Data Acquistion (not the sjmarcy kind)
 

Hey guys,

If you were looking for a Data Acquistion system for your car, what sensors and readings would you like to see in it? I am hoping we can go beyond the regularly available stuff like a GPS unit and an accelerometer. Which are relatively quick and dirty. I am talking about things for tuning and functional measurements.

Examples:
Tire and Brake temp.
Wheel Speed
Suspension position/movement
Steering Angle
Throttle position
etc. etc.

Raytek CI or MI series

Piggyback on the ABS sensors. Will require cleaning the optics after every race.

Penny-Giles. Expensive, do not last very long

Celesco string pot on the steering rack

Duh?

Those are ridiculously expensive for IR temp. sensors, whats so special about them?
Have you considered using a hall effect sensor rather than optical for wheel speed?
Never really done suspension displacement, but I know a lot of racing teams want that. I will look into methods of measurement.
Really expensive here too. What are they doing with these sensors that requires a cost higher than most microcontroller boards.


Reverant have undertaken building a data acquisition box? What are some of the issues to consider that may not be obvious at first? are the sensors prone to damage due to the environment they are in? is there a lot of noise and filtering that is required? I am baffled by the expense of sensors like the IR and pot sensor, when its likely the actually IR sensor they are using is no more than $8-$20 to buy.

I'm not going to argue whether they are expensive or not; they are. You get what you pay for with these items. I've tried several cheap-ish IR temp sensors; the problem with most is that they either don't output a 0-5V signal or they don't have the proper ratio (1:1, 1:5, 1:10) to properly read the temperature (ie of the tyre) or they don't have a quick enough refresh rate for accurate measurement (this is especially true for the optical distance measurement method Vs the linear pot).

The Celesco string pots are expensive because a) they are accurate and b) they are durable at 1,000,000 cycles if I recall.

I went with $15 optical sensors for ride height (at 1000ms refresh rate it is useless for shock travel) and the Celesco string pot for steering angle. I have not found a decent solution for tyre temps.

I haven't used analog IR temp sensors. I used one that was I2C/SPI/PWM. In my use I found it to be very accurate, and its about as close to 1:1 as you can get, with some minor deviations dependent on power supply. Not saying its the perfect solution, but it works great if you are willing to invest the time for programming I2C/SPI. The big issue is with mounting, especially on front tires.

The 1M cycles seems a bit low compared to a couple of other brands I have looked at. Unfortunately, the 981 HE I was planning to use has hiked up to $55 a pop compared to the original price of $23. Good thing though I found it on sale closer to the original price. Its good for 5M/10M cycles depending on the form factor used and probably dependent on environmental variables. Accuracy is available in 1% or .5%, which on even a simple 8-bit ADC comes out to a fraction of a * or position, especially if you do multiple reads for an average. I know Hall effect sensors tend to be less happy with temperature than a pot style sensor.

I think a magnet and a linear (rather than rotary we are talking about above) hall effect would make for a pretty good shock travel sensor. Plus, due to its "mechanical" behavior, its going to be as fast as you want to read it, but distance maybe a factor in this.