Atari 8-bit robot the beginning...robotic arm

I'll start a blog on this both here and on AA, one that will be on-going for a while, possibly years. But for now, I just wanted to show a bit of it's beginnings, the robotic arm. I'll be seriously reading and studying the 'Arm Your Atari' article covering two Analog issues, IIRC, #54 and #55 though I might 53/54 or even 44/45 or something close or inverted.

I don't know how much from that article I'll be using, since it covers heavily modifying the old toy Armitron robot and connecting it to servo's and strings/wires, etc. But I believe the software listings, programs, will still work with my modern robot arm, but I'll be wiring the servos directly to the Atari's joystick ports (just like the article) using my Atarilab Interfaces though instead of wires directly to the pins or using DB9 plugs. But I'll outgrow that quickly, it's just a study and starting point to get my creative juices and brain ticking along the lines of robots. Eventually I'll have lots of sensors and camera(s) etc. and will probably have to move to Arduino and other kit boards and the Atari will control them as they control the robot. The 800 will be the main brain and manager, but will probably offload most work to other electronic boards.

So far I am on step 15 of 42 steps in assembling the robot, using instructions in Chinglish with very small and poor illustrations showing the Chinglish steps and having to figure some things out that don't make sense and they don't show nearly as detailed illustrations as needed. I really wish they had just included and exploded schematic of the entire robot that I could easily work with and have no need for Chinglish instructions or other poor illustrations at all. But I've only put in about an hour's work on assembly, and I'm just forging ahead with what I have. A couple hours an evening and I should have it completed in a day or two. An example of the Chinglish I am dealing with: "After install the bottom servo bracket. Start to install steering wheel before install steering wheel." And I really don't believe "steering wheel" is even the right name of the part they mean. It is probably called a 'control mechanism or some other' that attaches to a gear the servo motor to allow movement in one of the six degrees of freedom of movement the robot arm has. But whatever, I'm smart enough to get it built with no instructions at all, it would just take me ages longer to figure out the proper way.

I'll use this thread to cover everything I do with the robot arm and the 'Arm Your Atari' article. After that, all else regarding it will be put into a blog covering the entire robot building process and using my Atari 800CX as a testing and prototyping tool through it all to the point I actually install my other 800 mobo in the robot itself and make it antonymous. And of course my Power Analog/Digital trainers (P.A.D) there will also have a big as role as the 800 in research and development too, for testing cicuit designs and power for robot parts that need more that the joystick ports +5V lines.

I've decided to make my own mobility base for the robot, it may be tracks and treads or wheels most likely, but that is a ways off in the project and I may end up giving it limbs for mobility if I think I can as I learn and move forward. But after seeing the prices of anything larger than a toy/kiddy educational robot base has made me decide I can make my own better, more powerful and cheaper than the ones selling for $500+ and anything under that price is a toy.

 

Good luck with the wanking arm Matt, don't forget to file down the sharp edges!

Seriously, what a great project, I'd have loved to have tried this as a younger person, might have improved my poor electronics. I remember as a young kid I had a robot that walked, flipped its chest open and shot sparks from 3 guns on it, was my fave toy for a long while. Always wanted to take it apart, but my parents would have not been happy.

This on the other hand is a fully fledged robot in the making and being controlled by the trusty 800.. Excellent..

Program it to stick a lump of dynamite up Putin's arse..

Again, seriously, I'll be watching this unfold... Great stuff..

 

Well, after actually reading through the Arm Your Atari, Issues 44&45 by the way, I have discovered that 95% of the article was about modifying the Armatron and adding servos and potentiameters. Since I have a modern robot arm that will have all that, once assembled there are only 2 things from the article I need to use at all: the pinnouts of joystick ports 1-3 and the Basic listing that will allow me to control the arm with peeks and pokes.

There is actually more information of use to me in the same magazines in the Bits and Pieces column that is covering interacting with the outside world using the PIA (PIAs in my case) and controller ports.

I will also have to look into the data sheets/specs of the servo motors I have that each have 3 wire ribbons to see what they are each for, but I suspect they will be the same I/O wires that the DIY servos/pots use in the Arm Your Atari articles and I will attach RCA cables to those words which will then plug into the Atarilab Interfaces connected to the ports.

Then it will come down to programming more advanced and complicated programs to make the computer control the arm autonomously instead of the simple program listing that requires keyboard input from me to move the robot. And so I will be moving on to more than just the arm in short order.

 

A man with a plan, good luck...

 

Very interesting, this is the kind of project I would love to do too! Maybe one day... Good luck, I'll follow your project.

 

Thanks guys. As always, I have the robot in the back of my mind, thinking what kinds of things want to give it, to make it actually useful, but also some stuff just for "bling" and a "cool" factor, like I'm thinking about using my lighting plasma ball for it's head. It has a setting to react to noise and I thought it would be really cool to have it flash with lightning when the robot talks (S.A.M voice and built-in speaker). I always like the clear "dome" heads sci-fi robots like Robby, or "Robot" from Lost in Space had with plasma/electricity and lights flashing. The plasma lighting ball would fill this want pretty well for what I have on hand without too much work and actually be reactive and not just randomly flashing stuff.

Anyway, I have decided that one arm just isn't sufficient and my robot will have two proper arms. I want two arms so it can really manipulate and carry objects.Though I may get a different type of robot arm for the second, so they won't necessarily match, I think I want to get an arm with a five-fingered hand, the arm I have now just has a claw, like R2-D2. Also I purposely want my robot to look like something built from the Hodge-podge of parts like it will be. It won't look like junk, but it won't be necessarily "sleek" either. Two different arms would help with that look.

For a brief moment I even thought about building my 3D printer into it so it can make stuff like Robby the Robot from Forbidden planet, though it won't be emeralds...but it's too big a contraption. But I think when I make the Wefax decoder device that I will install it in the robot with the short-wave radio and a 1020 plotter for print out. It will also have a screen in it's chess, a B/W 5" screen from an old karaoke machine I salvaged. So if no 1020 is installed a read-out will still be on the screen.

Another idea is adding in a Voice master so I can have voice control of it, etc.

I think it will be about 3-4 foot tall in the end.

 

The robot arm is fully assembled. Now it is time to get a data sheet on the servo motors just to know what the three wires coming out of each one is used for, then start making wire harnesses going from the servo wire connectors to RCA plugs at the other end ready to plug into the RCA jacks on my AtariLab Interfaces.

From there, it will be all about programming, both autonomous pre-programmed movement and and interface program to control the robot arm manually with joysticks, it will probably require two joysticks to allow for all the movement controls in six degrees of freedom. This is were my two modified Starmaster joysticks will most likely be used (one borrowed from the 7800) as they have the suction cups to hold them in place so I can operate both sticks at the same time, and I believe the yoke-like grips will make it much more comfortable than standard CX40 straight sticks.

Not only are the manual controls for me to have a bit of fun with, but if I get good and proficient with them then instead of just typing in movement coordinates, a program will allow me to manually move the robot arm and record those inputs to be used in programs speeding up programming and less math and geometry to worry about calculating.

Now that it is assembled and I have a better idea on different ways the arm can be mounted to a chassis or body, I think that one of these arms mounted on either side of the robot chassis/body will work well for manipulating and holding onto objects, and I probably won't try to use an arm with a five-fingered hand, as that just complicates things for me, and the 8-bit Atari that will be running the robot. It will leave both I/O more free for other things, and far less complicated processing for the 6502 to require for manipulating things. Besides, it's going to be more like a Star Wars "Droid" type robot (like R2-D2, not C3-PO) and not a humanoid style biped or an android.

Six servo motors on the arm to enable the six degrees of movement and claw manipulation.

The thread will continue, though it may be a few weeks before another update as I still have some work to do on my 800CX before I can get started using the arm with it.

 

Your bid to recreate a T2 model takes a step closer

You'll be back!

 

From your friends at Cyberdyne Systems. "Manufacturing the future today."



Now you got me thinking about weaponizing it...maybe a taser, then if I have an intruder I can send the robot out via remote control as a first line of defense...Hmm...might not be a bad idea, especially since the only dog I own that would actually attack an intruder as a first defense can't stay in the house because he attacks the other dogs.

They are good for an alarm system, otherwise they are too friendly. Though they've never been in a position to protect their master's if we are being attacked and I don't know how they would react. But my hound dog Charlie, definitely, without provocation, just due to being a stranger on the property. But he stays in my office trailer at night as a guard dog and is on a leash next to it during the day.

Not that I've ever had an issue of burglars so far, generally they would look for richer neighborhoods and are too lazy to go way out in the middle of nowhere, but you never know and I live in just the type of area that escaped convicts would look to hide or criminals on the run...1 in a million, but...

 

Hi Matt
Well you are getting on, sorry I've not been about much over the last 6-8 months.

Servo wiring. you may have found this by now, two leads for power nominally 5V but there are some that will go to 7.2V foe a pair of Li-Po battery's in series.
the third wire, you need to send a pulse train between 1mS & 2mS the center position being 1.5mS
That's an awful lot of bit bashing to do even on just 8 servo's (bits) whilst still running the Atari operating system.

have a look at this https://www.hobbytronics.co.uk/servo-controller-12ch-ht?search=servo
and this https://www.hobbytronics.co.uk/servo-pro-12ch?search=servo

these can control a number of servos as seen, the problem is it wants an i2c interface for a 6502 have a look here

http://www.c64os.com/post/i2c6502
http://retro.hansotten.nl/6502-sbc/lee-davison-web-site/i2c-bus-interface/

and especially this as you have an extra 6520PIA


https://www.seeedstudio.com/Atari-8-bit-I2C-Interface-(A2I2C)-g-1127014



more here
https://atariage.com/forums/topic/135092-i2c-interfacing-with-8-bits/
https://systemembedded.eu/viewtopic.php?t=32
https://en.wikipedia.org/wiki/I²C


there are some servo based chip that can use SPI interface https://sbc.rictor.org/65spi1.html
this is in hardware, best to connect via 2nd PIA

sorry as its not as easy to use the pot lines to read the positions.

 

Wahey, Graham has popped on

Good to see you...

 

Hi Paul
yea been out of it for sometime looking after Pam, also had a minor stint in Hospital myself about a month back, only 5 days and sorted out now. Diabetes playing up.
Hope you and Cindy are Well, p.s. hows Uni going for Serena. I'm a pleased Dad as Kat just got results for a MSc

 

Welcome back @Graham! It's great to hear from you. And yes, I've already got the data sheets for one, and know all of that, actually, I already knew one was ground, one was +5V and one was signal 1-2ms, I just didn't know which colors were which until I saw the data sheet.

And have already been doing research on interfaces, which I will be building my own, at least to start, no Arduino's or Rasberry Pi's yet, I've already got a circuit drawn up for using a 555 timer and through PIA, etc. I will take a look at some of the links above and see what other possibilities there are, but like the last time we were speaking of the dual-PIA board and the robot, you are getting way too far ahead for me, it will take me a while to go back and go over all you have told me already.

I'm taking baby steps, and soaking it all in, starting with 1+1+2. I really want to build my own interfaces and such from scratch, not follow someone else's designs or commercial ready-made boards. I am starting with basic components only. At least to start with to experiment and learn with the arm. Later on I'll get more sophisticated, look at all the possibilities (basically all the great stuff you have posted).

I'll only being doing a bit of experimentation with the arm and the servos that came with it anyway. I've already decided I'll be extending and reinforcing the arm(s) and I'll be getting better servos. +5V doesn't cut it if I want it to actually be able to manipulate and carry anything more than, IIRC from the data sheet, 2.5grams of torque. Maybe 12V servos, but I'm hoping to get some 18V servos at least, and bolstering and modifying the arm as needed.

I obviously won't be relying on the motherboard to power the arm(s) but will be using at least 18V Lithium Ion rechargable batteries. already decided to make my own chassis/axle/wheel/trac setup and most likely will be using 18V/24V cordless drill motors to operate them, and another battery or two for the upper body and arms, again, probably 18V minimum

I agree that once I'm at the point of wanting the arm to be working while the OS is doing other things and I've have other mobile parts and sensors and whatnot (whatever I decide I want the robot to be able to detect and react too) that it will be too much for the 6502. I have been considering that. And several plans come to mind. And which one I go with will depend on how well things work out as I go along.

One possibility I might consider is more than one Atari motherboard in the robot, working together in a local area network using the second PIA's new pin functions (sorry it's late and I can't remember atm) taking the place of the pins used for pots on the first Pokey, with dual Pokeys in all motherboards used, and of course this opens up even more ports to control more as well. I have two 1200XL's and an 800 motherboards at my disposal (not including my every day 800 and 1200), so I could have up 2 or 3 in a network.

Another thought is not using my second 800 mobo at all, and instead using just my two extra 1200XL's with Rapidus 65816 CPU boards networked.

The third option, which is cheaper, but I don't know if it would be a better alternative or not, is converting the CPU's of the 1200's and the 800 over to 6809's. Actually I'll be using the newer better Hitachi 6309's that are 100% compatible with more features. I already upgraded my Coco 2 with one. Of course I'd have to add in my own 6309's and the pull-up board like I did in my Coco 2. With the Liber809 board found here: http://www.cloud9tech.com/


I appreciate all your help most definitely, and I'll definitely be asking you lots of questions, clarifications and verification on stuff, but I need to look through all you already sent in the PM's last year and links above, take stock of it all and then I'll be ready. If I can possibly have 80's 8-bit's as the answer to processing power before relying on more modern alternatives, I will strive for that.

But I really, want to teach myself, and learn everything and have total understanding and comprehension, and not just build something by using a bunch of ready-made boards and interfaces like it's all a kit I'm piecing together from different sources. That's last resort if something I need to do is beyond my abilities or facilities. Building it all slowly, from scratch over a few years too, not something I will throw together in days, weeks or even months.

So please, bare with me and I will ask for your very generous assistance and definitely superior advice, etc., but I really need to be able to take my time, one step at a time, and not have overload of info all thrown at me at once. I'll get to it, as I'm ready for it. I'm on chapter one of this robot story that will be 50 chapters long.

Thanks for the great info and thanks for understanding that the entire point of this project is not to have a cool little robot to play with, but really learn everything it takes, inside and out, hardware and software and I/O on my own. That is the point of the project with a real working robot that's more than a curiosity toy as my "graduation" present to myself.

It may be a year before I'm done building DIY interfaces and learning to program it all and testing and experimenting with just the robot arm alone. I'm not even thinking about going beyond that until I can program just the arm to manipulate stuff, and I'm starting from scratch basically learning Basic and Assembler again since it's been about two years since I was teaching myself and have now forgotten all of that!

At this point, I just wanted to finally assemble the arm instead of it sitting in pieces in a box, and it is now back up on a shelf. This was just a short break from the soldering iron and such. As I still have a bunch of repairs and upgrades I'm in the middle of for 4 different computers and three different consoles! This is still including my 800 that only has the Dual-PIA installed in it and I still have to connect it to the new ports and mount them. And about 2-3 weeks to get them finished before I'm too busy for my hobbies, for the most part, until next winter.

And please don't take this as some kind of brush off to you @Graham , I want all the help and advice you can give me and linked information, but just not information overflow which is what I have now between all the posts regarding the Dual PIA and robot before already. Let me catch up with all the stuff you gave me before and now in the post above, along with getting my feet wet on my own with the robot arm and temporary experimentation. Even the DIY interface boards I use will be temporary for learning purposes most likely. When I'm ready, I'll come to you for all the knowledge and advice I can get, with plenty of questions and ideas to go over with you, if you are willing to help, sort of like a professor. But this is an independent study course and I need to start out on my own, at least until I'm beyond just one robot arm. Once I build my interface circuit with the 555 timer and play with it a bit, I'll look into all the links you left me above for other alternatives.

Oh, and by the way, no matter which way the project goes, I have decided to use the PS/2 keyboard interface you gave me. I have a PS/2 to USB adapter and I have a wireless keyboard I'm using right now for the robot. Though if I stick with 6502's or 65816's I'll be using a Fujinet on-board for remote control too. So I'm still leaning toward 2 or three 6502's for all the bit banging as the cheapest way forward allowing for full use of Fujinet and Incognito on the 800 and U1MB boards in the 1200's , etc. Which is still possible of course with Rapidus boards installed, but quite expensive. There are alternatives with 6809/6309's but not as good as fujinet. Might even be able to just use Fujinet and bluetooth to network all motherboards and still save the PIA's for other things, or a combinations. I know Fujinet has come a long way since my 1.0 with a 1.6 out now, and tons of new features that I haven't even looked into yet also. There is tons of new updates and apps now even for Fujinet 1.0

 

Congrats to Kat, great work there!

Obviously, I hope Pam is doing as well as can be, and you look after yourself as well as others. Blooming blood sugar is a pain, it's one thing I've not suffered with (but I'm sure it will catch up with me at some point), been borderline but never in to needing treatment. Apart from that this damn long covid is causing hell for me and Cindy, Cindy far worse, always tired and aching. I'm not far behind, but still fighting and making sure Cindy does not get worse. Serena just gone back to uni after an extended weekend with us, her fatigue is showing through, any set of long days, and she is totally tired for days, but her work is great, tutors very happy with her, and we support all we can..

You folks take great care, our best to you all..

Paul..

 

I think I'm slowly talking myself in to getting a Fujinet, I keep hearing all sorts of wonderful noise about it and will have to look at all the features to see if it has a place here. I have to be careful, as it's something I have to plan for (unless I win the lottery). Initially I had no use for it, wifi is no big thing for me, but I keep hearing of new features that keep me circling it.

Depending what is on it now, it might be a birthday present..

 

Even though I've had Fujinet for probably about almost two years now, I haven't used it for over a year with my 800 apart for that long going through all of it's upgrades and mods. And I've heard noise here and there since and it sounds like with all the new soft ahd hard developments It will be like starting over for me too. Especially since I've forgotten what I did know due to lack of use.

I haven't had time to read through all the new threads in the sub-forum and usually don't keep up on the latest unless I'm currently using something, as I just forget it all again anyway before I'd get back to it. I can't wait to get back to it and catch up though, and start building my own...I guess 1.6 version now, instead of the 1.5 I planned a year ago, for my 1200XL, mounting it in an my old 1030 case along with an SIO hub since there's plenty of room for both in the 1030.

It's actually the same situation with Fujinet and information overload I need to catch up on like I was just telling Graham with all his generous help with the Dual PIA and robot stuff. But I'm still so far behind on repair and upgrade projects for multiple computers, consoles and peripherals it has to wait until they are all done for me to catch up. I don't even have a working 8-bit atm to do so. As even though my 800 is done with it's upgrades and mods (except for connecting the extra controller ports mounted in an external case I still have to design and print), but my PSU board has a bad regulator or whatever for the +5V I need to repair now, it went out during all the upgrades and testing either by natural death or something I did.

But I still hope to have my 800 and 1200 both working within the next couple of weeks. I'm still busy with home improvements too though, trying to get it all done before my work takes off in a few weeks, and issues with my mother who could go at any time. When she passes, not that I'm looking forward to it, but it will open up a lot more free time, as well as once the house it finally done too. But I also have to turn my focus to start getting all my business machinery up and running and all in order for the new season too. Just not enough hours in my day currently, even with me cramming as much in as I can and averaging only 5-6 hours sleep a night.

 

Just back from Hospital, nice quick visit. I just read through some of what is new and it did nothing for me, I was hoping for some new fancy stuff but CPM, Python stuff and some Coco related stuff just wasn't what I was hoping for. No disrespect, but it's all a bit too geeky for me. I did all the geeky stuff I wanted to do back then, now for me, it's about having some fun which is why I stopped coding a while back, any sort of game or util is almost certainly already out there and coded to a much, much higher level than I could do. Add to that where I find it difficult to stay focused these days, which from reading science sites is almost certainly down to COVID, there's already a study that confirms this.

So as quick as the Fujinet sounded ok again, it's lost its appeal.

I think it's just me fancying a treat and feeling a bit sorry for myself, I would not mind, but I've got stuff to do such as repairing a C64, installing that 32 in 1, building a cable and just generally tidying up the flat of clutter, but the sec I think of doing it I just give up. I dunno what is going on with this COVID, but I've never been a lazy person, but suddenly I'm just out of sorts. I am doing stuff, I just redid the bathroom, painted it, demoulded the window frames and some other little jobs, but I'm so shafted from doing it.

I guess the idea of the Fujinet was a sort of pick me up but when common sense kicked back in I realised it was a silly idea.

 

I don't care about CP/M on Fujinet, I already have three drives with upgrades ready to install, so in theory I could have 3 CP/M machines already if I wanted to give one drive to 3 of my computers. But one is enough as it is a curiosity for me, There are better apps already on the Atari for productivity. I got the upgrade for all three because it also makes the drives even faster and more features too just using them as Atari drives.


I don't know what Python is, or at least if I did, I forgot, the name does sound familiar.I think it's a cross or multi system OS, is my guess?

The Coco stuff you mentioned interests me though, now that I have one. That will mean being able to communicate and transfer files and networking games between the Atari and Coco, and fits right in too, with possible options for a multi-processor robot as Fujinet for Coco would allow the possibility of mixed processors/systems in the robot too, each handling different aspects of the whole and communicating via BT or Wi-fi, instead of me bit-banging the PIA ports for everything (and of course the Coco has 2 PIA's of it's own on board! Leaving ports and PIA I/O available for other things besides networking possibly.

But, that is all speculation and ideas and I have a lot to do before I even get beyond using my desktop 800 for research and development of various areas of the robot and then when they are ready to come together into an actual robot, that's when I start looking into multiple computers on board and communicating with each other wireless.

 

Python, multiplatform scripting style language..I thought the Coco stuff would interest

Have fun Matt..

 

You read through some fujinet stuff on Atari Age 8-bit fujinet subforum? I went scrolling and skimmed a topic on multi-platform, saw stuff for CBM and Coleco Adam and even PC and Apple II, but nothing about the Coco. I didn't see a Coco specific topic in the first two thread topic pages. Or did you go to the Fujinet site? I'll see if I still have the old bookmarks...

 

I was doing a bit of research on servos. looking for more powerful ones than what came with the robot arm, as I want the arm(s) to be able to lift a bit of weight. I did find out that I can get some much more powerful servos that still run on 5 volt power. My current servos have 2.5kg of torque, but I have seen some reasonably prices 5 volt ones between 20kg-30kg and even higher, if I did go up in voltages to 7.2V or 12V, etc. but for the arms I do think some 20kg-30kg servos that are the same dimensions and voltage is probably the way to go.

Of course I will get the arm move as-is and actually see what type of weight it is capable of grasping and lifting/holding with 2.5kg of torque under real application and then decide just how much weight I want my robot to be able to manipulate and carry. And then see how strong the arm chassis really is and if is strong enough to handle weights acceptable to me or if I need to reinforce the arm chassis. Right now that translates into about 5.5 pounds of weight, if it can actually lift and manipulate that much weight, then replacing them with some 20kg servos would definitely be enough for me and probably be over-engineered for my uses, but I'd rather over-engineer it.

 

Oops Matt, I may have read Coleco as Coco

 

And I would have got a Coleco Adam computer instead of a TRS-80 Coco if it weren't for the fact that the Adam requires it's printer for the PSU to everything and complete systems are selling north of $500. I almost put that much into my Coco by now, but it's not all at once up-front, which doesn't fit my budget.

But the Coco got better support and has a much larger library, if small compared Atari and minuscule compared to Commodore. And was also and is well supported with hardware, legacy and modern, like the Atari. It does surprise me that they are doing a Coleco Fujinet before Coco, but they must prefer the Coleco. I'm sure the Coco will get a Fujinet in due time though, and there is still plenty of that before I'm at the point with my robot to be concerned about it yet.

 

I guess the Adam' was something that Thomas Cherryholmes was intrested in, so he could write the software to match (Z80 I think) I'd guess that with the serial IEEE port that Commadore use for there drives, Fujinet could be ported over, The coco use yet another processor 6809 like a super version of the 6800 & 6502 before it.

Re Servo's Usually there is a trade off in servo performance, high speed low torque compared with Hight Torque and low speed, depends on the output gearing. always best to go with Metal geared servo's, there are of course several sizes of servo, however if you are limited by the servo frames you have your likely to end up with high torque but very slow speed. also if it's a chinese spec it's unlikely to be true anyway. bit like the cheap LED lights I purchased 36W LED ... the PSU is rated at 12V 1 A ie 12W MAX.

 

Yes, the 20g 5V servos I have my eye on are metal geared. Yes, when I go with cheap Chinese stuff, I always test before installing/using them, it's a crap shoot for sure, but generally it's just basic components like 74LS chips and sockets, resisters and capacitors.

Fortunately, I don't need my robot to move fast, I just need it to be able to manipulate some real weight, and the servos I want will be able to handle roughly 40 pounds. With two arms, reinforced by me, it should be able to do some real lifting and carrying. I just want a robot that is actually useful, if limited in ability, and not just a toy and curiosity that can move a cup or chess piece only.