Why I hate power meters

Peak event, May 3rd.

Today:

PT SL Torque offset 324
PT SL stomp test: off by 12.7%

Yeehaw. Now to see if I can get a response back from Saris and an act of God to get a powertap calibration turn around to Japan inside two weeks. &(*^%*^

__________________
Cycling Blog - Training with Power

Being a new SL owner I'm curious. How often does this happen? How long since your last service? How many miles/hours of you?

How did you notice the change, where your numbers looking GREAT, but too good to be true?

Also if you know it is off by 12.7%, could you not just take that into account post ride so you could no tbe without you PT for the event?

I felt a great difference between PE and wattage showed. (much lower wattage than PE). I couldn't really tell if it was legs or power meter, so when I got home I did the research and found out the info about the torque offset and stomp test.

SL data:
Purchased: Oct 2006
Serviced: Never
Mileage: 10344 km (the CPU is downstairs)
Hours: 350.1

Interestingly, I checked some ride files tonight to the ride file from today (same hills, etc). A 2.475 km hill I did in January was 8 min 5 sec at 340 watts. I did the same hill today, 2.475 km @ 8 min even - 338 avg watts. This leads me to believe it has been off at least that long. In which case I'll continue working with the consistent readings and chalk up today's PE to fatigued legs.

I'm looking at purchasing a backup PT Pro to have when I send the SL back to recalibrate. Worst case scenario I resell the PT pro and get most of the money - if not all of it - back.

__________________
Cycling Blog - Training with Power

That's a good question though. I dunno. I don't know if I can assume that the percentage is the same all the way up, or if the absolute value is different. I suppose if I check across the gear ratios on the stomp test, it will tell me if it is off a consistent amount of torque or a consistent percentage of torque.

Going to check that now....

__________________
Cycling Blog - Training with Power

Welp... God only knows. I guess practice makes perfect on the stomp testing. This time I got a road shoe, clipped in. Re-weighed myself and set to business with the wife holding the brake.

__________________
Cycling Blog - Training with Power

Well those numbers are ALOT better. Looks like your most error is in gear 13/52 at 3%. Looks like things are ok with your hub. So I assume it will not be going in for service now?

I am going to try the stomp test when I get a chance, but so far I have been very happy with my PT numbers vs. my Computrainer numbers. (as long as I do my part and properly RRC the CT)

My experience with the "stomp test" is that it is really hard to get consistent and correct results using your body weight and trying to balance. When I did this, I too thought my PT was reading ~10% low.

I then borrowed a couple of 25lb plates (weighed them to get the actual weight) and figured out a way to hang them from the pedal spindle while the bike was mounted in a trainer. I could then raise the weights to the point where the crankarm was horizontal just by SLOWLY turning the rear wheel while observing the torque reading and iterating back and forth (VERY SLOWLY!) for the max reading. Doing this, I found that the torque reading was within 1 N-m of the calculated, i.e. within less than 0.5% of the expected value.

Here's a thought....how old is your chain? When I was running my PT wheel in conjunction with my Polar, I observed a change in the difference between the 2 PMs of ~5-10W just by putting on a brand new chain. This difference doesn't show up in a "stomp test" since the chain is static...but once the chain starts moving, the driveline frictional losses are there.

That's kind of an "unspoken issue" with the PT system, i.e. that the power readings can be affected by driveline wear.

Good news that things seem to be working ok again (although it would have been nice to have that extra 12% wattage instead ).

Maybe your math was off on the original test, possibly off by 1 cog? Also, make sure the cranks are perfectly level during the test.