I've come across the aDyno application for my Android (HTC Hero) phone. Had a play with it last night, and although its a freebie and still in development, I'm fairly impressed. The results seemed very realistic and matched what I'd expect them to be.

It uses the built-in gizmos (G-sensor / accelerometer, GPS & digital compass) for all the measurements after you input details about the vehicle (weight, front surface area, max / min engine revs and speed at the given RPM etc). The phone needs to be fixed securely to get an accurate reading, and it calibrates itself before taking any measurements.

It can do a dyno run, starting from 2000rpm up to max in any given gear, or a timed run from a standing start measuring 0-30, 0-60, 0-100, 1/4 mile etc with an estimate of the wheel horse power.

Lots of fancy graphs and stuff as well, which it saves for later reference.

While it’s all good fun...I was expecting another smartphone app gimmick, so pleasantly surprised at the accuracy of the results last night.

So just how accurate is this likely to be?

Anyone ever compared these sorts of things to the real deal?

Ross

Edited 2 July, 201016 yr by rossm

It says my desk has 20HP.

Guess it depends on the accuracy of your phone's bits and bobs. E.g. the common GPS chips operate at 5Hz, that's +/-0.4 secs error. There's other places error can creep in too.

Good enough for me though

It says my desk has 20HP.

You must have one of those 'hot desks' that seem to be getting introduced! Yeah, my right arm had the power of 8 horses when I first had a play with it. Once calibrated to the vehicle and securely mounted, it become more realistic though.

Guess it depends on the accuracy of your phone's bits and bobs. E.g. the common GPS chips operate at 5Hz, that's +/-0.4 secs error

Yeah, definately - not all Android's were created equal. I believe the Motorola effort has a jumpy G-sensor which causes some problems with Apps like this. I've just realised that you can run the App with GPS disabled so not sure how much the GPS chip plays a part (probaby for the reason above). It seems to be more based on acceleration, and some sort of algorythm calculated with the G-Force.

It also reports back its data sampling rate - typically 40ms for me, so it could mean a 0.04 second tolerance presuming the sampling rate was accurate.

OK. So where would I get my frontal area data etc.....

OK. So where would I get my frontal area data etc.....

Area = Length x Width

I used a measuring tape. Frontal area of 2.8 sq metres on an Octavia.

Some of the other settings are a bit trickier though, like pitch and roll characteristics, drive train losses, rolling resistance and drag coefficient. Who knows how you get that lot!

However, so far - the more accurate I make the car data settings, the less accurate the results are becoming. 536 wheel horse power from my Octavia might explain why foo's desk has 20!

It's a cool idea. If it can get even approximately close it'd be really really cool.

Skoda say 2.19m2 frontal area for a pre-facelift vrs, that helps explain you're fire breathing Vrs Ross Dunno for post FL

It's harder than that though, from doing the sums for aircraft, you need to know the temperature, the current air pressure, there's also the mystical "Cd" which is described in some textbooks as "magic" it's a fudge factor -- we can approximately measure it with a few experiments but not exactly the kind of thing you can DIY too easily (this is the drag coefficient it's asking for).

Remember with your sampling rates there's a +/-0.04ms error on the recorded start, then there's the same error on the recorded stop, so it's +/- 0.08ms, but still seems a bit too good to be true :p