I bought a used road bike that had a Polar CS200 computer already installed. I’m curious how accurate the speed and distance functions are on these non-GPS units. Has anyone taken the time to actually test how accurate (not just the Polar, but biking computers in general) these things really are?
How Accurate Are Speed & Dist. on Biking Computers?
1 - 20 of 47 Posts
Joined
·
32,560 Posts
If it is not using GPS--speed/distance is as accurate as the wheel rollout measurement (measurement of the circumference of the tire under rider load). The speed sensor only counts how many times the magnet passes. You multiply the number of wheel passes by the wheel rollout and you have distance, you divide that by time and you have speed.
I fully understand who it works, my question is how accurate are they. The manual tells you how to enter the tire size, that which is stamped on the tire. If they are setup per the manual with proper tire inflation, how accurate are they?
If they tend not to be accurate I know I'll need to tweak the tire size.
If they tend not to be accurate I know I'll need to tweak the tire size.
Joined
·
32,560 Posts
It depends on how much you weigh, inflation pressure, and the tires you use.missin44 said:
How much you weigh and PSI alters tire deflection. You sit on your bike-you deflect the tire shape changing the rollout length a bit. No two tires are the same circumference to start with either. The inaccuracy in wheel-sensor-based computers derives from those factors. There is no universal answer therefore as to how inaccurate they are.
Generally, your bike computer will be a bit more accurate than your car's speedometer presuming a reasonably accurate rollout measurement. Of course, your car is not measuring wheel rotation it is measuring transmission speed.
Joined
·
32,560 Posts
Your car mileage is more inaccurate than your bike computer, assuming a decent rollout on your bike.francoaa said:
Your bike computer takes a wheel circumference and calculates distance from that. It actually measures the distance you travel from known and measured quantities.
Your car measures the transmission speed. Assumes appropriate inflation. Assumes weight of load for deflection calculation. It also assumes tire AND wheel size. Your car odometer/speedometer NEVER actually measures distance. It infers it-albeit good enough for most purposes. You also are unable to correct for flagrant flaws in all the assumptions in your car.
Joined
·
1,854 Posts
Most cars base speed/odometer on a gear in the transmission. If your tires are not inflated at the manufactured spec, not the original size, have any casing deviation etc.. they can be off too. The method that bicycle computers use is the same as a measuring wheel, which are pretty accurate.
Joined
·
2,352 Posts
When measuring circumference, inflate tires appropriately, put chalk mark on floor and tire where the tire meets the floor.(you'll need a friend here) Under rider load, move the bike in a straight line until one full rotation is achieved. That length is your accurate circumference.
Avocet used to claim 99.9% accuracy with their cyclometers, presuming you performed a rollout measurement. I think that's accurate enough.
I went on a ride today with some friends. Afterwards, we compared distance numbers. Between the two of us, the variation was 2.8%. Close enough. I have no idea whether the friend used a rollout measurement; nor can I remember if I did.
I went on a ride today with some friends. Afterwards, we compared distance numbers. Between the two of us, the variation was 2.8%. Close enough. I have no idea whether the friend used a rollout measurement; nor can I remember if I did.
Although I've used the much-touted "roll out measurement" to calibrate computers in the past, the last 6 or 7 Ive done (many bikes in this family) I've simply used the pre-programmed code for the tire size. I've found it to be plenty accurate. When checking with my Garmin gps, it's generally within a few 1/10 mile in a 20-30 mile ride. I have absolutely no need to be any more accurate than that. I could care less if my ride is 24.8 or 25.3 miles.... It's "about 25".missin44 said:
I remember once checking against the mileposts on a local highway on a 30 mile ride and it was within 1/10 mile every 10 (similar to the above GPS comparison).
My advice is to just use the pre-programmed code and you'll be fine. The roll-out measurement is a waste of time and only for those who think precision is much more important than it is ... and they probably don't get significantly more accurate measurements from that method anyway.
I use a Cateye Microwireless (the best computer I've ever used) and I think it's VERY accurate. I was within 2/10 mile of someone with a GPS on a century. For all I know he might have looped around the parking lot once.
In other words, bike computers are as accurate as car speedometers, maybe more so, so long as you're putting in the right wheel measurement.
In other words, bike computers are as accurate as car speedometers, maybe more so, so long as you're putting in the right wheel measurement.
Joined
·
5,480 Posts
Inaccuracies would really lie in tire differences. I'm assuming the original calibration by the factory would be based on a "typical" 700x23c tire. There's really too many differences between brands and sizes to get the 100% correct readings. Some computers allow exact, manually-inputted measurements.
The value of such inaccuracies is little. It will get larger over longer distances, but quite frankly, you can't always squeeze in the extra mile to make up for it on the route just like that. For speed, you can use whatever numerical values as references for pacing.
The value of such inaccuracies is little. It will get larger over longer distances, but quite frankly, you can't always squeeze in the extra mile to make up for it on the route just like that. For speed, you can use whatever numerical values as references for pacing.
Joined
·
24,573 Posts
It's only as accurate as your roll out measurement. You must weight the bike, so you get the proper "tire drop". One way to do it is to lay a tape measure on your driveway, get on the bike and push with your toes, three revolutions of the front wheel ) or back, if you have a rear wheel pickup)
Divide this number by three, convert to mm, and this is your circumference. If you want to be dead on, make this measurement, three or four times, and take an average.
Wireless bike computers aren't as accurate as "wired" computers, because power lines can interfere with the signal.
Divide this number by three, convert to mm, and this is your circumference. If you want to be dead on, make this measurement, three or four times, and take an average.
Wireless bike computers aren't as accurate as "wired" computers, because power lines can interfere with the signal.
Joined
·
9,699 Posts
I'd guess they are very close to 100% accurate for measuring the way they were programed to.
I have one 50 mile ride I do quite often and I always pump the same tires to the same pressure before the ride and the measurements are always within 2/10th of other times and even that small variance I would explain by not going in a perfect straight line and not taking corners the same every time.
So who cares as long as it's consistant.
The average speed is all messed up on those things though. It's generally way to low so I add a few MPH before I post about it on the internet.
I have one 50 mile ride I do quite often and I always pump the same tires to the same pressure before the ride and the measurements are always within 2/10th of other times and even that small variance I would explain by not going in a perfect straight line and not taking corners the same every time.
So who cares as long as it's consistant.
The average speed is all messed up on those things though. It's generally way to low so I add a few MPH before I post about it on the internet.
Joined
·
87 Posts
Accurate enough. 1-2% if you ballpark the circumference, better if you measure your tires with a measuring tape.
Joined
·
93 Posts
ok i have been wounder this also
does it matter here on the spoke you put the magnet if it lover on the spoke it would read faster is what i thinking
is this true or not
does it matter here on the spoke you put the magnet if it lover on the spoke it would read faster is what i thinking
is this true or not
LOL!! The average speed is your average speed and is just as accurate as your measured distance is, it's just that when you are starting and stopping a lot it has a big effect on the average.
I use average speed over particular routes to measure performance and improvement, and although it varies from day to day according to wind and how often I need to stop, I know that when my average speed over a certain route is 0.5 - 1mph faster consistently (i.e. from week to week) I am fitter. This is well within the range of error on the cycle computer if I have not done a roll out test. I get a 2% difference on the roll out test from the standard recommended figure for the tyre size (the recommended figure overestimates my speed and distance).
The method I use for the test is to pedal the bike slowly for one tyre revolution indoors while using a wall for support (if you get the distance from the wall just right you can be almost, but not quite, balanced on the bike the way you would be when riding with your shoulder touching the wall). Repeat several times and average.
Not... the computer just counts the number of wheel revolutions and uses this to measure distance and speed based on the tyre size. The number of revolutions is the same whether you are looking at a point on the spoke nearer to the hub or nearer to the rim. The magnet will be traveling faster if it its nearer to the rim, but this isn't what the computer is measuring.
Does the hour hand of a clock point to a different time at the tip from what it does at the bottom?
1 - 20 of 47 Posts
