SparkFun Forums 

Where electronics enthusiasts find answers.

Have a good idea for a new product for SFE or Olimex? Let us know!
By Tre
#5515
I don't know how feasible this is, and thats why I am here to ask questions of the experts! I am thinking about a new project that would monitor a car's acceleration, speed, and other parameters. Essentially, I would like to a make a cheap affordable device that could measure a cars 1/4 mile time, 0-60 time, breaking distance, etc.... I was thinking that this could all be done with an accelerometer, RTC, and a GPS module. However, I have some serious concerns about the GPS module. Will it accurately and quickly be able to gauge 60'? Most cars that I would run this with would have times 3.7-4.5 seconds so the GPS would have to be able to refresh fairly quickly.

Do any of the experts have any thoughts on this?
Thanks!
By gambrose
#5553
I'm no expert but i have seen modules that do this and they use GPS and accelerometers.
they can be very accurate as they are not affected by tire distortion which effects typical car speed sensors.

you would not require a real time clock as the GPS module has one built in that is configured by the GPS to accuracy of 50 nanosecs.

GPS modules tend to out put every second although you can get them to out put more regally you won't get the resolution you require. but that is what the accelerometer is for.

I've got a 2 axis accelerometer as i thought that would be enough but it never occurred to me that if it is tilted then it will start registering gravity. :oops:
By Tre
#5559
Yeah, I was going to use 2 acceleromoters mounted orthogonally to each other. The biggest challenge in using the extra accelerometer is calculating the acceleration in the x and y directions. Mathematically speaking, you want to go from R3 <x,y,z> and translate the vector to R3 with <x',y', 9.8>.

Most GPS modules that I have seen only update at 1Hz (some at 5Hz). This is far to slow to use for that kind of accuracy. I would be better of trying to integrate the acceleration on a discrete time basis. Right now I am thinking about ditching the GPS module entirely.
By gambrose
#5581
yes you get your accurate resolution from integrating your acceleration but you still need the GPS to keep the integration in check.

if you integrate a small error it becomes a large one if you do this over a long time i becomes a huge error pretty soon you car is telling you that you are approaching the seed of light while you are still sitting in the car park.

so don't ditch the GPS.