Motorbike acceleration calibration

[pentavalle]

Hi,
I just tried HLT on a racetrack with my motorbike using an Android phone, I'm happy with results but I think I did something wrong with acceleration calibration because calculated lean angle was too low.
Phone is intalled with a tilt angle of maybe 100° (believe me, is the only possible way of mounting), so top of the phone is lower than its bottom, orientation is "portrait". During manual calibration the first 2 steps were automatically detected but I wasn't able to complete third step without forcing it: I accelerated reaching maybe 40kph and than stayed at costant speed for a few seconds waiting for phone feedback but nothing happened, so after a few tries I forced calibration while moving.
Should I accelerate more? Consider that bike has racetrack suspension setup and weight transfer is minimum when riding at low speed, so phone tilt doesn't change so much.

Thanks!
Realy happy with your software, waiting for v20 because video overlaying with Android is too slow to complete in v19!

[Harry]

It is hard to run individual calibration for motorbikes. In case you can't use the standard motorbike acceleration preset (see animated illustration 4 in Quick Reference), it is probably best to calculate acceleration and lean from GPS: select a lap from Lap List, select Recalculate and Acceleration. Adjust the slider so you get a smooth curve but without going too far.

Harry

[pentavalle]

Tried motorbike preset with same bike and mounting setup (100°tilt, no easy way of changing it) but lean angle was mostly wrong, sometime displaying bike turning while it was going straight or rapidly changing lean angle while it was constant.
The only way of having some sort of correct data was recalculating basing on gps with strong filtering due, maybe, to the not precise fixes of BT747 at 5Hz.

That bike is pure racing so it's not easy to make tries because I can only drive it in a racetrack, so I tried with another bike with a "flexible" RAM mount to replicate tilt conditions and noticed a, at least for me, strange behaviour: lean angle apart, longitudinal acceleration shows bike accelerating even if it's not running, all depends on mount tilt angle, keeping phone on a flat table shows, as on picture, that bike is accelerating forward:

Keeping phone in my hands and making simple tests shows that to have a perfect 0g acceleration on all axis I have to keep phone on quite 70° tilt but different tilt angles affect lateral acceleration, so the less the tilt angle, the less the lateral acceleration while still having phone standing still in my hand.
Pressing "Individual" button shows a text message indicating that a flat mount is best for a motorbike, so I am a bit confused....reading acceleration chapter PDF document didn't help me understanding while I'm not able to have a correct lean angle calculation on my bike

[Harry]

Using the Motorbike Preset, LapTimer will consider a tilt of 70 degrees neutral. When changing tilt, the effect is the same as when braking. This is the reason you get a negative longitudinal acceleration when putting the phone flat on the table. It is a tilt of 90 degree. Same when raising the phone upwards - this reduces the tilt angle and you will get positive longitudinal acceleration.

In case you have problems mounting the device with the angles shown as illustration 4 in Quick Reference, I'd suggest you use the GPS base calculation. Smooth the result so you do not get spikes to the wrong side of a corner any more.

There should be no statement a flat mount is best for motorbikes (except if written in context of a comparably "flat" 70 degree tilt). LapTimer's acceleration measurement does not work at all when the device is flat.

- Harry

[gplracerx]

Harry,

Do you correct for the reduction in sensitivity in lineal acceleration when tilted at 70 degrees? That's a cosine function and 70 degrees gives you only 34% of the sensitivity at zero degrees. I would think that the computing capacity of modern phones would allow you to use data from all three axes and any phone orientation would work once calibrated.

A 100 degree tilt is going to be even worse at 17%.

[Harry]

Yes, all axles are used and rotated according to the expected tilt etc. Once this is done, the longitudinal direction is extracted. Same for lateral for cars and downforce for motorbike lean angle.

- Harry