Tilt Calibration (Was: 10 Coast Downs and 4 Mile Ride...)

[Norm]

The iBike 2 instructions indicates that once you've built a profile from the coast down and 4 mile ride data and uploaded that profile to the iBike, that will configure the iBike with the the following:
1) areo coefficient
2) fric coefficient
3) wind speed correction factor

But in step 8 on page 26, it also adds tilt correction into the discussion.

QUESTIONS:
Is a tilt correction factor sent from the profile to the iBike?
If yes, then why do we have to reset tilt at the start of each ride?

To me, it seems like tilt would be a one-time setup, assuming that the iBike is mounted rock-solid. It also seems that the iBike 2 software could determine a more precise tilt correction factor than the manual process with the 180deg flip. I'm missing some piece of this puzzle: which one?

Travis, could I get a ruling on which values are sent from a profile to the iBike?

[travispape]

The iBike 2 instructions indicates that once you've built a profile from the coast down and 4 mile ride data and uploaded that profile to the iBike, that will configure the iBike with the the following:
1) areo coefficient
2) fric coefficient
3) wind speed correction factor

But in step 8 on page 26, it also adds tilt correction into the discussion.

QUESTIONS:
Is a tilt correction factor sent from the profile to the iBike?

Not explicitly, but that information is wrapped into the fric const.

If yes, then why do we have to reset tilt at the start of each ride?

In order to ensure highest possible accuracy for your power readings on the road--esp. important for doing your intervals.

To me, it seems like tilt would be a one-time setup, assuming that the iBike is mounted rock-solid.

Rock-solid is better, but we don't take it for granted.

It also seems that the iBike 2 software could determine a more precise tilt correction factor than the manual process with the 180deg flip.

That's true

I'm missing some piece of this puzzle: which one?

I don't think you are missing anything significant. Do the 180deg flip proceedure to get the best power accuracy on the road, but in any case, the iBike 2 post processing make the result even better.

Travis

[Norm]

If yes, then why do we have to reset tilt at the start of each ride?

In order to ensure highest possible accuracy for your power readings on the road--esp. important for doing your intervals.

It also seems that the iBike 2 software could determine a more precise tilt correction factor than the manual process with the 180deg flip.

That's true

I'm missing some piece of this puzzle: which one?

I don't think you are missing anything significant. Do the 180deg flip proceedure to get the best power accuracy on the road, but in any case, the iBike 2 post processing make the result even better.
Travis

Thanks for the reply Travis. You're a font of iBike knowledge.

I'm looking for a bit more technical explanation though. Why do we expect the tilt value to drift over time? Is the accelerometer sensitive to environmental factors (temperature, barometric pressure, etc.)?

I understand the need to re-zero wind at the current barometric pressure. I don't yet understand why my tilt would change from ride to ride.

[racerfern]

I understand the need to re-zero wind at the current barometric pressure. I don't yet understand why my tilt would change from ride to ride.

My iBike is mounted rock solid using the banjo shaped piece of foam. On smooth roads, the unit will not budge. However, slam a pothole, expansion joint or something else unexpected when you are leaning on the bars and the bars themselves can move within the stem mount. Not much, but enough to throw the tilt slightly off. Of course the software corrects it.

I just did the Tour de Palm Springs century. 50 miles into the ride the road was so rough that the tilt was thrown off. Heck, I was almost thrown off the bike. At the next SAG I reset the tilt. Even though I knew the software would fix everything at home, I wanted to "see numbers that were accurate."

On local rides from home I check the tilt on the same spot I pulled away from at my house. It satisfies my curiousity that in fact nothing moved.

Sorry if this wasn't the technical answer you're looking for, but it's been my real world experience.

[travispape]

Thanks for the reply Travis. You're a font of iBike knowledge.

I'm looking for a bit more technical explanation though. Why do we expect the tilt value to drift over time? Is the accelerometer sensitive to environmental factors (temperature, barometric pressure, etc.)?

I understand the need to re-zero wind at the current barometric pressure. I don't yet understand why my tilt would change from ride to ride.

The accelerometer is not significanly affected by environmental conditions.

Depending on your mount, it might not vary from ride to ride, especially for the stem mount.

My first iBike setup with the standard size handle bar mount would sometimes vary a few tenths day to day. There was one ride where I hit a frost heave buckle in the road and I made a mental note of the mile marker. Later when I post-processed the file, I saw a marked change in tilt at that point in the ride. At other times I think it was just my old, tired stem and handlebars creaking to a new position.

On my new bike (Specialized Tarmac), I have noticed much less day-to-day shift on either the stem mount or the handlebar mount, but it still happens every once in a while. Note that if you mark your floor and do your tilt cal in the same spot every time that you can make a quick check of your tilt reading and if it reads like it always does after a tilt cal, you can skip it that day.

Also, within a given ride you will change the tilt of the iBike as you change your position on the bike. This has nothing to do with the pre-ride tilt cal procedure. This type of tilt variation is mostly due to the flexing of the bars-stem-steering bearing, but also due to tire squish. If you move from the drops to the hoods for a climb, or if you sit up to take a drink, just the change of weight distribution on the bars will cause a measureable change of tilt on the iBike unit. The slope measurement is that sensitive. That variation is removed in the iBike 2 post-processing.

Travis

[Norm]

Thanks Fernando & Travis! That's good info for an iBike newbie.

It's impressive that the accelerometer can detect bar/stem flexing, tire squishing, etc. and equally impressive that iBike 2 can identify and filter out these variances.

It's also good to know that the environment doesn't effect the accelerometer. I've been checking the tilt in my apartment after inflating my tires and I haven't noticed much day-to-day variance with my stem mount. I've wondered if the little bit of variance I do see is due to operator error/imprecision in the manual tilt calc procedure.

Of course, the bike is unloaded during the manual tilt calc. Might we ever see the day when iBike 2 will set the tilt calc in the iBike based on ride data (loaded bike) and notify us when the iBike tilt is out-of-tolerance when compared to the iBike 2 tilt calculated from the ride file? We could then save the manual tilt calc setup for on-the-road emergencies.

But perhaps in the end more tilt precision won't make that much of a difference on the road. How much of a watts error is introduced by each tenth of a percent your tilt is off?

[racerfern]

But perhaps in the end more tilt precision won't make that much of a difference on the road. How much of a watts error is introduced by each tenth of a percent your tilt is off?

Of all the variables that affect power I'm not sure which affect things more, wind or tilt. But I think tilt is even more critical than wind based on my observations. There is no doubt that the absolutely most important factor in determining my power output is me.

[Norm]

But perhaps in the end more tilt precision won't make that much of a difference on the road. How much of a watts error is introduced by each tenth of a percent your tilt is off?

Never mind. I just read in http://forum.slowtwitch.com/gforum.cgi? ... st=6882760,

...the error I quoted due to handlebar mount out-of-level was a worst case analysis; that is, if your handlebar mount is off by 10 degrees AND you are ascending a 15% grade, then the hill watts error ONLY is 1%; for hills less steep the error diminishes; on the flats, there is no error at all.

It looks like there's no need to get too torqued up about the accuracy of the tilt. Correct?

[Site_Admin]

I'm not quite sure what your points are here, but that thread is pretty old, and most of the cynics' questions have been answered.

FTR,

Do the CD's and the o & b and set your profile. It'll help with accuracy.

If you don't have time to do a tilt and wind cal, go ride anyway, and you'll get accurate data when you download. Doing the tilt and wind offset definitely helps, however.

[Norm]

I'm not quite sure what your points are here, but that thread is pretty old, and most of the cynics' questions have been answered.

No point -- just questions. Just trying to understand the how's and why's (along with potential future improvements) for getting the most accurate watts number possible. I haven't seen the old threads on this board; for some reason I though the 4 mile out-and-back was a new procedure.

My thinking was that if tilt isn't changing from day to day, then iBike 2 would be able to calculate a more precise tilt number than the manual tilt calc procedure and that might lead to a possible future improvement in on-the-bike watts measurement. But it appears that (1) tilt does change day-to-day (and sometimes even during the ride), esp. for those without a stem mount; and (2) a little bit of slop in the tilt number doesn't translate into much of a difference in on-the-bike watts (on my setup, a 0.2% difference in slope on a 12% grade climb shows up as a 10 watt difference).

Sorry if I'm wearing everyone out, but I've never been real good with blind obedience.

[Site_Admin]

I'm not quite sure what your points are here, but that thread is pretty old, and most of the cynics' questions have been answered.

No point -- just questions. Just trying to understand the how's and why's (along with potential future improvements) for getting the most accurate watts number possible. I haven't seen the old threads on this board; for some reason I though the 4 mile out-and-back was a new procedure.

My thinking was that if tilt isn't changing from day to day, then iBike 2 would be able to calculate a more precise tilt number than the manual tilt calc procedure and that might lead to a possible future improvement in on-the-bike watts measurement. But it appears that (1) tilt does change day-to-day (and sometimes even during the ride), esp. for those without a stem mount; and (2) a little bit of slop in the tilt number doesn't translate into much of a difference in on-the-bike watts (on my setup, a 0.2% difference in slope on a 12% grade climb shows up as a 10 watt difference).

Sorry if I'm wearing everyone out, but I've never been real good with blind obedience.

You got that right....

When you set the profile, and do everything pretty much the same, you'll be fine. Do the tilt cal and do the wind offset and do the temp stabilization and then go ride.
I don't believe that tilt changes from day to day, and if it changes during the ride, then you've got the mount set up wrong, or you've got a slow leak on one or both of your tires.

And if you pick up your front wheel while climbing a 17% grade, you're going to get bad watts, no matter what.

Quit obsessing and go trap some data.

[Norm]

Quit obsessing and go trap some data.

Going out now! Skeptic mode off.