your PT recording interval is set to something longer than the minimum?

-Dave

Yes. I think I set it to 5secs awhile back.
 
tonyzackery said:
Yes. I think I set it to 5secs awhile back.
Yep, that'd do it. The CPU displayed peak is at the hub sampling rate but the peak in your files would represent the best 5 second average power.

-Dave

[edit] Actually I can't recall whether the PT averages samples at lower recording rates or simply drops samples in which case your recorded peak would actually be just one sample out of each 5 second block which 4 out of 5 times won't actually be the peak.
 
daveryanwyoming said:
Yep, that'd do it. The CPU displayed peak is at the hub sampling rate but the peak in your files would represent the best 5 second average power.

-Dave

Thank you Obi Wan...
 
daveryanwyoming said:
Yep, that'd do it. The CPU displayed peak is at the hub sampling rate but the peak in your files would represent the best 5 second average power.

-Dave

[edit] Actually I can't recall whether the PT averages samples at lower recording rates or simply drops samples in which case your recorded peak would actually be just one sample out of each 5 second block which 4 out of 5 times won't actually be the peak.
The PT discards all data except one second of data every 5 seconds.
5 sec recording interval is definitely not recommended and significantly compromises the data analysis. Even at 2-sec recording, it discards every other data point.
 
And here I was thinking I needed the extra long recording time afforded by the longer sampling rates. Heck, I download the thing after almost every ride as it stands anyway...I've switched the sampling rate back already...
 
tonyzackery said:
And here I was thinking I needed the extra long recording time afforded by the longer sampling rates. Heck, I download the thing after almost every ride as it stands anyway...I've switched the sampling rate back already...
Yeah - it's not a very smart "feature" of the cervo CPU. It would have been better had it simply averaged the previous 5 seconds and stored that but alas, no.

FWIW - at 2 sec "sampling", the data isn't usually too bad (e.g. it wouldn't normally have a large impact on calculation of average and normalised power), but beyond that and I would take it with a grain of salt.
 

Similar threads