Lateral/longitudinal data incorrect?

subawang · Post by **subawang** » Wed Aug 08, 2012 5:32 pm

I got back from a lapping session at Gingerman and want to do some analysis of my laps (using v15.0.5 profession with iphone4S). If I export the data and plot the lateral (x) vs longitudinal (y) acceleration in order to compare to a traction circle, it looks like all the data is shifted and rotated.

Question - why is my data not look right and how do I fix it?

I suspect that the device not be mounted perfectly parallel/perpendicular to the car is throwing off the calibration. Do I need to adjust the calibration? Even so, I don't even see data for when the car is accelerating?

I've tried to attach an image (why is xlsx not allowed?) for a given lap showing the problem. Thanks in advance for help.

On another note, with only 1 Hz sampling, this app really is not so useful for data analysis. I figure the 1 Hz is limited by the gps update which is slow for the iphone's built in gps, but aren't the accelerometers 10 Hz or more? Is there plans to somehow save the higher sampling rate of the accelerometer for data analysis. I'm not looking to upgrade to external gps since I am happy with the accuracy of the actual lap timing, but I wanted this app more than just lap timing.

Post by **Harry** » Wed Aug 08, 2012 5:45 pm

How do you plot your data? Keep in mind that while latitudes have the same distance across the globe, you need to factor longitude depending on your latitude.

Acceleration is not currently used to interpolate / extrapolate position. There is work in progress, but I cannot give a schedule yet. Currently, if you want finer plots, you need an external mouse. Please search the forum first, there are plenty of discussions already.

Harry

subawang · Post by **subawang** » Wed Aug 08, 2012 6:00 pm

Sorry, I not talking about position on the earth but acceleration forces. The terms used within the software are lateralG and linealG. But the issue is that the traction circles is not displaying correctly.

Post by **Harry** » Wed Aug 08, 2012 6:33 pm

Ok, sorry, didn't see that. Have you checked calibration has been correct when recording. The lease you should see is that not lineal and lateral acceleration are nearly zero while standing still. The device must not be mounted flat, most other tilts and turns are compensated by LT calibration. Please check the Acceleratuon view plus the Assistent you get when pressing the small wrench.

Harry

subawang · Post by **subawang** » Wed Aug 08, 2012 7:13 pm

Hi Harry,

I appreciate you trying to answer my questions. I think the issue arises from mounting the timer in a way which is flat, but not perpendicular to the front to rear axis of the car. If you imagine mounting the phone to the windshield in the center, but angled toward the driver. All acceleration will show 0, but as soon as you accelerate or brake the values will be rotated slightly about the origin.

I'm thinking this is my problem, and a recalibration at this orientation is necessary. Or another option would be to mount the phone perfectly facing the rear of the car - not oriented toward the driver.

Even so, this does not really answer why the data seems shifted in addition to rotated. But to try and separate out the issues, how should I go about calibrating the software for a slightly off-center orientation?

Post by **Harry** » Wed Aug 08, 2012 8:29 pm

Step 3 in LT's calibration eliminates the rotation that results from the turn to the driver. Please note that step #3 is triggered only in a narrow corridor of driving straight and decelerating. You will have this on track only, e.g. when braking at the end of the start finish line.

I have no chance to analyze your data (on the road) but suggest to check proper calibration steps next time you are on track.

Harry

subawang · Post by **subawang** » Thu Aug 09, 2012 1:58 pm

Harry,

I agree that the calibration must not be right. I've read a few posts where you address that, but I don't see how I have any control over it.

Within the acceleration mode, I see that step 3 is driving/braking in a straight line. You mentioned that this can only be done on track, and if I understand correctly, will be and can only be done automatically. So there is nothing the user can do to mess this up, or correct it if need be.

Can you tell me what exactly you are suggesting by checking the calibration (i.e. if no good, what should I do)?

Post by **Harry** » Thu Aug 09, 2012 3:06 pm

To check calibration:

While standing still in the pits, both lineal and lateral acceleration should be zero or nearly zero -> verification step #1 and #2 succeeded. Step #3 can be verified by watching the Acceleration View (Accel tab): there should be a green checkmark (as stated above, you will nearly not be able to get there in public traffic), and when braking, the red arrow should move up straight (not left or right).

- Harry

subawang · Post by **subawang** » Thu Aug 09, 2012 3:21 pm

And what if the calibration is wrong? I thought the software auto-calibrates while on track, so if it is wrong can I fix it? Can I force a manual calibration and have it hold while on track?

Post by **Harry** » Thu Aug 09, 2012 5:10 pm

You simply tell LapTimer to start over automatic calibration (re-calibration):

Press Reset in the Accleration View (press the small wrench first)
While in the Video View, touch the acceleration circle
In v16, you will see a Max Acceleration Gadget in many places, touch it to re-calibrate

- Harry