The accelerometers

[Validius]

1549's bot uses sort of hybrid skid steering. We have the regular wheelchair wheels on the front and some big solid rubber tires on the back. needless to say, our center of rotation is constantly changing, even in the middle of a turn.

My question is this: If i mounted two accelerometers on the bot, one 1/4 bot length from the front and another 1/4 bot length from the back how hard would it be to mathematically figure out how far i turned?

Also, do i even need to have two censors?

:edit for spelling:

[Mark McLeod]

A gyroscope is more appropriate. The accelerometers won't give you an angular rate change which is what you are looking for on turns.

[Validius]

Originally Posted by Mark McLeod

A gyroscope is more appropriate. The accelerometers won't give you an angular rate change which is what you are looking for on turns.

Then what are the accelerometers good for?

[Mark McLeod]

They are good as tilt sensors, so you know if you are starting to fall over, or if you are on the stairs from last year or the ramp from the year before (and so on).
They can also be used to detect when you run into something.

[Validius]

what kinda output do they send?

[KenWittlief]

wait a minute, which accelerometers are we talking about here?

the ones that use to be provided by FIRST were angular accelerometers, or in other words, solid state gyroscopes

they do tell you the rate at which your platform is turning in degrees per second (change of angle is acceleration)

if that is what you are referring to, you only need one, and it doenst matter where you put it, because the entire robot rotates at the same rate.

[Joe Ross]

Originally Posted by KenWittlief

the ones that use to be provided by FIRST were angular accelerometers, or in other words, solid state gyroscopes

Look what you get for taking a year off

This year FIRST provided a linear accelerometer from TI, used in automotive applications. It senses +/- 1.5 G. They didn't provide any type of angular rate sensors.

[KenWittlief]

in that case, they would not be able to measure rotation.

search the forum for gyros and you will find lots of threads on commercially available solid state gyros that do measure angular rate of turn.

there were many threads on them last year. Its too late to design them into your machine this year, unless you make the mods at the event you are attending - It would be tough to write a PID loop on the fly

if you have a practice bot you could experiment at home, but anything you build has to be done at the events at this point in the season.

[Validius]

so the ones this year are solid state gyroscopes?

How do i hook them into the bot?

[Mark McLeod]

Originally Posted by Validius

so the ones this year are solid state gyroscopes?

How do i hook them into the bot?

What came in the KOP this year are accelerometers not gyroscopes.
The TI accelerometers connect to the analog inputs, but be careful with the wires, because the colors are mixed up from what you'd expect.
-Black TI accelerometer wire goes to power, the RC analog input center pin
-Red accelerometer wire goes to ground, the RC analog input outside pin
-White accelerometer wire goes to signal, RC inside pin

See the spec sheet FIRST provided http://www2.usfirst.org/2005comp/Spe...elerometer.pdf

The readings will be:
~512 if it is oriented so it points to the side (when it starts to tilt, it starts to sense gravitational pull)
~700 if it points up (& measures gravity)
~300 if it points down (also gravity)

For more info on accelerometers see: http://www2.usfirst.org/2005comp/Manuals/Acceler1.pdf