What you’re now working on is what I’ve been slowly working on myself for the past couple of months.
Thankfully, I went with building an SES biped, so I haven’t needed to reverse-engineer like you will with the Nova’s micro.
I can point you to some well-known products, and throw out some ideas, but I’m severely lacking in any hard code, as of yet.
From what I gather, gyros are great for measuring where the biped is, relatively.
I say relatively because they tend to drift.
This is the main reason that you’ll see most people going the IMU route (using both accelerometers and gyros).
An IMU uses an accelerometer to correct the drift of the gyro.
Which accelerometer/gyro is best?
Well, take your pick.
I chose to go with accelerometers from dontronics.com (which haven’t arrived yet because of an SSL problem with the site that Aaron is trying to work out).
I went with those buggers because they had all the necessary caps/resistors onboard, and they broke out the accelerometer into a nice DIP package for prototyping.
Oh, and at 15 bucks a pop, they’re quite cheap.
I haven’t found a good source for gyros, yet, so let me know if something catches your eye.
I’m eyeing sparkfun’s nice gyro breakout boards, but I’m keeping my eyes peeled for now.
You mentioned quantum pills…
In my last revision of my biped, I used .2 inch FSR’s from phidgetusa.com.
I can’t complain in any way about the FSR’s themselves, as they were quite responsive, small, lightweight, low cost, and fairly durable (superglue eats through them, though, so be careful!).
The only problem with them was mounting them.
I tried mounting them between the standard biped foot and another aluminum plate , but the foot was not perfectly level, which threw everything way off.
I’m thinking that the glue that I used to mount them also effected their pressure response.
Last time, I had them arranged in a bridge, since I needed to read 8 with only 4 ADCs.
This time, I’ve got a micro with 11 or so ADCs, so I’ll be able to read each one individually, which will allow for software correction of hardware imperfections.
I’ve been playing around with the idea of having the IMU and the pressure sensors work independantly of each other.
Perhaps a micro on each foot could read the pressure sensors and correct the balance using and x and a y servo in each leg.
Meanwhile, the IMU would be reading the gyro and correcting a waist-bend servo and a waist rotate servo to get it’s x and y correction.
I’m not sure, though, if that’s going to produce a more stable bot or a bot that looks like it’s continuously having a seizure.
That’s where the fun comes in.
