I have the Bluetooth stuff for WALTER now. I just have to figure out if I can use HSERIN/HSEROUT for it now. I don’t know what pins HSERIN/HSEROUT uses on the Basic Atom and it doesn’t seem to be documented anywhere. Until I know what pins HSERIN/HSEROUT uses, I will use standard serout on P12 and P13.
I have added the configuration for the Bluetooth module in WALTER’s software. Now I need to see what sort of initialization it requires, according to the datasheet, and add that to WALTER’s software.
I may not even be able to add the initialization and I/O routines for Bluetooth, due to lack of program memory. WALTER is down to just 188 bytes of program memory now.
I’ve been tinkering with my LPC-2148 ARM7 Proto Board. I have gotten FreeRTOS to build successfully using the LPC-2138 port. The LPC-2138 is very similar to the LPC-2148.
Now I want to get uIP (TCP/IP) running as a task under FreeRTOS. I've already gotten uIP to build separately. If I can get this to work on my proto board, WALTER will be able to be a fully connected robot! :smiley: I'm planning to add an ethernet interface to my proto board also.
I think I am on the road to being able to start developing th C based software that will allow a real time behavior oriented architechture.
8-Dale
I just had a brainstorm the other day! I have an Olimex LPC2148 Proto Board, which has a small prototyping area on it. This area is just large enough to add a 40 pin socket and some headers.
I want to add a Hammer to the prototype area, with some headers. The Hammer would connect to the LPC2148 using I2C. I2C #0 is brought out to headers near the prototype area, and the LPC2148 has two I2C buses. The Hammer would naturally be the I2C Master, with the LPC2148 running FreeRTOS/uIP as the slave.
I think this would be the PERFECT new “brain” for W.A.L.T.E.R.! I will have the much desired complete ARM9, full Linux (not uCLinux) capable controller (Hammer) with a nice ARM7 (LPC2148) slave MCU running a nice little multitasking kernel.
I’m already starting to learn how to use FreeRTOS, and have been starting to flesh out the routines I will need to run a true subsumption behavioral based system with it. I am in the process of searching out and grabbing up as much of the needed code (Open Source, of course!) I can to use in this project. My FreeRTOS v4.7.0/uIP builds clean using Crossworks for ARM (Linux version) and works on my LPC2148 board now. I just have to add more tasks, initialization for I2C and other hardware, etc.
You are seeing Alan’s setup, not mine. Alan is making the decks for me.
I am not going to have the new decks before Christmas, but they will be a nice way to start the new year. I have been wanting to get these made for many months! I would rather wait and see everything get done right than have a hurry job done and have things not right. I know Alan will do the best job possible.
I’ve been thinking a lot about the new “brain” for W.A.L.T.E.R. While I would really like to put a Hammer on and connect that to my Olimex LPC2148 Proto Board, I can’t do that right now. I will probably not get the Hammer until March or April at the earliest.
So, I am considering going a little different route. I might add two dsPIC30F4012 (28 pin DIP) and an MCP23017 (28 pin DIP) I2C I/O expander to the proto board. This would give me more processing power, two quadrature encoder modules for the motor encoders (one on each dsPIC), and more ADCs which I need for Sharp IR rangers and my 3-axis accelerometer.
The LPC2148 would be the master controller and communicate with the MCP23017 and dsPICs using I2C. I can run a slightly cut back version of Pete’s (sapian59) dsPIC I2C slave code on the 4012’s, and have already got this done. I just have to tweak the hardware initialization for the dsPICs to what I need.
I also need to get the PIC I2C Slave code done, so we can use regular PICs as either I2C Master OR Slave, selectable via jumpers. The select code already works. I just need the I2C slave code to complete it. I am also adding this selection code to Pete’s dsPIC code.
I also want to add socket headers for my Bluetooth DIP module. I believe there will be just enough space on the proto board’s prototyping area for three 28 pin DIPs and the bluetooth module. I will probably have to put the necessary three pin I/O headers and other I/O connectors on another board.
I am in the process of rebuilding W.A.L.T.E.R. with the new decks now. I already have him mostly disassembled. Time will tell if I get him back together right, but I am thinking of making some changes the new decks will allow.
You can make extra 1" holes in your decks if you buy a Greenlee punch. I think I see a layout of 4 holes in one of the two utility decks. You need a better camera!
Yes, one of the top deck has the four large holes already. That’s what I used as a template to draw the circles on the utility decks.
I have to figure out how to fix the motor mount holes also. They are too far apart, but that could be an error in my CAD model. I have not checked that yet. This is the first time I have had something made from my CAD drawings, so I expected some glitches.
Yes, I know. One of these days I will surprise everyone and get a new camera. I want a Nikon P3.
Ah, OK, I will have to correct that in my CAD model then. If that is the only error I made, I think I did pretty well. All the other holes look dead on so far. I think I will just fill in the existing motor mount holes and redrill them. It looks like the holes are centered properly, just too far apart. Now that the MMT-03 is available, I might just get a pair and put in some SES holes between the current motor mount holes. I think there is plenty of room. I can use any SES bracket for a hole template.
Yes, this material is very easy to work with a dremel tool on. That’s why I decided to stay with this rather than switch. I’d still like to get a laser cut set of lexan decks for W.A.L.T.E.R., and I think I have found a local guy who will do it inexpensively. He has one of those monster automated milling setups, and other cool machines. He’s offered to work with PARTS (local robotics club) at cost and make the time for stuff. PARTS is undertaking a RoboMagellan project.
I don’t think I know anyone with a printer who could print a 1:1 drawing for me.
W.A.L.T.E.R. is in pieces all over my desk right now, and I have to find one of those little hex tools so I can remove the hubs from the motor shafts. I have to completely disassemble each of the motor assemblies in order to be able to remount the motor mounts on the new wheeled deck.
Once I get all this completed, W.A.L.T.E.R. is going to be one cool robot (and much larger).
I did a test fit of the wheeled deck with the top deck, and it looks like we have perfect alignment. I even test fit one of the motor assemblies where they will be mounted, and the entire wheel fits completely within the radius of the body, just like it is supposed to fit. I actually designed the wheel wells to accommodate up to 6" diameter, 3" wide wheels. I also test fit a servo into the servo cutout and it fits perfectly! The decks look really nice put together. I think the 4-40 hardware will work fine and I am going to stick with it.
I just have to find and fix the motor mount hole problem in my 3D model now and find that durn little allen wrench hex socket tool so I can finish disassembling the motor assemblies.
I am really excited about getting W.A.L.T.E.R. rebuilt using the new decks!
So far, I am really happy with how this has worked out. It’s my first design created from my own models. It’s also the first time working with somebody to make the design real. I failed to communicate to you that the four larger holes in the decks needed to be in all three decks. My fault there.
The motors are GHM-04’s and I have the QME-01 encoders for them also. I need to get new casters for front and back, and I would like to use free wheeling omni-wheels. I have an idea for retractable and self leveling casters.
I will see what I can do about getting better pictures.
I may just keep W.A.L.T.E.R. as a three deck robot and use one of the decks without large holes as the top deck. Having the four small SES holes around the hex spacer holes is going to work out very well. I am going to have to stock up on hex standoffs and see if I can find some longer than 1 7/8" (2 1/4" would be perfect). I do NOT want to use any more nylon standoffs with metal standoffs.
Good. I need 2 1/2" clearance between the second and third decks. I just have enough clearance using 1 7/8" spacers between the bottom and second decks. The second deck just does clear the tilt servo for the front arm with the pan servo mounted in the bottom deck.
I tried looking at Jameco, but their search engine really sucks. I’l try Digi-Key.
I’ve completed the 3D model of the new four deck body for W.A.L.T.E.R. I’ll need at least 24 standoffs that are at least 2 1/2" long, 4-40 threaded hex aluminum. Now I can start adding stuff like the motors and front mounted pan/tilt (needs a 3rd DOF).
Unfortunately, I let a bit of magic smoke out of the old style ABB that holds the Basic Atom that is W.A.L.T.E.R.'s current “brain.” The Basic Atom survived though, but there seems to be a problem in the power circuitry of the ABB and I can’t get anything on the VS terminals now, which powers the SSC-32.
I’m going to order two new Bot Board II’s. I’ll go ahead and get an Atom PRO to replace the overworked Basic Atom at the same time.
I did find what I need in 4-40 hex standoffs at Digi-Key. They have 1 1/2" F/F, 1/2" M/F, and 3/4" M/F aluminum hex standoffs. To get anything longer than 1 1/2", I would have to move up to 6-32 hardware, which is a possibility at some point.
I finally got everything mounted on W.A.L.T.E.R. - sensors (3 Sharp IR, a PING, and an IRPD), and the motors are mounted. The new 11 1/2" diameter decks definitely make W.A.L.T.E.R. bigger.
I need more standoffs and hardware to attach the third and fourth decks. I also don’t have the clearance I would like between the bottom and second decks using 1 7/8" standoffs. This would be a good time to convert to using 6-32 hardware and standoffs if I am going to switch. I think 2 1/2" between decks would be perfect.
Until I know what pins HSERIN/HSEROUT uses, I will use standard serout on P12 and P13.
I just have to add more tasks, initialization for I2C and other hardware, etc.
building W.A.L.T.E.R