Polar Power measurements - is this right?

Larry Klassen

I installed the power measure from Polar, it always has a reading despite not being within the distance as they say. I had posted on another website and others said it does not need to be as close as Polar recommends to the chain (I would have had to build the sensor up by about 1.5cm to get it within the distance they say, then the chain would be running along it at all times in a small rear/small ring combo).

First off, I have only used this on a magnetic resistance trainer (Minoura) as it is now to cold to ride outside here in central Canada. Anyway, I get readings that seem fairly accurate, in the mid-200's, during good effort (I'm not world-class or anything). I don't care how accurate it is, as long as it is internally consistent. At any given cadence, though, staying in the same gear and staying roughly the same cadence, it can vary by 50watts - it can average out at say 220 or so, but drop to 185 or go up to 250. Is this normal for this thing? Why so much variability, or is it that I don't have a smooth enough pedalstroke? Do I just go for the average over longer times (ie. over 10 minute or more periods) as a more accurate measure, not really looking at individual moments?

Am I making this clear? The power seems to vary a great deal despite sustained effort. Also, the power doesn't seem to go up all that much despite going up alot in effort, going to a bigger gear with same cadence, etc. - any thoughts, anyone? Thanks in advance for any advice/help.

Mansmind

I feel your pain lol


I have the Polar unit also. It can be quite finicky on trainers. I've noticed the same variations your seeing, the average seems to work out well, but it's hard to see where you're at many times.

I think:

1) At certain wheel speeds, trainers create a vibration that confuses the unit. You can feel a trainer feel "rough" then smooth out with a little more speed, or vice versa.

2) I've had better results using a higher resistance setting on the trainer, and generating wattage at much lower wheel speeds. Resistance that gives you ~300 watts at 20 mph.

3) Pedaling style seems to create some of the fluctuation also, but I don't think it affects it nearly as much as the vibration.

4) There is a 5 second period between each measurement. As you make adjustments trying to get into the "window" you're looking for, it's very easy to miscalculate since the reading of what you're doing now, doesn't show up for potentially another 5 seconds.

I get MUCH better results on the road. Polar uses chain tension and speed (as I'm sure you know) to calculate power. My guess is that chain tension can vary greatly even in one complete pedal stroke most people apply the majority of their power in the 70 degree range (out of 360). Granted you have two legs, so you're really creating that tension in 140 out of 360, but that still leaves a lot of the stroke in which a much lesser amount of force is being applied, which would reduce the tension in that part of the stroke. Variation would be hard to avoid I think.

John

Larry Klassen

Those are all thoughts that I was thinking as well - I really mostly needed to hear from someone else (any others out there?) that are experiencing the same thing I am, and that it isn't way off or I'm doing something wrong. I can live with it as long as I know I'm not doing something wrong, and it is the same experience as others.

I am getting a Fluid2 trainer this next week, so I suppose that could theoretically help at least a little. I'm sure it won't be nearly as good as if on the road, but hopefully some improvement will occur. My mag trainer tends to vibrate at higher speeds (I think it isn't as smooth as it used to be), so that is likely playing into it as well.

Thanks for the input, I appreciate anyone else who has experience as well.

michaelk001

Nope. I've used it on Fluid2 with the same results - very poor. Get a PowerTap!

Weisse Luft

Yes, the power measurements are problmatic on trainers. You need to find out what equivalent road speed produces accurate results and live with it because the other features of the PPS outweigh this limitation.

Pedaling index is one such feature. This helped me fix a right knee problem and now I ride with no pain after longer rides.

I did a time trial where I lost my wheel magnet but I still had power readings. Ignoring speed and concentrating on power HR and cadence was far more helpful than watching speed. By plotting elevation changes, I was able to estimate speed on the data file.

Mansmind

UPDATE!

I discovered a problem with my installation of the power sensor. The chain speed sensor can be very difficult to place in such as way so that it doesn't intefere with anything. I THOUGHT I had it (after looking at pictures of other people's installs, etc.)

I discovered that while in the smallest cog on the back wheel, the chain sensor was actually in contact with the rear derailer (Ultegra 9 speed). After playing around with it for a while, I ended up placing the chain speed sensor
so that it's just below the area on the derailer that almost completely covers the chain. In this position, it's actually reading chain speed as the chain is on the pulley rather than it being part on/off depending on gear selection. It doesn't seem to interfere with any gear I'm currently running, large or small chain ring.

I tested it on the road yesterday, and found the power readings to be significantly more consistent, although it didn't seem to change the overall power reading.

I tried it on the trainer this evening, and while I still see more variation as compared with road riding, it's MUCH better, and much easier to control. It also seemed to give me a very consistent reading regardless of which gear I chose to generate power in. I still saw a weird peak or valley now and then, but not nearly as often as before.

Of course, if all you guys are already at this point, then I'm just catching up. If not, it might be something worth looking into.

Let me know if it helps!

John

spomanski

I've been using the polar power attachment for some time and I have some observations.
First off you cant get an accurate reading on an indoor trainer. The magnetic sensor needs for the chain to vibrate as it would on the road.
Secondly, you'll get different readings from the unit from the small and large chainrings. This is due to the different distance from the unit on the chainstay and the chain itself.
I just purchased two power tap computers. One for the bike i keep on my indoor trainer and the other one for the bike i use out on the road. The Polar watches are great, the software that comes with them is excelent, but the power attachment leaves much to be desired. It cannot be used for real serious power based training. But I"ll say this. If you want something to measure your progress from month to month it can be usefull. All you have to remember is that you need to be in the same gear from measurement ride to measurment ride. Then you'll get some reliable readings.

Weisse Luft

Anyone try using a cyclocross tire on rollers?