G-Force / Predictive lap timing / Cornering Speed / Crashing

[sheaven]

Hi,

I am new to Harry’s Laptimer and had the chance of testing it during the last weekend on track. I use an iPhone 4S with Dual XGPS150A (firmware with 5hz).

I like the look and features of the app but had several issues while using it:

G-Force
The accelerometer view typically showed max lateral and longitudinal acceleration of 0.7. This was during hard turns or when braking where I would have expected g-forces of around 1.2-1.3.
When looking at the recorded laps, the g-force maxed out at 0.7 as well.

When using the recalculation method based on GPS data for recorded laps, g-force data maxed out at 1.3 – as expected.

It seems that the g-force meter has a significant time lag. This becomes very obvious when testing as follows:
1. Open accelerometer view
2. Put the iPhone on a table on an upright position
3. Do the calibration of the accelerometer (step 1 and 2 seem sufficient); g-force meter should show zero
4. Lay the iPhone down at the table (g-force should equal 1)

Result: The g-force will increase to 1 but it takes around 5 seconds to increase to 1.

This is obviously a very long time lag for on track driving.

Testing the g-force with other apps, changes happen instantaneous (as expected from the 10hz accelerometer in the iPhone).

I understand from other posts that some lowpass filters are used to smoothen the data which makes sense. But the resulting time lag makes the g-force data pretty meaningless.

Am I doing something wrong or is this the way it is implemented?

Predictive lap timing (lap gap to reference)
I know that this feature is built in and worked during some earlier tests on a public road.

On track, I had not been able to see the predictive lap timing. It always showed the gap versus the best lap in absolute time, i.e. at the start of a new lap, the lap gap to reference was always the actual time for the new lap less the total lap time for the best lap. Example:

Best lap was 2:00 minutes.
New lap actual time at 1. turn: 20 seconds (lap gap to reference: -1:40)
New lap actual time at 2. turn: 60 seconds (lap gap to reference: -1:00)
New lap actual time at finish: 122 seconds (lap gap to reference: +0:02)

In the morning, we were driving sections of the total lap and I did some testing with the app. I deleted these recordings before going the full lap. I assume that the predictive lap timing did not work as the app was somehow confused with earlier data. Can this happen? How can I delete this old data? I deleted all recorded laps…

Cornering Speed
Cornering speed typically was too low, i.e. car speedometer showed 45kph while the app showed cornering speed of 29kph. I will test this again with GPS speed calculated instead of measured. Faster corners seemed to be more in line.

Crashing
The app crashed multiple times, especially when driving to the pit lane (near finish line). Is this a known bug?

[Harry]

Hi,

G-Force is sampled at a low update rate currently. Together with the named low pass, it takes some time until you see peaks. Due to the low sampling rate, peaks may be missed in addition. v16 comes with a standard rate of 30 Hz and can be tweaked up to 100Hz. If the lag is critical depends on the using pattern I believe. I rarely look into the accelerator view while driving, especially in the corner... I look at it just after the corner, so I can learn for the next lap. v16 will come with a "last max" gadget that keeps the peak for some more seconds.

On predictive lap timing - please check with v16 again. I did some changes but have problems to reproduce the situation.

Cornering speed - expect your car to report around 5% too much. GPS is a lot more accurate on speed, but has problems to grab peaks that occur only for a short time (tight corners).

Version 15.0.5 should be pretty stable, so this is not a known bug. You may send me the crash logs using the description here: http://aplus.rs/apple/how-to-find-crash ... ta-and-xp/

- Harry