That’s a good idea - I’ll give it some thought…
Pete
Pete,
Have you considered a home made nuclear reactor for power? You could have your bot run for more than 20 years. 
not a bad idea mike, but then again, it would either be tethered to the reactor and be limited to moving places, or pete would really have to beef up his quad 
maybe you could wheel your biped, like make a rollerskating biped with continuos rotation servos. yay 
so many ideas but so little skills. 
Update:
The wheels and motors are in place.
The wheels are 3" R/C plane types.
The two motors are Autotrol gearheads.
The 3rd wheel is a rotating plastic caster, probably made for a roll-around cart of some sort. The wheel is about 1.75" diameter.
Tonight I got the I2C communications working between the main PIC-brain and the sensor board’s PIC. At the moment, every 5 seconds the main PIC tells the sensor PIC to do a ‘sweep’ with the IRPD mounted on a small servo.
Nick: If you’re reading this: When you’re ready, I’ve got a nice set of C routines to do I2C between 2 PICs. It sends little multi-byte ‘packets’, that include a checksum, so that if the comms don’t work for some reason, the protocol will notice so that you can re-try if you want to.
Pete
Well, that sounds swell…
But I don’t have anything to I2C to, yet.
As yet, I’m using the UART to talk to the bluetooth and an I/O to talk to the SSC-32.
Since neither of those buggers have an I2C connector, I’m SOL.
Hold that thought, though, since I don’t doubt that I’ll eventually be whacking on more PICs.
I’m thinking that I’ll eventually have a PIC on every limb and on each foot.
Each limb’s micro will be doing IK and the two foot micros will be reading the pressure sensors.
All of this will have to be sent to the main micro (the one I’m working on now, probably) and then sent to the SSC-32.
I’m thinking that I2C will play a large role there.
The new wheels are fully functional!
Here are pics:
geocities.com/saipan59/robots/wheels1.jpg
geocities.com/saipan59/robots/wheels2.jpg
geocities.com/saipan59/robots/irpd.jpg
In the first pic, the board sticking out just above the tire is the new motor controller. The schematic is here:
geocities.com/saipan59/robots/motor.jpg
The HW worked on the first try. I had a little trouble with the SW, because I had set the IIC address wrong.
The motor board accepts IIC commands for Stop, Forward, Reverse, Pivot Right, and Pivot Left. At the moment, there’s no PWM - the motors run ‘full speed’, which isn’t very fast (a couple of inches per second).
Pete
I forgot to mention that the irpd.jpg pic is a view of the IRPD sensor mounted on an HS-55 servo. The ‘mounting bracket’ is wood (maple).
I used masonite to make the motor/wheel mounts. I know it’s low-tech, but wood is a very practical material: easy to cut/machine, fairly light, and very cheap.
Pete
Wow, wow, and more wows.
By the looks of all those wires it looks very complex. I have a few questions:
Are those regular DC motors? How go you control them with no encoders? Also, do I see two microphones on the top of your bot? They look black. How have the mics worked for you.
The motors are “Autotrol timer motors” - I got them from a surplus dealer a few years ago. They were cheap (about $4 each I think), and they’re all metal, and seem to be good quality.
No encoders needed. When I get around to doing navigation, I’ll do it primarily by ‘landmarks’, rather than ‘dead reckoning’ (using direction and distance). This means I won’t need to know exactly how far I move in a certain direction.
In this respect, the bot is ‘female’… I read somewhere that women tend to navitgate more by landmarks that they remember along the way; men are better at keeping track of distance and direction.
Yes, those are condenser mics on the sensor board. I haven’t written the SW to do much with them yet. Eventually they will be mounted several inches (or more) apart from each other, so that he (she) can determine which way a sound came from.
Pete
Very nice. The mics look like they have some kind of black shroud on them, or are they all one piece?
Yes, the mics have a black rubber ring around them. They’re the kind you would find inside a cordless phone, or any other consumer gadget that would use a mic. They’re similar to the Radio Shack mics that you use, except that they are a 2-wire type instead of a 3-wire.
Pete
Very nice work, Pete!
Your woodworking is spectacular.
I can’t wait to see your spider floating around on it’s wheels with it’s legs waving around above the floor.
Made some good progress today. I had been fighting an I2C problem, but all is working well now. The main PIC-brain board is now talking to the sensor board and the motor control board. The sensor board scans with its IRPD sensor, and returns a ‘map’ of what it sees back to the main board.
Here is an output from the sensor sweep, converted to ASCII for us humans to look at. Each row of 's represents a value returned from the IRPD at that point in the sweep. There are 40 measurements.
The longer the '**’ is, the closer the object is.
The blob in the middle of the sweep is my hand, about 1.5 feet away.
The second big blob is my body.
**
**
*
**
**
**
Meanwhile, the motor board accepts commands for stop, forward, reverse, pivot right, and pivot left.
Pete
pretty sweet Pete
you have any videos?
Pete, that is awsome.
I have a question regarding the mics you use. Your mics are 2 wire where mine are 3 wire. Lets say that I would want to use 2 wire electret mics in my design, would I wire it up like this: ?
+5v
|
|
|---|-------------o To Op Amp
| |
|---|-------------o GNDThanks,
With a 2-wire mic, you must have an R and C. I’ll try to make a picture, where ‘M’ is the mic, ‘G’ is ground, and ‘5’ is 5V:
5
|
R
|
M---+---C-----
|
GIn case the picture doesn’t work, the text explanation is:
The mic is connected to +5 via a resistor (about 1K-2K).
That same connection on the mic is connected to the + side of a cap (about 5uF-10uF). The other side of the cap goes to the op-amp (typically thru a resistor first).
So, the mic receives power through the same wire that is its output.
Pete
Thanks Pete,
I left out the Cap and Resistor to make my ASCII art easier.
I figured that was the way to do it.
It is helpful to get a data sheet for the mic element and find out what it’s current draw and recommended bias voltage is to pick an optimum resistor value. You don’t HAVE to, but you may find you can optimize the response of the mic quite a bit by doing so. I have an automotive voice recognition system that uses several different microphone types (user dependent) and we program one of 4 different bias resistors from an 8 volt source depending on which mic is used. The resistors range in value from 10K to 680 ohms.
Another thing to remember is that your power supply is now AC coupled into your amplifier just like the mic is. Not sure how much it’ll matter on a little robot but in a car where supplies are derrived from the vehicle and/or coach battery this is an important thing to keep in mind (especially for voice recognition, heh.)