I have at the outset to say that I approach discussion of power metering in the context of cycling performance and training with something of a pinch of salt. For many years, I've used heart rate monitors (most recently a Polar S720i wristwatch and bike computer unit) to monitor exertion and effort during my training sessions and in racing. I have a pretty good understanding of where my training levels are (I am still using Pete Read's five levels, for this 0 to 4), and can judge not only where my limits are when racing on that basis, but also can judge when I'm not completely recovered and shouldn't train. This has actually worked pretty well, probably most notably in 2001-3. In recent years I've found that my working life has begun to limit opportunities for training (hopefully addressed now by moving to train before going to work!), and I've been seeking ways to better monitor progress.
In planning my 2009-2010 winter training programme, I'm a bit uncertain as to where power monitoring will fit in (though perhaps interested readers might leave suggestions via the commenting system). I've always felt that the advantage of HRM-based training is that you're getting a readout of physiological response to a training load, which seems to me to be rather more important that knowing what power one's generated. This is, I confess, an element of novelty spurring me on to train more seriously!
I recently bought a Polar CS600X bike computer/HRM, with an associated Power meter system - here I present an initial review of the unit. I later bought the add-on Polar G3 GPS unit for use with it - I'll review this separately as I haven't had a chance to use it at the time of writing.
Out of the box, the CS600X looks little different from any other modern bike computer - it's the usual streamlined shape with a large-ish display and several buttons. Annoyingly mine came without the bike mount. I suspect this was in error, but emailing the supplier elicited no response and I then left the country for a couple of weeks. I ended up buying several bike mounts anyway, as I plan to use it with several bikes. The unit is supplied with a speed sensor and the Polar W.I.N.D. power meter, which includes a cadence sensor. As supplied, the CS600X has already been "introduced" to these transmitters - the introduction process seems to be one where the parts exchange identification code to prevent cross-talk between two cyclists.
Installing the speed sensor is simple - much like other wireless bike computer systems. Since I decided I'd be using the system on my turbo trainer, I set this up on the rear wheel. Entering the wheel/tyre diameter is nothing unusual, as is entering one's personla data. And indeed, this can be done using the included Polar Protrainer software.
Installing the power system is a fair bit more complex, but not terribly hard for someone with reasonably good bike tinkering skills. Essentially it's in two parts. The simple part is the battery holder, which mounts on the right hand seat stay. This connects to the main sensor unit on the chain stay by wire. A second wire extends from the power sensor to the rear derailleur, where it attaches to a chain speed sensor. Attaching that sensor is a bit fiddly, and requires replacing the jockey wheel bolt with one that has a thread to which the sensor screws. This was the only real hitch, as my ancient Mirage mech seemed to have a strange bolt thread (or it was just badly corroded). In the end the only usable rear mech that I had kicking around was a 9 speed Record one, which does seem a bit over-specified for a turbo trainer bike.
There are clearly stated instructions for positioning the power sensor on the chainstay - this needs to be positioned correctly fore and aft as well as with respect to how close it lies to the upper chain run. However, this isn't particularly difficult, though it can be fiddly. Finally, the cadence magnet is attached to the crank arm.
As best I can tell, the system works in a manner analagous to an electric guitar pickup, and you need to enter a bunch of data to do with the number of links, the chain weight and the BB to rear hub distance. That would have been fine, but I was using a dirty old chain, so things got a bit messy! I have no idea how accurate this system is - anecdote would suggest that it's less so that the more expensive offerings on the market. I'm not particularly bothered whether 200W on this system really is 200W, but rather that the system is consistent and that values reported are proportionate to the real values.
Having set the thing up, it does look a little cumbersome, with wires looping around the place. I'd be unhappy using it on a sleek time trial bike, and indeed I'd imagine it might be susceptible to damage when transporting the bike.
In use, the first problem I faced was scanning through the different displays, which emphasise just how much data are available from a computer like this. Some of the screens seem unneccessarily complex to me, such as the default Power display, which has got all sorts of stuff like left-right pedalling balance and the altitude display, which includes inclinometer displays. You can of course customise at least some of these displays, though the "Zone Lock" display, which usefully shows what training zone one's in and how long one's been there seems a little refractory to customising. Unless, of course, I'm missing something.
I've been using this device for a week now, and have mostly been sorting out how best to use it, and such things as display customisation. I would say that thus far I've used it for HR-based training only. But on the other hand I'm getting a fix on what power I produce at a given intensity level (here I'm using a blurred perception-HRM derived idea of exertion level), and I am sure I will be able to use power values more appropriately as this winter's training proceeds.