There have been several questions coming up recently regarding tilt and riding tilt, and I thought I would try to give as brief of a summary as I can about the important points to remember regarding tilt.
Tilt calibration is done from the setup screen on the unit itself. The recommended method is to do it off the bike, but if you decide to do it on the bike, you must always do your tilt calibrations while on the bike. Again, it would be better to do this while off the bike in order to be as repeatable about it as you can. Also, don't be in a hurry--be still. Keep the bike fully upright near its balance point. The accelerometer is sensitive enough to pick up the minor movements of the bike.
Yes it is true that when you get on the bike, the tilt of the unit will change due to your body weight. This is called the riding tilt, and it is ok to have some riding tilt. The fric coefficient takes into account both the rolling resistance and the riding tilt in the firmware power calculations--this is why you don't need to worry about being on the bike to do your tilt cal.
(It is hardly worth mentioning, but it is true that the slope measurements you see on the unit while riding will under-report slope by the amount of your riding tilt. Again, this does not cause an error for the power reported on the road. Once the profile has been attached to the ride file in iBike2, the slope numbers will be adjusted.)
Riding tilt is measured on the day that you do your coastdowns and 4 mi ride. Depending on several factors including rider weight and bike stiffness, it can be close to zero or -1% or even more negative. Due to this riding tilt, you would expect to see the corrected and uncorrected tilt lines diverge in the window for processing calibration rides. The following screenshot shows a good measurement of riding tilt.
- good 4mi ride.png (45.82 KiB) Viewed 7992 times
On a daily basis when you download your ride files, the riding tilt is removed from the slope data as soon as you do something that causes the profile to be attached, usually either pressing the Auto Analyze button or selecting the Detailed tab page.
When you select the detailed tab and press the "Analyze Tilt & Power..." button and then press the "Advanced..." button, the graph that you see on the top is after the riding tilt has been removed. Ideally, the green line in the top graph is already close to 0 slope; however, in general there will still be some tilt error and the graph on the bottom shows the result after removing it.
Boyd's screenshot earlier in the thread for his own ride shows a good example of everything working well. Like he pointed out, the vertical scale is "zoomed in"--the magnitude of the numbers show little tilt error.
There are a couple of things that might have gone wrong for his teammate's ride in the other screenshot Boyd posted. The tilt calibration might have been bad on the day he did his calibration rides, or on the day of the ride, or both. If for day after day, the teammate sees the same behavior of the green line, that is a strong indication that the riding tilt saved in his profile is bad. In contrast, if the tilt error behavior is different day to day (but always bad from the beginning of the ride), that would probably be an indication that the rider isn't doing good tilt calibrations before some rides. Again, when doing the tilt cal don't be rushed, hold still, be balanced, keep the handlebars steered straight forward, don't be buffeted by the wind, be sure to get the wheels on the exact same spots when you tun the bike 180 degrees.
If you see the slope make a change during the course of a ride, like in the attachment below (and to a lessor extent in Boyd's good ride), that is an indication that the iBike unit or its mount is physically tilting slightly during the course of the ride. In the screenshot attached below, it happened suddenly at roughly mile 6. In Boyd's good ride, you can see the slope change in several places, but all of them are small. They also happen to cancel each other mostly for the total ride. If you consistently see the slope of the green line changing mid-ride like in the screenshot below, that is a good indication that you need to work on making a the mount more stable. The slight changes in Boyd's ride are par for the course and likely nothing to worry about.
- slope change.png (46.89 KiB) Viewed 7991 times
Keep in mind that it only takes a small amount of mount shift to show up as a tilt error. For numbers, say that the double-stick tape between the mount and the bars squishes in a wedge-like fashion by only 0.004", which is the thickness of a piece of 20lb paper. Given the 1.25" width of the mount, this results in a 0.32% tilt error and would create a very visible change in the slope of the green line. There are any number of places in the bike that might "creak" a little and cause slight tilt errors that can be detected by the tilt error. Given the hand-numbing vibrations that a bike experiences on the road, it is to be expected that you will sometimes get actual small changes in slope of the iBike unit.
Travis
I just fixed it.}