tommyturbo wrote:
I too get some green lines that stay pretty flat for 4-6 miles then start sloping. I thought some of these had something to do with the ride itself. I generally ride about that distance, stop to meet some other riders, then take off again. I do nothing other than stop and if the unit goes to sleep, I wake it up when it's time to ride. I'm on eggshell smooth roads with no potholes (luckily). I wonder if barometric changes are causing some of this? I guess we'll have to wait for Travis.
I've read through the tread and actually I don't have too much to add. It looks like Fernando and Ron have helped sort out how to edit a profile in order to remove a consistent day-to-day trend in the top green line. (Top green line = the difference line you see when you click the "advanced..." button in the "edit tilt & power..." window.)
Regarding the need to edit riding tilt in your profile
Riding tilt is measured during the 4 mi out & back ride done after the coastdowns. The assumption is that the riding tilt is the same for both the coastdowns and the 4 mi ride, so if anything is happening to break that assumption, then you find yourself in the position that you need to edit the profile to remove a consistent day-to-day trend in the top green line. Some possibilities:
* A biggie: redoing your tilt cal between the time that you your coastdowns and your 4 mi ride. Do one good tilt calibration only. I'm just making sure...
* Another biggie: Removing your iBike from the mount and putting it back on between the time that you do the coastdowns and the 4 mi ride. Again, I'm just checking.
* Unlikely, but listed for completeness: Atmospheric pressure changes significantly during your calibration ride. Normally, this would be accompanied by wind, which you want to avoid in the first place when doing coastdowns. I doubt this is a problem in your case, but there it is.
* Another unlikely item included for completeness: If there were something loose in the front end of your bike (head tube, stem clamp, etc.) and it causes any play, the iBike will detect the tilt changes.
* As I've mentioned a couple of times now, I'm wondering if it is possible that if the steering bearings suddenly change position from their nominal position, it would slightly change the tilt of the stem & bars. Remember that all of the weight & torque of the front end of your bike pass through the bearings. Why would it change between doing your coast-downs and the 4 mi ride? I don't know--I'm just throwing all the possibilities out there.
* A problem that I think is more common than people want to admit: a mount that is not really rock solid. Sometimes it just really takes some patience to get the mount attached to your bike in such a way that it absolutely does not move. (2 part epoxy paste has done wonders for me in this regard since my bars and stem are not nice, round shapes.)
* When you push the center button to do a trip reset before doing the 4 mi ride, the mount could tilt slightly. Even if the mount stays rigid, the iBike could tilt within the mount some. I don't think it would change by much, but the iBike is sensitive to very small tilt changes.
* It might be that you are distributing your weight differently on the bike during the coastdowns than you do during the 4 mi ride. Actually, you know this has to be the case to some degree since you have to pedal during the 4 mi ride but don't pedal during the coastdowns. The force you exert on the pedals bears some of your torso weight, reducing the downward force on the bars from your arms. How much depends on your particular geometry and how hard you go at it during your 4 mi ride. I would say keep your effort during 4 mi ride well below your max effort to keep your weight distribution as close as possible to your weight distribution during the coastdowns.
Regarding a "Top Green Line" that suddenly changes tilt during a ride
Many of the items in the list above apply here too. I think in most cases, the tilt of the unit has actually changed. Checking the tilt at the end of a ride doesn't really dismiss this possilibity for a couple of reasons:
* It might be that in braking and jostling the bike at the end of a ride, the mount "pops" back into its original position. Again, I'm just throwing all the possibilities out there.
* The iBike display only show tenths of a percent on tilt, which is too course to disprove mid-ride tilt changes.
Some possibilities other than tilt changes:
* Like Ron mentioned, atmospheric pressure changes.
* Another possibility is that you turned a corner and changed wind direction. The altimeter reading is slightly sensitive to the wind due to the Bernoulli effect on the small hole on the bottom of the unit. Sometimes you can see a jump in the altimeter data at a point in the ride where you stopped suddenly.
* Perhaps as a battery gets low in voltage, it might cross some threshold that causes one of the sensors to change characteristics.
Again, I think the most likely explanation for a top green line changing tilt mid-ride is that iBike device actually changed tilt for some reason.
Tom, I think I've mention most of these ideas to you before in our previous phone calls and emails, but these are the best I can offer--we're not hiding any secret flaws of the iBike that could account for your issues with tilt. As unlikely as each above possible source of error might seem, you have to keep in mind that the bicycle and iBike mount & unit experience very large g forces during the course of a normal ride and it only takes a tiny change in tilt angle or altimeter data to create a "glaring" feature in the "Top Green Line". The green line is a filtered difference between larger signals, so tiny flaws in the data show up as large features in the plot.
Just keep the overall accuracy of the iBike in perspective relative to all of the sensor measurements and calculations that go into calculating the power. We've picked lots of low-hanging fruit in improving the accuracy of the iBike in the last year. In fact, we've picked some higher-hanging too, and we'll continue making improvements as the opportunities are created.
Travis
, but there seems to be some potential for working around one of these limitations (with known caveats), and it would be immensely helpful to us if we could get some input from you guys on this.
