Polar 720i vs Computrainer readings



bikemery

New Member
May 25, 2003
2
0
0
I recently installed a Polar 720i Power kit for a customer and compared data while using on a Computrainer. The recording was set to 5 second increments on the Polar. The Computrainer responded to changes immediately and there was a bit of delay with the Polar. The Polar was 15-25% higher in wattage measurement shown compared to the Computrainer. Anyone with similar experiences?
 
Originally posted by bikemery
I recently installed a Polar 720i Power kit for a customer and compared data while using on a Computrainer. The recording was set to 5 second increments on the Polar. The Computrainer responded to changes immediately and there was a bit of delay with the Polar. The Polar was 15-25% higher in wattage measurement shown compared to the Computrainer. Anyone with similar experiences?

I installed the Polar power option on my bike 4 wks ago. I also own a computrainer. I had significant difficulty with the installation. I own a Kestral 200sci and the rear deraulier cable on my bike is routed along the top of the right rear chainstay- exactlly where the large power chain sensor has to be mounted. I was able to route the cable over the sensor but this could be affecting my readings. Two obsevtions. 1. I get different readings depending on the gear/chainring I am in. For example in my 39/ 23 the reading on the polar is consistently lower than the computrainer but not by very much. As I move into the middle cogs (eg. 17,16,15,14) while I am in my 39 chainring the Polar lags the Computrainer but it is fairly close in reported wattage when it catches up (w/in 5-10 watts of the Computrainer). As I go into the lower cogs (eg. 13,12) the polar starts to show higher readings than the Computrainer). If I am in my 52 chainring the polar shows wattage 20-25% higer than the Computrainer when I am in the 16,15,14,13,12. This is realy not acceptable. There are some anomalies on the road as well but harder to isolate or explain. The overall average wattage shown after downloading the data appears to be close to what I would expect based on speed/exertion on some of my standard/regular rides. I have added spacers,weighed my chain etc but all to no real avail. There are no warnings before you buy it about problems with different bikes. In the fine print in the instructions they do state that it may not work with certain bike configurations. The tolerances re keeping the sensor w/in 30 mm of the chain is very hard to accomplish even with a very standard configuration such as I have on my bike (ie. 12-23;39,52) The power package has been a real dissapointment but I do love the 720i watch. Its worlds better than the xtrainer plus was.
:mad:
 
I e-mailed Polar Finland and they say the height of the chain above the power unit has to be tailored per bike to get the most accurate readings. My only issue is that the average person does not own a Computrainer and a SRM to run comparisons to see if their Polar Power Kit is reading accurately. The value of anjy of these is once installed, you can see from one workout to another the changes YOU have. How it compares to other units is not really relevant at that point, just how was that workout value for you compared to other workouts with that same bike and are you improving or not.
 
I agree, but there is no comparability between different gear ratios. What I mean is that if I change the gear ratio and change the cadence so that I maintain the same power output on the computrainer I get a different power reading on the Polar. This makes comparing one workout to the next almost impossible. For example, if I do an out and back 10m time trial in 25 minutes on two seperate occasions and the 1st time I do the whole ride in my 52/17 and the second time I do it in my 39/12 the Polar will show a significantly higher average wattage for the first ride. This should not be the case. Also as to trying to set the chain sensor so that the chain is w/in 30 mm of the chain at all times results in my chain hitting the sensor in the 39/14-12 (with out keeping presssure on the pedals to keep the chain taught the magnetic pull of the sensor pulls the chain down onto the sensor) and outside of the 30mm range if I am in the 52/18-23. Using electrical tape or a piece of plastic like a chain protector saves the sensor but it does npt prevent the chain from getting pulled down towards the sensor. This causes the chain to get jammed up in the rear deraullier if you backpedal even just a little while this is happening.


I wish Polar would correct these problems. I have owned Polar watches/HRMs for over 15 yrs and purchased many for friends and family( at least 20 watches/HRMs). I have always felt their products were of the highest quality. This power option needs lots of improvement. I hope they offer some workable solutions to the current problems. Not to mention that the average consumer is not going to want to have/be able to go through all the work necessary to modify the setup to "improve" results. I say "improve"- as currently configured the system is unlikely to produce results anywhere near their claimed +/- 5% accuracy no matter how many spacers I use and no matter how many other compromises I make -like not caring that my chain drags along the sensor. The weight of the unit is the only compromise I should have to make.

Setting this unit up and attempting to minimize ridiculous readings requires creativity and lots of fabrication. The kit does not provide any material other than your standard thin rubber spacer to raise or level the sensor. This is not acceptable from a company like Polar



Art
 
Originally posted by ART
I agree, but there is no comparability between different gear ratios. What I mean is that if I change the gear ratio and change the cadence so that I maintain the same power output on the computrainer I get a different power reading on the Polar. This makes comparing one workout to the next almost impossible. For example, if I do an out and back 10m time trial in 25 minutes on two seperate occasions and the 1st time I do the whole ride in my 52/17 and the second time I do it in my 39/12 the Polar will show a significantly higher average wattage for the first ride. This should not be the case. Also as to trying to set the chain sensor so that the chain is w/in 30 mm of the chain at all times results in my chain hitting the sensor in the 39/14-12 (with out keeping presssure on the pedals to keep the chain taught the magnetic pull of the sensor pulls the chain down onto the sensor) and outside of the 30mm range if I am in the 52/18-23. Using electrical tape or a piece of plastic like a chain protector saves the sensor but it does npt prevent the chain from getting pulled down towards the sensor. This causes the chain to get jammed up in the rear deraullier if you backpedal even just a little while this is happening.

<Snipped>

Art

I have the same problem with power showing different variations in power across gears with some showing 70W higher than others for the same speed test. I am awaiting feedback from Alan Cote on of the Polar power meter developers on ways forward with this problem as I have the chain height within the 30mm tolerance above the chain tension sensor across all gear ratios. I think the issue is to do with the chain speed sensor incorrectly reporting the chain speed and thus the error in powers recorded.

I did 2 x 20 mins training session where the 1st I did avg speed of 18.9 mph with a power of 278W and then a second set a week later (rested) with avg speed of 21.9 mph with a power recording of 253W. The second session felt a lot harder and recording a lower wattage really dented the confidence in this system.

You may find Robert Chung's web site at this URL useful for Polar Power set up and testing

http://mywebpage.netscape.com/rechung/wattage/index.html

Regards,

Rick
 
I also hae the power option on the 720i, and have had similar problems. Inconsistent readings, not comparing well to computrainer.

I also am having a problem getting the device just to stay put to even register a reading. That is, the cadence maganet that goes on the end of the crank has to be so close to the power sensor to work, that I have it kind of on an angle, teeter tottering at 45 degrees on the 90 degree edge of the crank, held in place by what is now a loose plastic cable tie.

If the magnet moves during my ride, all power, cadence and %PI readings go to zero.

My bike is a pretty standard Trek 5200- anybody have any ideas.

Also, any other Trek 5200 (or 5900) owners out there have any luck with putting in the spacers under the sensors? What height of spacers worked best? Any help/input would be greatly appreciated!


BTW My view is that the computrainer is so much more stable and believable when it comes to the power things, but its great to go out for a day w/ your 720i, break your butt out and come back, download, and relive the experience on a Polar plot, with all the bells and whistles!
 
Originally posted by solidengineer
I also hae the power option on the 720i, and have had similar problems. Inconsistent readings, not comparing well to computrainer.

I also am having a problem getting the device just to stay put to even register a reading. That is, the cadence maganet that goes on the end of the crank has to be so close to the power sensor to work, that I have it kind of on an angle, teeter tottering at 45 degrees on the 90 degree edge of the crank, held in place by what is now a loose plastic cable tie.

If the magnet moves during my ride, all power, cadence and %PI readings go to zero.

My bike is a pretty standard Trek 5200- anybody have any ideas.

Also, any other Trek 5200 (or 5900) owners out there have any luck with putting in the spacers under the sensors? What height of spacers worked best? Any help/input would be greatly appreciated!


BTW My view is that the computrainer is so much more stable and believable when it comes to the power things, but its great to go out for a day w/ your 720i, break your butt out and come back, download, and relive the experience on a Polar plot, with all the bells and whistles!

solidengineer the Polar power system depends on the Cadence signal to be working to deliver power, PI%, L-R values as this is used for averaging. The supplied cadence magnet tends to slip down the crank arm due to the tapered shape. One solution is to use a magnet that fits into the back of the pedal spindle or use several zip ties to secure the cadence magnet in place, a form of harness, to stop it slipping back up the crank arm.

As for the Polar system on Egrotrainers then the power readings will always be inconsistent. This is what I had back from Alan Cote one of the systems developers:

A few things to consider. I don't know what trainer you have, but the friction between trainer roller and tire changes most everytime you install the bike, causing variations in resistance from session to session. Some fluid trainers change resistance as the fluid heats up too.

Also … the variation you saw in Polar's reported power across different gears can happen in some conditions - mainly if the sensor-chain gap is too large, or in some odd cases on trainers it can happen regardless of the air gap. But the gear variation is not an issue on the road from my experience. I suggest noting which gears are the "bad" ones on the trainer, and avoiding those for intervals. Ideally, stick with the same gear from session to session for the best comparison. Looking at the above you didn't use the same gear in session 1 and 2 - yes, this shouldn't matter, but for now it does on trainers.

Regards,
 
Thanks for the suggestion. I will look into the pedal spindle idea. Isn't the internet great. Advice from thousands of miles away - from someone with like interest!

Take Care

Marvin:)