Will Polar make a good power meter someday?

Alex Simmons

Hmmm. I'd go with that except at lower powers it seems OK, it's only when power goes up that it looks wrong.

Maybe an endless reel of the Dr Who sound track pointed at the cockroach

__________________
http://alex-cycle.blogspot.com/
http://www.cyclecoach.com/

jcjordan

As I have had a CS600 with power now for about 8 months I thought it would be good to put my 2 cents in.

I have found the unit to be fantastic and highly accurate as long as you are careful with your set up. I even borrowed a mates powertap and compared the readings of a ride up Black Mt in Canberra (both units running at the same time) and the two units were only 2-3 watts in difference. The polar was about a 1/8 sec slower to pick up changes, but I think that would only effect the top 1% of riders.

There are only two things that you really need to do with the polar to make sure it works well. Firstly, make sure that you weight and measure the chain every time, don't rely on default values. Secondly, find a reliable set of batteries. I have found the storage unit does not really like rechargeables, but does like Energizer lithium.

__________________
"Pain does not last, bones heal and chicks dig scars"

Tom Anhalt

Well...at a given cadence, a higher power will mean a higher chain tension. This also will mean a low chain vibration amplitude as well. If the "environment" is too "quiet", the chain isn't getting "excited" enough to vibrate at a larger amplitudes. The inductive sensor can't tell the difference between changes due to the chain going up and down vs. changes due to things like the pins of the chain passing the sensor. The circuitry just does it's job and locks on to the largest signal...if it happens to not be the chain vibration...well, things will be off.

Take a string or a rubber band and stretch it between your 2 hands. Pluck it to get it vibrating and then increase and decrease the tension and look at what happens to the amplitude of the vibration and you'll see what I mean.

On the road, the random inputs of the road surface keep the chain vibrating at a sufficient amplitude so that this isn't a problem (random and impulse inputs tend to excite ALL frequencies equally in the range encompassed by the random input). This is why I speculate there's a difference between "on road" and trainer performance with the Polar.

Now...with the fact that the Polar unit actually has a chain speed sensor already, I would think it would be somewhat trivial to design the circuitry to reject any signal that's the same frequency as the chain speed. But, then again, I'm not a signal processing engineer and it may not be that easy...



If the input energy was high enough, that might work. However, I have no idea what effect it would have on the sanity of the rider forced to listen to the "loop"

Hmmm...I know that there are white noise CDs available...I wonder if I can rig up an experiment to prove/disprove my theory? First, I'll need to make sure that the noise is in the same frequency range as my expected chain vibrations and then I'll have to figure out how to get the energy into the chain....

footsie

I have found that duracell rechargables work fine, I place them in before each ride and charge them straight after so far no problems

lanierb

My Polar was okay on hill climbs but wildly inaccurate on the flats. I think the main issue was that it read differently in different gears (despite hours and hours of install tweaking). I like the Polar head unit better than my PT head unit so it would be nice if Polar came out with something more reliable.

Tom Anhalt

Just out of curiosity, but how did you measure the differences? Also, what was your chain to module distance in the big ring (I'm assuming you were in the big ring on the flats)?

Alex Simmons

Got it ! Ta

It would have to be one hell of an amplifier-speaker set up I'd imagine

__________________
http://alex-cycle.blogspot.com/
http://www.cyclecoach.com/

jcjordan

Accuracy does not seem to be effected on my unit in regards to flat or uphill, even works well on the trainer with.

It does sound like you don't have the paddle set up right. This is the primary reason for any faults and if you have non straight stays like I have on my bike the set up can take quite a bit of time to get right.

__________________
"Pain does not last, bones heal and chicks dig scars"

RChung

If you had a way of knowing which chainring you were using you wouldn't need the chainspeed sensor.

lanierb

Chain module distance was close. I used some pieces of plastic to lift the chain sensor up very high -- to the point that it rubbed when using the inner chainring on smaller rear cogs and always sucked up the chain at stoplights. It was still unreliable in the big ring. This was on my Trek Madone 5.9. On my old bike (Trek 2300 aluminum) it worked better. I eventually measured the differences using a powertap but I had already figured them out by doing some tests using 3 different gears and holding a constant speed for a minute in each one on the same course. Another thing I found was that the accuracy was better when my chain was dirty/old. A clean new chain somehow led to errors (again on my Madone). It was all too weird so I bought a powertap (eventually 2) and everything has been great since then. I'm confident I tried every option on the setup and had it setup "right". At the time I read everything everyone had to say on the issue including trading e-mails with the inventor (was it Coates or something like that?). It just plain didn't work on that bike. I suppose it's possible something else on the bike was "singing" and causing interference.

jcjordan: how do you KNOW that accuracy is not affected by flat or uphill? Have you actually tested it using something reliable? It's not that easy a thing to do without a powertap/srm to compare. How do you know it isn't, for example, reading 5-10% low on hills relative to flats? This could be hard to know for sure from PE.

veloventoux

The simple true reason why it doesnt work on some bikes is , as I've said before, that the shape of the chain stay doesnt allow the paddle to be // with the chain. If you install it on a bike that has straight chain stay than it works fine.IMHO the chain has to "cover " the entire length of the paddle in order to give reliable data.