MIT/FIRST DIY Segway

[chris31]

Congratz to the guys who were working on the DIY segway as a summer project. You are on Engadget and now the front page of digg with 282 diggs.

http://www.engadget.com/2007/08/21/c...po-diy-segway/

http://digg.com/mods/Clever_students...apo_DIY_Segway

[ZZII 527]

Thanks. And thanks to these guys:

http://www.waylandfirst.com/albums/a...rs/img-84.html
http://www.waylandfirst.com/albums/a...rs/img-81.html

(Ignore the "we'll own them" caption... They were feeling a bit competitive, I guess, being down in Atlanta and all.)

Whatever you told our HS students down in Atlanta convinced them that they should convince me to let them make a Segway. So this is your project, and for that matter all of CD's and FIRST's project, too!

-Shane

[Akash Rastogi]

What was the hardest part about building your segway? Was it more of a programming project or mechanical? I figured that the harder part would be programming the controls and functions of the gyros. I was just curious. Awesome job btw.

[dtengineering]

I can see some awards being handed out to the first team to show up with a seg-bot at a Regional. Not to take anything away from this awesome project, but from watching the video, it really makes me appreciate just how smooth the real Segway really is. It will be interesting to see how close the DIY one gets... is it just a matter of tweaking the control loops?

Jason

[Salik Syed]

Pretty cool, although I saw the same thing a few years back... did you totally design from scratch, and did you use only first parts?

[Qbranch]

I'm amazed they got enough power at a reasonable current with 12volts... I think the segway runs like 48 or 96 volts to cut down on current/wiring weight.

Anyone know what motors they used? PWM drives? (Being MIT, they probably put their own FET bridges together)

As a note... i'm not sure you could build this with the current FRC control system... the 120Hz max chop rate on the victors would probably make your servoing relatively rough (Segway probably has an update/chop rate of several KHz)

Amazing project... now if I could just pitch having to build one to my boss so I had the funds....

-q

[chris31]

Originally Posted by Qbranch

I'm amazed they got enough power at a reasonable current with 12volts... I think the segway runs like 48 or 96 volts to cut down on current/wiring weight.

Anyone know what motors they used? PWM drives? (Being MIT, they probably put their own FET bridges together)

As a note... i'm not sure you could build this with the current FRC control system... the 120Hz max chop rate on the victors would probably make your servoing relatively rough (Segway probably has an update/chop rate of several KHz)

Amazing project... now if I could just pitch having to build one to my boss so I had the funds....

-q

They used the small CIMs with the Banebots gearbox and I believe they used Victors. I think they have all of the information on ther epage I just have to go back and read it all.

[ZZII 527]

It is built mostly from FIRST parts, and with very little interference from the MIT contingent of the team. It uses the standard KOP Victor 884s. 120 Hz is fast enough. I thought I read somewhere that the commerical Segway controller checks its sensors at 100 Hz? The bigger problem with the 884s is the neutral deadband, but we have a software fix for that.

We just posted a big folder of more detailed technical documentation, which I'm sure you'll all appreciate much more than the general public, having knowledge of the components:
http://web.mit.edu/first/segway/segspecs.zip (13MB!)

[Heavy Metal]

awesome pretty good
i dont think its fair to point out theres looks like a stripped version of segway
they built something top-notch engineers build and upgrade
its pretty for just MIT

[Salik Syed]

Originally Posted by Qbranch

I'm amazed they got enough power at a reasonable current with 12volts... I think the segway runs like 48 or 96 volts to cut down on current/wiring weight.

Anyone know what motors they used? PWM drives? (Being MIT, they probably put their own FET bridges together)

As a note... i'm not sure you could build this with the current FRC control system... the 120Hz max chop rate on the victors would probably make your servoing relatively rough (Segway probably has an update/chop rate of several KHz)

Amazing project... now if I could just pitch having to build one to my boss so I had the funds....

-q

I think it's very doable, as far as power needs are concerned I don't think they should be much more than your average FIRST robot, especially considering that most people don't ram into things or push on a Segway. (Albeit Segways need to go faster, and probably weigh ~40-50 lbs more too... assuming we are dealing with lightweight nerds like myself)

120hz is actually a relatively fast rate for this type of application, a few kilohertz seems like overkill, even for the REAL Segway


Build two and tell your boss he gets to keep one!

[joeweber]

Looks great, you should be proud of what you have acomplished. Now that you have a base idea that works to build from you can do almost anything with it. I bet it cost you alot less than the real thing. Keep building.