I'm trying to figure out which is more exact, a cyclometer or a gps. I'm using a Sigma cyclometer and a Garmin Etrex HCx in tracking miles, time, speed, etc. When I use them both at the same time they never have the same data. Anybody testing them with a greater amount of detail as to which is more exact? Any thought welcome.

How accurately have you measured your wheel circumference?

You are measuring different things. A cyclometer measures the rpm of the wheel as it passes by the sensor. since a rotation is a fixed distance, information can be calculated.

A gps measures point to point at timed intervals, so the gps records your position on the world, and assumes a straight line. Normally timing is 3 to 5 seconds and some gps systems allow you to change the time to increase battery life. GPX should be +/- 3 feet.

Even with gps, it depends on the frequency of data logging and the map topography used in the background.

I ride with Cyclemeter, a friend of mine rides with MapMyRide, over a 40 mile ride, he rides about 0.5 miles further.

I tend to like the gps system.

Old Asian proverb, "A man with two watches never knows what time it is."

GPX??? GPX is a file format. GPS is a satellite based positing system. Most better GPS bike computers include the capability for a wheel sensor because it improves speed, distance and auto-pause functionality beyond what can be obtained by GPS alone, due to GPS's inherent limitations and recording granularity.

Positing system??? Everyone knows it's a Global Postulation System.

Hey,
I've checked my wheels several times using the "mark on the ground method where I sat on the bike, followed a straight line and then measured the distance of the two marks. Luckily my garage is 50ft long so the measurement was smooth and straight.

I was not aware that the newer GPS's had wheel sensors also. I'm getting behind on my Technosavy

it was a late night for sure.

IMO a wired computer is the most accurate. I've used a hand held GPS ( not cycling specific ) and it seemed to show some distance traveled even when I was not moving.

Since you've done an accurate wheel rollout measurement the simple bike computer should give you more accurate results for distance and speed than any device that depends on GPS measurements. I still prefer to use my GPS since it gives me a computer record of where I went, and about how much elevation gain, plus it can tell me how to get back or where the nearest decent restaurant (or grocery store, etc.) is located. But for just knowing how fast I'm going or how far I rode any $10 bike computer will give me a more accurate answer.

I use both (strava for GPS on android) and my cyclometer to check the total distance while riding. Although they are off by a small bit, I do not really care about it. Who cares if they are reporting 3 km difference on a 70km trip

Both could be perfectly calibrated and still give you different answers. (Although I'm happy to get my cyclecomputer to within 1% of a measured roll-down mile.)

The GPS takes a measurement every 100 ft. or so, and calculates the straight line distance between them. The (perfectly calibrated) computer measures how far the wheel travels. So any wobble adds distance to the computer, but not the GPS. On a larger scale, if you go around a corner, the GPS could be off by cos(45 degrees), worst case, while the computer correctly measures how far you rode around that corner.

It really only matters when you're trying to follow a cue sheet. And the important measure there is how far off either is from the measuring tool(s) of the person(s) who put the cue sheet together.

I took my Garmin 800 out of a maiden voyage tonight and the speed display fluctuates wildly (jumps around by 1-2 mph) when in fact my speed was not changing as indicated by the seat of my pants and my cyclecomputer. No, I don't have the stupid cadence meter. Lots of other annoying features.

My Garmin 705 is set to one-per-second recording. Here's a ride that headed north, up the hill, and made a right turn on the return leg. You can see the distance and speed at each one second interval. It's pretty accurate, even showing which side of the road I was on.

If I'm riding downtown or with tall trees near the road, the limited satellite signals affect the position calculations, and the route drifts off the roadway.

15 mph is 22 feet per second, so the recordings would be that far apart.

Mapping from My Tourbook, free software.

All GPS software needs some smoothing adjustments to the raw interval data.

If I upload the same ride to Strava, ridewithgps, and My Tourbook, I never get the same total elevation gain on the three reports, and often even the average speed for the ride is different.

From March 15, 2014:
ridewithgps: 46.3 miles, 2758 feet, moving time 3:20:40
My Tourbook: 46.3 miles, 2982 feet, 3:26 moving time.
Strava: 46.5 miles, 3126 feet, moving time 3:22:18

None match, but it's all reasonably close.


~~~~~~~~~~~~~
For instance, here's another ride's raw data, with spikes in the speed, and big spikes in the feet/hour calculations.



~~~

The My Tourbook software has adjustable smoothing settings. Too little, and the results are spiky. Too much, and all the minimums and maximums are averaged away.

Here's my settings that don't over-average:

The velocity data is only useful in its raw form like in real time.

OP: The greatly fluctuating speed output of my GPS Garmin 800 is annoying to me. Likely better with a cadence sensor

My favorite feature on the maiden voyage last night was being told I was off route when in fact I was right on the right road that I cycle every day, then it routed me in the wrong direction/wrong road. Who knows why.

I have a Garmin 810 and an iPhone 5 with Cyclemeter. Last year, I bought the bluetooth sensors from wahoo fitness and went exclusively with Cyclemeter. I like it better for the better display, the increased functionality and it seems to be just as accurate as the Garmin. Since I have to carry a phone anyhow, it's just one less thing I need to have with me. Since I use the quadlockcase mounting set up on the stem, then it's one less thing in my jersey pockets. What's not to like?

Cyclemeter seems to be quite repeatable. If I ride the same route, I get the same altitude gain and the same distance (varies a tiny bit, but I presume I do too). So I'm pretty happy with it.

J.

If you don't have a cadence meter, there will be a data lag because it takes time for the info to reach from the satellite to you. A cadence sensor is by far the most accurate way to measure both speed and distance.

I have an older Cateye wireless bike computer and I usually ride with Strava running on my iPhone.

I find the bike computer very accurate, WHEN it registered. Being an older wireless unit, it will occasionally get interference and speed shows up as 0 or something much lower then I am traveling, which impacts distance. This is typically enough to loose a few tenths of a mile to a mile on a longer ride. I assume this would not be an issue with a wired computer or a newer ANT+ wireless computer, but who knows.

Strava seems to give me very consistent distances every day. The only catch is to make sure it gets my location locked before I start or it might mess up the first part with a sudden big move. Having wifi on and giving it 5 seconds or so usually takes care of this for me. I have tried some similar apps as well (map my ride, etc.) with good success as well.

My only complaint with Strava accuracy is elevation - its not consistent, maybe a 5-15% variation on the same route I ride regularly. no idea why. On rare occasions, its way off, 20-30%. Typically on these days it puts me on the map as though I rode 20-50 feet off the side of the road, which is often down a big drop and messes with elevation readings I guess. This isn't common though.

I guess a bike computer with an altimiter would be more accurate.

Not sure how well a dedicated GPS bike computer would be in comparison - proboably better then either of the above as it could have sensors plus GPS, and software could smooth the two out.

If your wheel circumference is accurate, the cyclometer will be more exact for speed and distance. The Garmin Edges can take measurements every second. In that time, your 700c wheel rotates about 3 times at 15 mph. That means the resolution of the cyclometer is 3 times than that of the GPS. Additionally, the variation in position of the GPS is 3-15 m (much larger than the error in the wheel rotation rate, which is close to zero).

Keep in mind that the GPS first determines position (earth surface coordinates) at different time points and computes speed (and distance) from that. Thus, the speed and distance are secondary results of a GPS. Given that the GPS is a self-contained unit (it doesn't require extra sensors), many people might accept a small reduction in speed accuracy for the convenience.

??? No, that's not the reason for the variation in speed.

You don't really understand the tool.

Garmin | What is GPS?
GPS.gov: GPS Accuracy

There is variability in the position measurement of GPS receivers. Garmin says +-15 m, the FAA says +-3 m for "high quality" (large, expensive) GPS units. (The variability can be reduced if the device can WAAS or differential GPS.)

Given the unavoidable variation, it's expected that the speed will "jump around" a bit.

A 1-2 mph speed is a change of 0.447-0.894 m per second (much less than the variation expected).

(At high rates of speed, like in a car or airplane, the GPS determined speed is very accurate.)

A variation of 5-15% in elevation is very good. Strava is using the GPS coordiates to determine elevation and GPS isn't very good for measuring small differences in elevation (the horizontal variation in GPS position is 3-15 m and the vertical variation is likely larger). Barometric readings improve elevation gain measurements (but I don't think many smartphones include a barometer).

I believe if you add the Wahoo Fitness RFLKT+ display to the system, it does add barometric sensors that Cyclemeter knows how to use. It also adds temperature.

J.

Interesting.

Wahoo Fitness - RFLKT Bike Computer- Bike Computer for iPhone

Abvio | 10.2.1 Cyclemeter with Elite

"RFLKT and RFLKT+ – Supports the RFLKT and RFLKT+ bike computers, including temperature and altitude sensors."

J.