Journal98 feb
From Simreal
February 1998
Tue, 3 Feb 98
It is amazing how quickly the days fly past. I've mentioned this before, but it still amazes me.
I've been researching serial communications through Windows '95... simple in theory, but with enough new features that it is challenging to learn. I also spent some time exploring ActiveX OCX controls, but for my needs, it seems that DLL's are better. And I've also been re-packaging and cleaning up the Robot Brain software.
The purpose of all these software machinations is... to control a robot from the desktop PC. Can't reveal much about the actual connection between the PC and the Robot, because that is proprietary to the company I am doing it for.
It looks like I will have a running, rolling robot to test my Brain technology on this next week though. And if that works, I will re-orient the Brain to work with the Boris Legs. I think I want a full-blown PC running the show on Boris, with MicroControllers as slaves (much like I have on this for-pay project).
One thing that I am doing now is working with the PIC16C74a MCU. I know, I spent a couple of hundred dollars ramping up on the HC11, etc, etc... but my client uses PIC. So now I have the MicroBench software, an PicStart Plus EPROM burner, an UltraViolet EPROM eraser, and so on. Once I actually get the PIC to run on my proto-board, it will be complete.
So the question is... do I switch to PIC? Or stay with HC11? PICs are more expensive than HC11 ($20 to $10). PICs have more I/O. PICS are faster (20Mhz to, what? 2 Mhz? 4Mhz? I forget). Many of the down-sides (EPROM vs. EEPROM) are now resolved due to my contract.
So maybe I will convert to PIC. Damn. Anyone want to buy some HC11 and BotBoard2 boards?
Thu, 12 Feb 98
I'm still *way* too busy. Days, I work on paying projects. Nights (and, heck, lets be honest; and some Days) I work on Robot-related projects.
This last week or so... hmm, today is Thursday, so two weeks now... I have had my hands (don't ask me how, <heh> <heh>, it helps having friends) on a pair of "Radiometrix UHF Radio Data Transceiver BiM Development Kits". Interesting. I thought at first that sending and receiving RS232 through the R/F would be easy... after all, my computer has RS232. My PIC has RS232. What could be simpler?
ARGH! Many things are simpler. It seems that you have to find *just* the right magic sequence of bytes to get the data stream to synchronize and fall into place. It took three days to find the first set, from PC to PIC. Did that work on the way back? Of course not. It only took one evening to find the second. Very scary stuff.
Anyway, I now have bi-directional communication through radio... which, as we all know, is vital to any RobotWars human-directed robot.
Right now, I am interfacing my Brain software with the code that talks R/F. Most of it is still synchronous -- that is, it waits for the send or receive to complete before going on with life. I can live with that for now. With a basic message length of about 16 bytes (including sync bytes), at an effective 4800 baud rate (don't ask), that isn't too bad. I can double the speed later, when I feel daring, and willing to risk the wrath of the syncronization bytes.
Hey, I found a buyer for a pair of BotBoard2 and HC11 setups! Hope he likes them...
This weekend, I intend to have my non-Boris robot spinning its wheels in response to Brain behaviors. That will be a major breakthrough for me -- my desktop computer directing the mechanical systems remotely. Darn handy. It is getting closer to time to re-start the Boris mechanicals...
Tue, 17 Feb 98
I spend Friday and Saturday sick, so my productive weekend was greatly reduced. When I hooked the PIC up to the robot platform, I ran into the classical problem of motor transients reseting the MCU. I added a few more capacitors, and I think it will be okay. I also ran into the annoying problem of the PWM output from the MCU, at its slowest, is about three times too fast for the PWM motor controllers. I have the "official hardware guy" working on the problem (since I'm just the software guy on this project, I don't have to); hopefully, in the next day or two, I will have a circuit that can take the high-speed PWM from the PIC, smooth it to a DC level, and then re-generate a PWM signal at the correct frequencies. A neat trick, and one that will be excellent for Boris -- since I need to drive the air valves at a low frequency also. Well, about 10 times slower than the motor controllers... so I am hoping it will work, we'll just have to wait and see.
It is sad when I go out into the workshop -- my test leg just sitting there on the bench. The cutoff wheel cold and lifeless. Our resident teenagers keep piling junk here and there in the corners of the shop, giving it an overall air of disused clutter. There are only five or six months left until the competition -- this is going to be tough.
