How does PerfExpert works ?
PerfExpert relies on the accelerometer of your phone to measure the actual acceleration of your car. This requires that you always leave from complete halt (0Mph), which allows the application to compute very accurately the instantaneous speed and distance. PerfExpert can compute not only car performances in acceleration times (0-60Mph, 1/4mile...) but also data such as power and torque of your engine thanks to the information you previously provided in a sheet (total vehicle weight, gear box ratios, final drive ratios...).
Resulting from long research and development, techniques used by PerfExpert to handle acceleration and compute engine power give reliable and realistic results.
What do I need for reliable measures ?
In order to perform realistic measures, you must give great importance to the measure conditions and the measure procedure in itself. By following these instructions you can get excellent results :
- Fill with attention all the parameters of your car sheet (search info on the net, weight your vehicle...),
- Choose a free track, flat, horizontal and as straight as possible,
- Although secure the phone in the car so that it is perfectly rigid with the car and in its direction of travel,
- Always leave from complete halt, if possible take a mark from the same point of the track,
- Ensure the absence (or near absence) of wind, that your tires are hot and properly inflated (manufacturer pressure),
- Follow the instructions in the Guide for the measure procedure!
How does PerfExpert know the engine rev, as it's not connected to the car ?
From the information you entered in the car sheet, PerfExpert know the overall ratio of your car transmission. So from the car speed, PerfEpxert can compute back the engine rev.
What results Accuracy can I expect from PerfExpert ?
This accuracy not only depends on PerfExpert or the smartphone used, but also on the rigor of your measuring procedure. It also depends on the quality of the fixation of your phone in the car. Following a perfect measuring procedure, using a good phone (with a good accelerometer), results relative accuracy can reach 0.5% in Timed Run mode and 2% in Dyno mode.
Do I need a good GPS signal or my measurements ?
No, PerfExpert does not use your phone's GPS, it is inappropriate for reliable detailed dyno measures.
Why does PerfExpert is accelerometer-based and not GPS-based ?
We choosed to use accelerometers rather than GPS receivers for reliable performance measurements. In fact after studying the question by considering the Dyno mode (PerfExpert speciality), here is the main reason that guided our choice : data retrieved by accelerometers (G-sensors) allows a more accurate results calculation.
Some explanations are necessary :
Basic principle : to determine the power curve of a car realistically, you must know precisely and as often as possible the car acceleration. Therefore an application providing power results must be based on a data acquisition system providing accurate data, and at a high frequency.
Acceleration is by definition an information provided directly by accelerometers. Conversely, a GPS receiver provides a position, which must be differentiated to get a speed, which is differentiated again to finally get an acceleration. With GPS's two operations (differentiations) are necessary to get an information that accelerometers give you directly. The differentiation performed here is a mathematical operation that consists in dividing the variation of a quantity by the one of time. The raw acceleration provided by an accelerometer and the raw position provided by a GPS must both be filtered to eliminate noise. Note that the nature of this filtering will vary totally according to whether we want to filter data from a GPS or an accelerometer, but the most important thing to know here is that a differentiation will amplify any noise. As we seen before, in the case of a given position from the GPS, we must differentiate two times... So you can already see here that technically the GPS is disadvantaged.
About accuracy and frequency, we performed measurements on the best Android and iOS phones of the moment and this is what we have noticed :
Modern smartphones electro-mechanical accelerometers have an accuracy from about 0.001m/s² to about 0.2m/s² when the best smartphone GPS reach an accuracy of one meter in the best conditions.
Accelerometers are between 10Hz and 100Hz, when GPS receivers are running from 1Hz to 5Hz.
In conclusion, accuracy and frequency of these two systems are in favor of the accelerometers, causing the same effects on dyno measure accuracy.
About Timed Run measurements : Despite its low acquisition rate, GPS's can be particularly reliable in "distance/time" timed run measurements like 1/8mile and 1/4mile, since in this case, a GPS gives directly the position, when accelerometer-based systems must retrieve it by double integration. And it can be fairly accurate for measuring "speed/time" timed runs, like 0-60Mph, since in this case, each system is "half and half" : to get speed GPS-based systems differentiate once the position when accelerometer-based ones integrate once the acceleration.