Sensor response time...

[randomperson]

Hmm.. does anyone know the response time for the optical sensors? I know that it says it in the paper for them.. but how long does it actually take before the controller will tell the program that the sensor is on/off? I guess a simplified version of this question would be what kind of sampling rate are we getting with the sensors?

[Dave Flowerday]

The best sampling rate that you can have (if the sensor is hooked directly to the RC) is only about 40Hz, due to the extreme slowness of the BASIC Stamp.

The actual sensor can detect changes much, much faster than this (1000Hz according to it's datasheet).

how long does it actually take before the controller will tell the program that the sensor is on/off?

Worst case the delay between the actual transition of the light sensor output and the time at which your code knows about it is 25 milliseconds. Also, if whatever you're doing with the light sensor causes a pulse that is less than 25ms long, you run the risk of the RC never knowing about the pulse in the first place.

Let me know if this didn't answer your question, and I'll try to explain it better...

[n[ate]vw]

We've been playing with automatic control using the light sensors and their response time seems to be negligible. Like Dave said, the delay is going to come from the code, not the sensors. If a light sensor is physically triggered at the time the loop is running, it's safe to assume that your program will know it.

[rbayer]

You get the same sampling rate as with everything else connected to the RC: 26ms minimum. If your code is excessive complex, this could be as high as 40ms, but I haven't seen much past that.

[Lloyd Burns]

The "response" seems to depend on setting of the sensitivy, which in turn depends on what is needed to discriminate between the red carpet and the white line.

On a red broadloom, I notice 'considerable' (and considerably varying) time between the light hitting the calculator-tape curve, and the robot's reaction to it.