A first layout of the controller for our summer project. I’ve been saying it’s easy to build a Segway for years and the team finally called my bluff, so we’re gonna do it. The controller will handle two PWM outputs (left and right motors), plus a few analog and digital channels. It is wirelessly programmable and debug-able. (In theory it could operate under remote control but I don’t know how good of an idea that is for a Segway.) And the strange-looking connector on the bottom right is for an ADIS16350 inertial sensor that I just happened to come across…
P.S. MechE’s shouldn’t do PCB design: If any of the EE-types here spot something glaringly wrong with the design, please let me know!
good luck! also you could probably use a gyro mounted to measure the z-axis to keep it stable. may i also suggesy building a Legway (lego segway - google it, theres a few links) to prototype to make sure you have a lot of the bugs worked out before you build it full-scale. theres nothing worse than making something to have it not work…
Are you using a ground plane with this board? There is no trace for gnd visible. If so, make sure you select the correct options when you send it to be made. Also, you may want to trim the board down and reduce some of that text, boards get pretty expensive pretty fast as they get bigger.
Oops: Yes, there is a ground plane. I turned it off so I could see what I was doing, but it fills everywhere inside the green outline and all the pads that look like they aren’t connected to anything are on that plane.
As for the cost, www.33each.com is my new favorite fab place. $33 boards up to 60 in^2 with no minimum quantity for students. Plus, they send you free popcorn and a Domino’s coupon.
Anyone know if the Schottky diode in series (D1) approach is okay for reverse polarity protection for something like this? I’m looking to protect the sensor - the PIC and radio are relatively expendable. Are there any other good methods of making sure the sensor (bottom right connectors) doesn’t get toasted?
Works like a charm. Um, not the Segway part yet, just the PCB. I can grab the 3-axis gyro and accel measurements off the sensor at about 300 samples per second and drive two PWMs, plus communicate everything to a PC wirelessly.
Hooking up the $275 sensor for the first time was scary. No matter how many times you check the pin diagram, it’s still an adventure.
