Mechanics and Construction > Mechanics and Construction

Driving straight for 4 meters

<< < (2/5) > >>

dunk:
i'm not convinced a gyro would be any more accurate than encoders (or any other method of ensuring the wheels were moving at the same speed such as dual differential drive).
gyros all have some element of drift.
which method would be more reliable depends on how smooth a surface you intend to drive on.

while a magnetic compass doesn't drift in the same way as the other suggested methods they are susceptible to metal in the surrounding area (or your friends teasing your bot with metal keys).

so in summary, it depends on the surface your bot has to travel on.
if it's a perfectly smooth playing field and you have the mechanical ability then i'd go with dual differential drive.
in 2nd place i'd put DC motors with encoders if you want to simplify the build.
(i've never played with stepper motors. anyone got any experience using them as main drive motors?)

if on the other hand you have to cross unpredictable terrain that might cause wheels to slip then i'd incorporate 2 or more of the above methods.
maybe a magnetic compass combined with a rate gyro. the gyro will only drift slowly so use the gyro when the bot is moving and may be passing metal objects but providing the 2 sensors agree, use the compass to calibrate the gyro.
or better still, combine all 3.
o, and add gps. i've always wanted an excuse to put gps on a bot.
and a trained monkey. monkeys can come in handy.

dunk.

ed1380:
what if all he needs is to put a resistor on a motor. unequall motors strength

trigger:

--- Quote from: ed1380 on February 20, 2007, 06:59:58 PM ---what if all he needs is to put a resistor on a motor. unequall motors strength

--- End quote ---

Won't work because that's still open-loop control. You need a feedback loop or else the slightest error will cause problems.

JonHylands:
A gyro isn't going to drift very far in 4 meters...

- Jon

Admin:
sounds like i need to write a tutorial about dead-reckoning someday soon, i reckon . . .

as dunk said, it very much depends on the terrain. but if you got perfect terrain, id vote for the gyro for simplicity . . .

the drift on the gyro should be a measurable constant, for a given 4m course. if you can test beforehand on that exact 4m course, then you can tweak to eliminate the drift.

Navigation

[0] Message Index

[#] Next page

[*] Previous page