Hi guys, I'm sure there's a similar thread somewhere in the forum but I can't find it. So I'm gonna start another one. =P
Just got a Polar S720i with the Power Output function. Firstly, a problem I noticed right off the bat is the receiver keeps disconnecting from the 2 prongs when I'm riding rougher roads.
Secondly, how do I test my FTP? Just push as hard as possible for 1 hour? What if it flunctuates throughout? Use the average given at the end? What are the basic things I need to test before doing power training? Thanks.
I'm not a Polar user, so I don't know about the hardware connections. As to your "getting started with power training" questions, I think the smartest investment you could make would be to buy Hunter Allen's and Andy Coggan's new book, "Training and Racing with a Power Meter." It will answer the questions you have and about a hundred more you don't know that you should be asking. Frankly, I think the power meter manufacturers should just include the book with their product. It is a much better "user guide" than any of theirs.
Hmm, I've never had any problems with disconnects. But I'd like to advice you regarding another thing. The average wattage in the Polar software does not include zeroes (when your not pedaling). For me that means that if I ride slowly out of town, press the lap button, do some tempo work (L3) and press the lap button again and ride home slowly through town I usually get about the same wattage for all three intervals. That can be pretty depressing (and also make the wattage reading pretty meaningless unless you constantly pedal). With Cyclingpeaks it looks much nicer. Of course there are other reasons for getting Cyclingpeaks as well.
On the other hand you can a pretty high FTP reading with the Polar. Just do a couple of sprints and then don't pedal for the rest of the workout:-).
On the other hand you can a pretty high FTP reading with the Polar. Just do a couple of sprints and then don't pedal for the rest of the workout
yeah, I downloaded the Cycling Peaks demo to calculate NP after reading "Training with Power", shortly after questioning the "avg. power" I got with my polar. Thing is I still get an NP too high... I have some training crits where my NP was over 400 watts for a 50 min crit, and at most my FTP is 340-360watts. I believe this is caused because I am taking really hard 1-2 minute pulls and or break attempts, then sitting in, and repeating. The zero readings should have been taken out with the NP calculation in cycling Peaks, so I don't know what is really going on. Polar is probably not good for comparing to others, but it will work to track your progress.
yeah, I downloaded the Cycling Peaks demo to calculate NP after reading "Training with Power", shortly after questioning the "avg. power" I got with my polar. Thing is I still get an NP too high... I have some training crits where my NP was over 400 watts for a 50 min crit, and at most my FTP is 340-360watts. I believe this is caused because I am taking really hard 1-2 minute pulls and or break attempts, then sitting in, and repeating. The zero readings should have been taken out with the NP calculation in cycling Peaks, so I don't know what is really going on. Polar is probably not good for comparing to others, but it will work to track your progress.
Hmm. The normalized power algorithm misses the mark occassionally, but I've never seen an error as large you're reporting. The way you rode can't explain it, as that is precisely the sort of scenario for which the normalized power algorithm was designed, and normally works quite well. I therefore suspect that you're either underestimating your functional threshold power, or it has something to do with your use of a Polar powermeter (for example, since the Polar only records every 5th 1 s sample, it's possible that this significant undersampling of the real data is somehow skewing the 30 s smoothing).
yeah, I downloaded the Cycling Peaks demo to calculate NP after reading "Training with Power", shortly after questioning the "avg. power" I got with my polar. Thing is I still get an NP too high... I have some training crits where my NP was over 400 watts for a 50 min crit, and at most my FTP is 340-360watts. I believe this is caused because I am taking really hard 1-2 minute pulls and or break attempts, then sitting in, and repeating. The zero readings should have been taken out with the NP calculation in cycling Peaks, so I don't know what is really going on. Polar is probably not good for comparing to others, but it will work to track your progress.
Can you output a PT CSV file with your version of CP?
Hmm. The normalized power algorithm misses the mark occassionally, but I've never seen an error as large you're reporting. The way you rode can't explain it, as that is precisely the sort of scenario for which the normalized power algorithm was designed, and normally works quite well. I therefore suspect that you're either underestimating your functional threshold power, or it has something to do with your use of a Polar powermeter (for example, since the Polar only records every 5th 1 s sample, it's possible that this significant undersampling of the real data is somehow skewing the 30 s smoothing).
Yeah maybe my power is that high but these files are from a year ago and I doubt it as I feel stronger now, yet have not had any stellar results this year... recently did a mediocre 22:30 10 mile TT on a road bike without aero aids, so no power house (but I was tired that day)... I did actually get my power unit back today and it is getting installed tonight so that I can do some more formal estimating of my FTP to begin to actually use the wisdom in the book you co-authored. 😄 You are probably on to something with the sample frequency as I had it set to 15 sec intervals.
RapDaddyo said:
Can you output a PT CSV file with your version of CP?
Here are a couple of sample files... they are over a year old as I have not had a power meter since then. One has a lap just over an hour with NP of 402 watts, the other with a lap of 28 minutes and NP of 421 watts. Polar reports average wattage of 431 and 419 respectively. Once I get the unit back on and working again, I should be able to get some better data, and interpret it with a little more knowledge thanks to the help of this forum and Allen/Coggen’s book.
You are probably on to something with the sample frequency as I had it set to 15 sec intervals.
Ah-ha! 🙂 Since your Polar was only storing every 15th sample, the 30 s rolling average would have been based on just two numbers, and therefore would have failed to adequately "knock down" very high values before the 4th order weighting was applied. The result would be an inflated value for normalized power, as the failure to "drag up" very low values (which would have happened just as often, assuming that your power was randomly applied with respect to the sampling times) wouldn't compensate, due to the non-linear weighting.
