I bought the Polar Power Meter a few weeks ago. Hey, why not...Since I already have the Polar 720i. I checked into several user forums, and the only complaints were regarding set up, with a couple over mechanical reliability.
Set-up: There were two issues, one that I was aware of ahead of time and one that was a surprise.
1) setting up the chain sensor requires a ruler and removal of the chain. This was the setup issue I knew of. The chain tension sensor uses a magnet attached to a strain gauge. It senses the defection of the chain under pedal induction (this is why you need to weigh the chain). It then uses that defection data, with cadence, chain speed (a separate sensor), road speed, and rider weight to come up with a power value. You need the ruler to ensure the distance of the chain to the sensor - which sits on the drive side stay - stays within distance limits across the gear range, or the magnetic strain gauge will not react to the chain deflection properly. You get spacers to accomplish this.
2) What I didn't prepare for was the lower jockey wheel bolt length. The system measures chain speed through a sensor attached to the lower jockey wheel, so they provide a special bolt that is unfortunately a little too short and will only grab about 1 thread. With a little judicious use of a dremel tool, I was able to shave about 100 mils off of the shoulder of the bolt so I got it to come though just shy of flush with the back of the cage.
It worked right away. So far the only problem I've had is that the wrist unit needs to be completely, perfectly, and firmly seated on the mount. I'm thinking this is the source of alot of the 'intermittent' complaints I've read about.
Since there wasn't any real calibration of the system, I decided to do a little test. Last night, I went out on a flat straight stretch of road with a small downhill run-in. Using the hill, I got the bike up to 30 mph and held it for the length of the stretch of road (slightly under 1/3 mile). The test took about 40 seconds, which translated to 8 samples on the polar.
I averaged 29.96 mph and 427 watts. Then, I went to the on-line bicycle power calculator, punched in all the data, and came up with a theorhetical requirement of 432 watts.
Granted, the numbers in the calculator are assumptions, but if you look at the actual co-efficients, the differences in the margin of error are not that great.
Another feature of the power sensor is pedaling analysis. It shows a right-left balance and what they call a 'pedaling index', which is "the ratio between the minimum and maximum forces of a single pedaling cycle". The polar website says 30% is 'very good', the higher the better. Over the several rides I had the new unit, I'm getting about 25% on average, since over 30% is 'very good, does that mean under 30% is _not_ very good?
I hope not to become a slave to the power meter, but I wouldn't mind becoming a slave to this: