BalBot

Yeah I have been studying them too. Now you have me envious. I like their design, it's like a "why didn't I think of that" kind of thing. My big balancing robot is a lot more complicated mechanically.

There is a caveat though. It uses a couple of IR sensors as tilt sensors. Thus it won't be able to go up ramps. I am really surprised they didn't use an accelerometer for the tilt angle measurement.

The IR thing is no big deal. I think of this as an entry point.

Really neat molded parts.

I have decided to do a write up in my column on this one.

Cool. Very tall platform. Do you know if these balancing bots work better with a taller platform or a shorter one? One imagines there is an optimal height. A taller platform might be more stable, right *up until* some critical angle of tip-over, after which it will be impossible to catch it when falling. Likewise, if it's laying on its side, a taller platform might not be rightable, whereas a short platform will pop right up. ?????

BTW, what h-bridge chips are they using on the motor controller?

- dan michaels

A taller design is easier. The difference between balencing a pencil or a broomstick. Likewise, more mass up above is more stable.

I am not a big fan of C, but I am going to plung into this one.

The bridge is some sort of allegro part, 3959 .

This thing is really really cool. Been playing with it all morning.

Mike

Where's your column ?

I do the personal robotics column in Nuts N Volts. Always looking for new talent. Looks like I found it!!!

Cool! I read that a good bit.

I've thought about writing something for Nuts N Volts and submitting it on the ER1 I got a couple of years ago. I wrote some software for it that got me a trip to the CES in Vegas. I really need to *finish* the software sometime, though! Maybe I could do that as a prelude to an article sometime ?

A motors/wheels/frame option would interest me!

Can someone give a brief rundown on implementing back-EMF sensing, and advantages/disadvantages compared to encoding?

Thanks,

Ian.

Back EMF is good for sensing the velocity a motor is spinning at. Not as good for sensing the position a motor is in, since you have to integrate the signal to do so.

Back EMF requires an A/D converter, and a current sense resistor. This is a common feature on microcontrollers these days.

Position encoding is just that, position based. You have to derrive velocity from it, which is easier. Most microcontrollers do not provide decoding specific modules on them.

Mike

Write it up and submit it to SERVO Magazine. The proper form would be an explanation of what you did, why you did it, and complete disclosure (instructions, code, etc.) so that other could do it, too.

Dan

jc-Atl wrote: