5X7 Matrix Driven By picAxe 18m2

It started as a simple way to communicate with a robot.  I wanted my robots to have faces...  I could have one that showed a grumpy face when it got stuck, or when batteries got too low, or just to have a fun looking little "bot" of some kind or another.

Of course, all this tinkering around with lights led me to a LED Matrix display I had lying around the junk box.  I got a couple of them for a dollar a few months(/years) ago, just cause I thought they would make good indicators for experimenting with robots.

There I sat staring at the new picAxe 18M2's that I got when I blew my 18X and wanted to keep experimenting.   The M2 versions of picAxe are "real" chips.  That means, they have enough memory to do some more interesting things with.  And there is the whole faster bus thing, and extra timers, and all that.  The chip is designed to run four (4) processes simultaneously.  Sort of a stubby propellor (8 processors) of course, you have to use different and interesting design processes.   The point is, when you run it in "normal" mode, not multi processing, you can really crank up the processor's operating frequency.

So now I am eyeballing the 5X7 dot matrix display with an idea in mind to using the 18M2 at 32mHz (instead of 4) so I can actually handle the timing of Rows and Columns in the 2057 LED arrays.  Oh boy, this looked like fun.  I could build a 5X7 unit for just over the price of the 18M2 because of the surplus 2057's.   I would let the drivers in the chip supply plus voltage so I could use the same chip to sink (20ma per pin on the chip - plenty for the 2057.

Now I know, I could have gotten the 20M2 (20 pins) and had enough pins to drive the 5X7 display with extra pins to drive more than one 2057 displays.  Like having only one display with a 20M2 on it while having more displays simple wired to the chip.

Things I learned:

• Average current drain is low when you are switching the power by sequentially enabling the Column lines.  The whole 5X7 array averages well below 20ma probably over time roughly 15ma, even though you are driving 35 LEDs in various configurations of on's and off's.
• Flickering becomes an issue at low speeds, some the videos are using 4mHz speeds, so you can see the flicker.  Since I switched to running at 32mHz, each pattern looks nearly flicker free.   That could be improved too.
• PicAxe's just got better for home hacker/experimenters.
• 20M2's can be had for less than four bucks!  More pins!  I'm switching! (I know, the socket costs more and all that, but it's pennies for the total parts costs.

How to build your own:

• Of course, the whole build it yourself project with the schematics and some "get started" software are online at CwhatIcanDo.com
• Go follow the guidelines and build one of these to put into something practical...

Weekend Project: Get Started With Robots

Well, it was time for a lot of cleanup. Google videos transferred to You Tube. I discovered early development videos as I attempted to get the dogbot to find the opening to a door, to running the bottom of a little $4 junkbot I picked up used.

It makes for a mighty interesting thing, the flat robot running around the room, seemingly headless.


This in turn reminded me of the handiest tool in my robot workbench. It's a motor driver board with an IR detector attached. Sometimes, I just hand wire it to an old robot or RC car chassis, adjust some software, and I can test about any platform available.

Simple yet handy tool. If you have a servo tester, then you are playing at the level that could use this subassembly.

Why the picAxe? Cheapest and quickest way to get going. That's about all.

Later, I'll do something with some of the other chips. It's not that I don't like programming in C...

picAxe 18m2 Robots Look Out!

Bot builders look out!

Yes, it's controlled by simple BASIC programs, but holy moly! Someone took all the complaints I had about trying to use picAxes for anything but single solutions and "fixed" the picAxe 18m2 chip! I bought one from SparkFun for Five Bucks.

It goes beyond the otherwise fun 18x. It all started when I tried to get my little black box robot to do a little more than just run around avoiding objects. Yes, the bot is cool, and the 14m is a great little chip to run it around, but that's about it. I really wanted to be able to read the adc coming from the IR Distance measuring device, but about all I could do was make the bot run around and do some intelligent avoidance. I needed more program space!

So I decided to go back to the old 18x chip I had used in the original 1st robot, the dogBot. It had enough program space to have interesting behaviors, but had problems because the timers couldn't run at the same time. So drive the wheels with a pwm signal, but you'd have to turn off PWM, to get the servos to operate. It was cool, but even with the program space, I could only get bursty operation: See this video, you'll get the picture:



So the dog would pull itself forward, shut off the pwm, then pull forward again.

Originally, I was trying to run the thing off a 14m, but the issue was the timers, not the chip itself. Eventually, it could find the door and do simple things. But it wasn't the greatest.




Now I'm looking forward to using the 18m2 to make dogBot come back to life, though I may change it's form (like giving it two-wheel drive and steering?) It will do things different.

By experimenting with a small platform, I was able to run four programs at the same time. Now I don't have to jerk and hack the PWM and Servos, but can run all at once and give a smoother operation. Basically, I've told the platform to constantly read the distance from the object it is to avoid.

This is not much of a video, but it shows the platform slow down and stop, then back up and turn, then it just shuts off and waits till I press what used to be the reset switch and it starts out fast, slows when an object gets close and stops reverses and makes a siren sound. Once it backs out of the way of the object, it constantly beeps (showing me it is in slow down mode). Right now it can't just drive away, but it's only a test platform, not ready for prime-time.




But here's what happens so far... It's pretty nice, a lot better than run and look and run and look some more. I think the picAxe 16m2 is gonna make my life a lot more easy and things are gonna be more fun than work with the multi-processing in the m2 chip. I here it still has EEPROM bugs, but I haven't needed the extra program memory in any of my simple bots. Maybe I'll like the chip so much... Well, who knows?

Yet another Cheap Robot

Lately I've been reading about Arduino's controlling the ever faithful Rumble Bots. They don't fight any more, they just run around under autonomous control.

Bot Bodies - Rumble Bot

In the early days (last year?) I was playing around with wheels and gears and all sorts of things when I stumbled upon an old Rumble Bot. I had to look 'em up because they weren't selling new ones... They have a card reader built in and all kinds of things that would make them a fun Robot.

I was just concentrating on using the body for batteries motors and gears when I saw Deans article via Make magazine. Dean used Arduino, I had used picAxe. When I added up the costs, I discovered Deans Arduino powered unit cost more than mine, so I thought I'd dig out the old videos and put this thing up as a proper project. So here it is on the CwhatIcanDo website complete with videos and step by step overview so you can build one too...

Dean's arduino cost $30
His Ping unit costs $30

Doing the same thing but replacing even the Motherboard means:
08m (if you want to be fancy) $4
08 motor controller board $15
Sharp IR detector unit $15

You can use the upper part of the body like Dean. One cool thing, Dean put touch feelers in the arms of his last conversion, it's a great idea. Use the rear switch already mounted on most models for detecting objects when backing up too!

CwhatIcanDo Projects

Motorola Droid - They Call it a Robot

Got my hands on a new Droid from Motorola. This is the finest piece of gear from moto in a long time. (Since they layed off many software developers.) I'm pretty sure it was google's guys who put the kitchen sink in Android, the Moto and Verizon took the Android, slapped in into a very decent piece of hardware and Voila!!!

Details can be found on C What I Can Do website. Follow this link to Motorola Droid: Article >>

You've got yourself quite a little handheld computer in the Droid. And I did a writeup on CwhatIcanDo Projects. I did a little javascript based script that runs in a browser on the iphone or the droid, doesn't matter. I didn't ask the droid to ignore the screen commands, so it actually jumps around a little, but hey, it works! At least enough to tell me it is possible to control things through a web server on the phone's browser. It works.


So what am I thinking here? Is Droid possibly a brain for robot control? I don't know yet, it may not be fast enough, but the I/O port should be able to talk to a robot motor controller via serial bus.

From Motorola Droid

The big question is can you use the GPS, and all the goodies in there to actually put the Droid into the Robot and allow Bot Control via DROID. Sounds like it would make a good commercial, a little droid with some wheels to get it around...

From Motorola Droid

Video Collection

Just an update at the 6 more months mark. Things are going slower than I hoped, but the site is growing in google PR ranks.



Virtual Robot Plays Online!


I am putting up a new project every month minimum.

This Will be a Year To Remember

We have a new president, so far he seems OK, but he's only been in office for 8 days. The site is running now and doing well as far as reliability is concerned. We are beginning to get a smattering of traffic, and our videos are getting some traction now. We have reached the moment of slow growth while we work out the financial deals and partnerships. Else, this site is going to end up a glorified blog...

Our goal for this year is to go "Public" and to be able to have our creators start to generate some income through partnerships or whatever. Considering the state of the economy, a good how-to website should do well, especially if the projects are fun, cheap and easy. And for the most part, that's what we'll be doing, building up the site projects and getting our accounting software melded into the framework.

Here we come 2009!!!

It Isn't Beta, But it is Open

True, we didn't hit the goal of Beta Release by Thanksgiving,

but it is a Public Alpha. No, I didn't get a design like I hoped, but we are ready. The site itself works just fine. You still need your own YouTube and/or google account and Picasaweb for images, but that is not much of an impediment. As yet, you cannot edit your parts list, only delete and replace. But I believe those are the only two bugs.

Before I call it Beta,

I need to hookup the whole "connect to CwhatIcanDoProjects on YouTube, googleVids, and picasa", so users take advantage of You Tube positioning as a group. And you need to be able to edit the Parts List AND the e-commerce piece needs to be hooked in.

But, there is plenty to do while in Alpha.
Tell a select group of known hackers, modders, and roboteers about the site and solicit their input.

Make our first deals with distributors, we have a "partial" deal with SparkFun to allow me to populate some projects, but there is only permissions for using SparkFun IP, we can't do a distributorship with them (yet). I'm certain I want at least one SparkFun like company, a small 'specialty' online distributor and my sights are on something like electronic goldmine or equivalent, and I really would like to get one of the bigs like Jameco.

SparkFun offers PC board production as well as 3-D modeling, done by batching, on the cheap. Perfect for creators with brilliant ideas to make into products. It's hard to find an equivalent to SparkFun, so I'll just hang with them for a while even though there is no current financial relationship. It's just one of those things with timing not quite right for SparkFun's business plans. We appreciate their contribution to the site. It will work out. Of this I am certain.

And, most important of all, I will take this time to focus on soliciting directly the best possible already published creators from around the internet. Look for a lot of kits of parts, and a chance to split commisions, if I am successful in this part of the effort. These folks can benefit the most directly right now, so they are logical pioneers for the site.

No. I can't yet call it Beta, but It's open for limited business, and subject to basic redesigns, but ready for those who are already on the leading edge.

CwhatIcanDo

Open For Public Alpha Testing

Open for Public Viewing (and participation) go to C What I Can Do now and you are permitted to operate the site and to freely come and go.

This is the time, before official Beta release - Where the design is likely the only major thing that will change - I hope.

From TrakBot: Build Your Own Track Drive Robot

Now is pretty much your last chance to get involved with CwhatIcanDo and become a contributor to the site's purpose and functions. You may just take a look around at the framework and some of the great projects. If it seems like mostly robot projects, that's because it is right now... Lot's of robot projects and a few little scraps here and there of placeholder projects.

I hope you enjoy your interaction with CwhatIcanDo.com!!
(No. I mean I REALLY hope you enjoy it!)

PS: Read the bug list in the previous post for outstanding bugs and the easy workaround.

Sneak a Peek!!!

Site Goes to Full Alpha

From TrakBot: Build Your Own Track Drive Robot

It is outrageous I know... I am offering a sneak preview of the new C What I Can Do website... That is correct earthlings, we are coming out of friends and family testing and I'm ready (though still in Alpha Testing) to open the site to more people. This is a chance to become a contributor on a brand spanking new website.

I'm taking the Site From Friends and Family only to a more generic level of Alpha Testing. There are still some bugs. I'll list the known bugs so if you decide to play around, you won't break anything or get frustrated. Remember, it's an Alpha WQebsite. It's not guaranteed to be bug-free. But the Alpha Testers, like you, get a chance to play around and make suggestions and help improve the site.

If you decide to give it a spin, remember you can't go through the "front door" yet, but you can go into the side entrance and have a look around.

YES! I wanna see this new website! I may even have a project or two in mind. I want to help you make it work better, so I'm coming in to the Alpha Test Site!!!

------------
Known Bugs

• You will have to have your own YouTube, google Video, or PicasaWeb account for now, we haven't hooked up the interface to CwhatIcanDo's accounts yet.
• You can't edit the parts list, if you mess up you have to delete the whole thing and start over with correct information. Sorry, for the inconvienience.

That's it! You were expecting a long list like Microsoft has when it comes out with finished products?

Andrei and I aren't Microsoft, so we are shooting for the fewest possible bugs by release time. I'm hoping we can open for Beta Testing around Thanksgiving...

So far so good...