SBC kit for education

[GARTHWILSON]

Quote:

- vintage and historical appeal: Apple II, Commodore, Nintendo

This might need some delicate handling. Without throwing away the vintage appeal, you'll want to make sure people don't automatically think "Old. Incapable. Slow," before their imaginations are opened up to what they could do with it.

Quote:

Needless to say, the idea of such kit brings many questions: PCB or WW? Display and keyboard or just serial port? I would be very much interested in your angle!

The PCB in quantity is inexpensive, and will prevent frustrations that may come from things not working because the construction was poor. There are plenty of tips about construction in the 6502 primer though, including why not to use solderless breadboards for this kind of thing.

I think it's good to have a keypad and LCD on the computer itself, and I put these on the workbench computer, even though source code gets edited on the PC with the full keyboard and hi-res monitor, and kept there on its disc drives. Even a minimal keypad and LCD are very useful.

Quote:

Yet, some people like me want to learn more about the underlying concepts these devices.

...and that's a very valid point. Too often, the argument today is, "Why don't you just get a ___? A lot of that stuff is already done for you." And that's exactly the problem. They try to get the mundane stuff done and hidden so you can "get on to the exciting stuff," and then the user never really does learn about the innards. Both have their place. One does not invalidate the other or render it obsolete. When I was starting out in 1981, we assembled our 6502 programs by hand and punched in the machine code into the AIM-65, even though the AIM-65 had a rudimentary assembler. I think it was valuable that we started that way, because we got a better feel for the very bottom layers. It would not have been doing us a favor to hide those from us.

Quote:

this is something that Daryl (8Bit) has already done with his SBC series. The ancestral unit in his series was 65C02 powered, and about as basic as one could get and still have a functional piece of hardware.

He did a very good job of it too. He would probably be glad to do another run of boards if you want to order it for your class. And as BDD alluded, the high price of WW sockets would justify the PCB anyway. (That's not to say there's not a place for WW. There definitely is, if you're making just one of something and know you will want to add to it later or might want to modify it.)

Quote:

most drivers known nothing about what goes on under the hood (bonnet) or inside an automatic transmission, but are able to drive the vehicle.

Not enough sometimes. Pilots have to know more, and go through simulations of engine failure and different things in their training. Fortunately the car's safety parts are pretty reliable; but, for example, once in awhile a car's brake hydraulics really do fail, and many (most?) people don't realize that the emergency brake is called that for a reason, that it's a cable connection to the back brakes that is immune to hydraulic or vacuum problems, and you can lock up the back wheels with enough pressure on it. Drivers' training ought to simulate failures enough that people don't totally forget in emergency situations.

Quote:

could fill the vacuum left by the Arduino microcontroller and the Raspberry Pi SBC. These devices have attracted a large crowd yet they do not explain how computers, processors and other digital hardware work.

Yes, those rank closer to appliances.

Quote:

The original Cray supercomputer was wire-wrapped and had a 12.5ns machine cycle time.

According to the pictures I've seen, I think it also had the wires in twisted pairs though, so each signal had its return twisted with it, which makes a big difference in performance.

Quote:

An optional serial to USB converter (FTDI) with a pretty, cross platform software interface(IDE, serial monitor, virtual TV screen, memory map) could make the SBC more accesible and easier to use.

Those should be add-ons which the computer doesn't need in order to get going. USB is not hobbyist-friendly by any stretch of the imagination. I have a contention with the name too. For one thing, how can they call it "universal" if there are a half-dozen kinds of plugs. For another, it's not really a bus if normally one port can only go to one device. Also, the spec. precludes having a hand-held battery-powered device being a controller. Also IDEs are kind of an appliance, hiding more of the development innards. I don't use one, yet I can write a piece of code on the PC and have it working instantly and see instant results on the workbench computer. It can be done with any computer with a text editor that allows "printing" a block to the RS-232 port. No special software, OS doesn't matter, etc.. The workbench computer takes in the text as if it were a line printer (ie, non-graphics mode), and compiles, assembles, or executes the code on the fly as appropriate. Eventually I would however like for use of a host computer to be optional.

[BigDumbDinosaur]

BigEd wrote:

Oh, terrible. Let's not do that then. It wouldn't do to show the registers and the instruction stream. That wouldn't be educational and helpful. Because BDD has spoken, and he's against it.

I'm guessing your interpretation of "front panel" and mine are different. Maybe you are referring to a machine language monitor, which can display those items and which should be visible to the user. However, that's really a software matter.

BTW, anything that I might say about this sort of thing is an opinion...not a fiat.

[ChuckT]

alkopop79 wrote:

Needless to say, the idea of such kit brings many questions: PCB or WW? Display and keyboard or just serial port? I would be very much interested in your angle!

I think if you want success, you have to go with low cost and the social media which is the Raspberry Pi. In order for some students to learn, they have to learn at home and $300 eprom burners are out of the question. The second issue is that you don't have the social media presence to promote 8 bit computers anymore like Compute!, Compute's Gazette, Run, Ahoy, etc. There is also the attention span factor; you can't get easy to use video or sound chips for the 6502 to interest students.

I went to college and students who had no idea or exposure to computers chose computer classes expecting the teacher to them them and the instructor didn't. If you don't know how to cook, giving people a bunch of ingredients from a recipe isn't going to help. They have to be willing to buy books, magazines and spend some time programming which is what we called "cookbook".

I spent several years reading about Microcontrollers, breaking the piggy bank and I still don't have time or know how to program a single microcontroller but I'm going to give it a try soon. You have to be young, have a lot of time and have no commitments. You have to be willing to learn new languages like C or Linux to get ahead today.

I got a Commodore 64 for Christmas and I spent each day trying to learn a new command until I got bored. I then took Basic in high school and again in college because the college made me. I learned another language called Pascal and I think it is from being fluent that you start to understand some things but all that learning didn't translate into understanding electronics, processors or microcontrollers.

There are financial and social reasons to make it easier to learn and each platform has its advantages and disadvantages but the days of Jim Butterfield coming along and helping me learn Apatco's computer with magazine articles isn't going to happen because those days are long gone. Instead, I have to spend the day at the library trying to figure out what books I can get my hands on to borrow so I can learn. I hope they didn't throw those old books out to make room for fiction.

You basically need a lot of geeks to come alongside of you and I think that you would have more success with the Raspberry Pi on the programming side. If you want to learn through building then pick up Arduino or another microcontroller where there are more tutorials. I think people can learn a lot about xbees and servos.

If you want to make the 6502 survive, you have to have kits and a culture to promote it. License a language. License a bare bones kit. Post tutorials. Create an online magazine. Write and produce some youtube videos. If you don't, you will probably be the last generation to use the 6502. Cars have computers in them and I think some of them are ARM products and not 6502. You basically have to promote sales by giving the next generation a product so if you want to be important, put your name on some projects, copyright them and patent them. Write a few e-books and put your name on them.

I'm learning ARM and the Arm Cortext M0 and M4 so I just ordered more parts and I'm ready to learn. I would be willing to learn with anyone but I also chose this alternative route. I'm willing to teach my son's clases what I've done but unfortunately it probably won't be 6502 related because the material available isn't geared for that so I've chosen ARM. I can be a great teacher but someone has to teach me first.

In order to have sales, you have to have service. Where is the service to teach and promote the 6502? Does it exist to the level that kids are learning it? I would say more service exists on the level of the Raspberry Pi.

[GARTHWILSON]

Quote:

If you want to make the 6502 survive, you have to have kits and a culture to promote it. License a language. License a bare bones kit. Post tutorials. Create an online magazine. Write and produce some youtube videos. If you don't, you will probably be the last generation to use the 6502. Cars have computers in them and I think some of them are ARM products and not 6502.

Much of that is being done right here at 6502.org. And actually, the 65c02 is going into new cars, and the 65c02 sales volume is in the hundreds of millions of units per year. It's just that it's rather invisible, being the processor at the heart of custom ICs.

As WHartung recently wrote about 6502.org,

Quote:

...the 6502. I don't think there is a resource dedicated to actually creating a machine from the ground up for any other CPU as comprehensive at this place is.

Much of it carries over to other CPUs, but as far as building boards and programming them and getting things to work, this is the best place on the internet. Everything else is mostly dealing with existing, legacy systems, or pre-made boards purchased from a vendor, or simply higher level languages that are pretty much CPU agnostic.

Dajgoro wrote,

Quote:

there really wasn't any other community that could help with the project. Some stuff can be found for the 68000, but nothing to compare to this community.

alkopop79 wrote,

Quote:

Strangely I could not find a single tutorial or article on 80xx series SBCs. Being one of the most successful series of microprocessors, it's ridiculous that there's nothing out on the internet. Even the 8086.org articles are useless

6502.org is developing a reputation which is spreading, as Hack a Day, Sparkfun, and other sites pick up our articles and give us more attention. Even though there aren't very many on the forum who post regularly, good things are happening here which I might not have guessed a few years ago. Rather than just slowing the death of the 65-family, it's growing, and it will outlast the Pi and other such products.

[alkopop79]

BigDumbDinosaur wrote:

alkopop79 wrote:

The Apatco kit got excellent reviews. What most people love about it is the manual/booklet that comes whit it. I guess if the kit is well supported with abundant and comprehendible information, it can be a success.

Ah, you've hit on the "Achilles' heel" of kits: documentation. Unlike creating the hardware, which can be very interesting, documenting it is akin to drudgery for many people and something to be reduced to a minimum. A DYI kit has to be very well documented in order for the user, who is trying to both learn and see a result from his/her effort, to maintain interest. That's in addition to the finished unit being able to do something more that scribble on a screen and accept some typed input.

Quote:

An optional serial to USB converter (FTDI) with a pretty, cross platform software interface(IDE, serial monitor, virtual TV screen, memory map) could make the SBC more accesible and easier to use.

You might as well then promote the Raspberry Pi if you want to go that route.

Why is that?

[GARTHWILSON]

Quote:

Why is that?

I partially answered that at the end of my 2nd-to-last post above, but to add to that, all these gee-whiz features add more layers between the metal and the user; and although they might do a lot of cool stuff, they also impose more limits because those layers are designed or programmed by people who did not take into consideration everything you might want to do with the gizmo, and there may not be enough information supplied to get past those layers and get the level of control you need for the job.

[alkopop79]

Makes sense however I never had problem with the FTDI converter. It's pretty straightforward. I can feel your sentiment about using a host computer. A keyboard and an LCD display would be more suitable. In fact, I don't like the idea of having to use a 1000x more powerful computer as a host.

[alkopop79]

ChuckT wrote:

alkopop79 wrote:

Needless to say, the idea of such kit brings many questions: PCB or WW? Display and keyboard or just serial port? I would be very much interested in your angle!

I think if you want success, you have to go with low cost and the social media which is the Raspberry Pi. ... The second issue is that you don't have the social media presence to promote 8 bit computers anymore like Compute!, Compute's Gazette, Run, Ahoy, etc.

I do have social media presence. Not me but the company I (occasionally) work for. It hurts me to say and write but Twitter and Facebook are powerful weapons of publicity and promotion (I do despise both). May I add, the DIY industry is before an explosion. Very posh high-street retailers (that British English for Wall Marts of the rich) showed lot of interest. Tesco (the other end) already sells Arduino.

[alkopop79]

"Since the object of this hypothetical kit is to get individuals' "hands dirty" with digital circuitry, insulation via PC-based software abstractions would tend to be counterproductive, methinks."

It's just helps the user to get started. The software would be an option, otherwise the user could chose the keyboard+LCD display combo.

[alkopop79]

Quote:

vintage and historical appeal: Apple II, Commodore, Nintendo

This might need some delicate handling. Without throwing away the vintage appeal, you'll want to make sure people don't automatically think "Old. Incapable. Slow," before their imaginations are opened up to what they could do with it.


I don't think so. Who thinks the mighty Commodore or the famous Apple II are incapable? The idea here is learning and doing. People buy DIY kits to make beer. Sure, they can just buy cans at the supermarket for cheap and you don't even have to wait weeks but that's not the same as brewing your own. Again, making beer is an ancient craft yet there are literally tens of thousands of micro breweries in the US. The SBC would not meant to be for browsing or video editing. It's about joy of making and understanding.

[ElEctric_EyE]

alkopop79 wrote:

.It's about joy of making and understanding.

Where is the joy after suffering for a year on my old project without internet help? There was no joy, until the discipline conquered the unknown mistake on my part, without the help of "voices".

[Dajgoro]

The think i would most like to see is a proper IDE environment for the 6502, like the 68k has the ide68k with a C compiler in it. I know that there is the cc65 but it is almost a year that i am trying to figure it out (it is not that i am trying with it every day), but even after i found toes documents, i still couldn't get it to work right for my system (in the simulator)...
If such a tool existed, i think that there would be a lot of cool project with a 6502 just to make them cooler, and it would gain more interest.

GARTHWILSON wrote:

6502.org is developing a reputation which is spreading as Hack a Day, Sparkfun, and other sites pick up our articles and give us more attention.

We definitely need more of that, because it seems nobody from my generation never heard about the 6502, every time i tell someone about my diy projects and the 6502 they are immediately confused, only collage professors recall something about the 6502.
Also many people think that it is like impossible to build your own computer, and that is like a taboo topic. So we should post more Youtube videos, Hackaday posts, and post into some other sites too in order to brake the ice/taboo.

[ChuckT]

GARTHWILSON wrote:

Quote:

If you want to make the 6502 survive, you have to have kits and a culture to promote it. License a language. License a bare bones kit. Post tutorials. Create an online magazine. Write and produce some youtube videos. If you don't, you will probably be the last generation to use the 6502. Cars have computers in them and I think some of them are ARM products and not 6502.

Much of that is being done right here at 6502.org. And actually, the 65c02 is going into new cars, and the 65c02 sales volume is in the hundreds of millions of units per year. It's just that it's rather invisible, being the processor core of custom ICs.

{Extra Deleted}

6502.org is developing a reputation which is spreading as Hack a Day, Sparkfun, and other sites pick up our articles and give us more attention. Even though there aren't very many on the forum who post regularly, good things are happening here which I might not have guessed a few years ago. Rather than just slowing the death of the 65-family, it's growing, and it will outlast the Pi and other such products.

How many schools are taking part with 6502.org? That is my point. Things would have to change for them to do so.

[whartung]

Seems to me that the Micro KIM-1 kit meets pretty much all of these requirements.

http://www.brielcomputers.com/wordpress/?cat=24

It's essentially a KIM-1 clone that you can assemble. Keypad and hex display. 5K RAM, single step function, RS-232 to replace the original TTY interface of the KIM. Needs a few more parts to enable the cassette interface, you can use the Paper Tape logic to save and load programs via a terminal emulator that can capture and send files (or simply cut and paste).

$109 + power supply.

Documentation seems OK, but there's no theory (at least in the PDF they link on the site). It also didn't seem to have a staged assembly procedure. If you've read Ciarcia's Z80 book, he builds the computer in stages, each to a testable goal to check your progress. These docs are more "stuff the components in this order" to go from bare board to completed project.

But that's something that could be worked on.

It is a compelling piece of kit, I think, around which a curriculum could be built.

I've contemplated getting one to assemble.

But I'd rather do one "from scratch", so that I know what all of the wires are for and what all of the components do.

I don't even contemplate it right now as I don't have the $$$ to get the tool set and components needed.

[BigEd]

Yes, the Micro-KIM is in this kind of niche too. But like the Compukit, which I started with, soldering it together is a task without much educational content - I had the schematics and gained physical familiarity with the various chips, but I suspect physically wiring a bus might be - might be - more valuable. The thing is, too much manual labour becomes tedious, and if you make a mistake you have to find and fix. Without someone to guide you, that could be dispiriting. Then again, it keeps the cost down. (See the Fignition too, for an example of a built-to-price minimal system with a helpful community)

Having the board in front of me and the schematics (or indeed, block diagram) to reference was valuable.

Those USB to serial adapters are excellent - Garth misses the point somewhat I think - you get a serial interface and your PC or laptop becomes a glass teletype. It's the simplest possible peripheral for that purpose and cheaper than QWERTY. (Hooking up a USB keyboard is a totally different problem - even a PS/2 keyboard is probably more difficult and that only gives you input.)

In order of increasing complexity then, you could have
1 - buttons and LEDs as a front panel
2 - keypad and LCD or seven-segment display
3 - serial-over-USB to a host machine
4 - onboard or PS/2 keyboard and VGA or RF out
(I'd go for 1 or 3 personally. The Fignition uses keypad and RF which illustrates the solution space.)

So, alkopop79, what's your current thinking? It's your project!

Cheers
Ed