_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Perhaps we could post up a minimal 6502 system which can be breadboarded, with a list of parts and links to where to buy them. (Probably we could come up with more than one such design!)

There's a little bit of a bootstrapping problem, in needing something in ROM to get started - unless you design in a switch-style front panel to single-step your bootstrap in. We could help there by selling pre-programmed ROMs which contain a bootloader (for a particular interface at a particular address...)

A similar problem exists with FPGAs. enso's willingness to sell pre-programmed devices helps a lot there. I don't know if his present design has enough I/O for this idea, but you can imagine a pre-programmed FPGA-based board could expose a databus and some address pins and enable signals, if not a full address bus.

I think the thing is to consider what you'd do if you had absolutely nothing: the first thing you need is a shopping list, and we can help with that.

We need to ease up on all the second-order advice about how to build industrial-strength high-speed systems, I think. 1 MHz and breadboarded, mean time to failure at least 10 minutes.

Cheers
Ed

Hi Garth,

First time caller here - I've been lurking for a while and enjoy all of the great info you and others have posted.

I don't think young people today are really all that different than in years past. Humans have been running the same firmware for 10,000 years. I think the difference is that adults no longer encourage these kinds of things. Some are worried that little Johnny might burn himself with a soldering iron or that your neighbors will think you're a pedophile if there's teenage boys visiting your house. Most adults don't tinker any more - manufacturers have done a good job convincing us that making things is "hard" and there is little practical value in repairing things when replacement cheap imported goods are so readily available.

However, there are still those among us, old and young, who have "the knack". Here is something I was inspired to write about a friend of the family recently: http://fuzzythinking.com/these-kids-today/

As for myself, I was a Commodore kid in the 1980s - I learned how to program from the display models in K-Mart. Eventually my parents bought me one for Christmas. With their continued support I was able to grow this interest into a career as a software developer. One thing that also helped was joining a local computer club, where I could meet other kids & adults with the same interests and learn from them.

As far as solutions, maybe a good thing for us to do would be to volunteer with organizations that help kids - Boy Scout troops, school tutoring, etc. Most of the kids won't be interested in what we do, but a few will. Once you find them, spend lots of time with them just explaining what you do and answering simple questions. Also, direct them to books and sites like this where they can find more info. I learned a heck of a lot from random books Dad would bring home from IBM.

Maker Faires and Hamfests are also magnets for techie-kids.

Thanks,
-JP

Great story - and welcome. I'd read somewhere that the TI calculator being Z80 has led to some amount of homebrew programming. It can only be a good thing.
Cheers
Ed

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I have to say that I am deeply disappointed. I remember dreaming of a world where everyone has a computer. The world we live in is certainly not what I envisioned .

30 years ago having a computer was empowering and stimulating. Today giving a kid a computer pretty much guarantees 8 hours of gaming and idiotic online blabber. Not to mention the phones.

Being stuck into a big barn and feeding from a trough is hardly progress. Even things like Arduinos and MakerFaires and open source are a combination of awful self-promotion and corporate greed. What happened to free thinking?

It is hard to be hopeful, but I think it's important to explain to kids that spending time with a closed platform or a plastic box with a closed OS is not an answer to anything, just a way for someone to gain market share and make a lot of money by going public.

My son is joining a Lego Robotics team. Another weird closed platform (the damned controller is hundreds of dollars???), but it's something; I am hoping it will stir up some interest in doing strange things with computers.

_________________
In theory, there is no difference between theory and practice. In practice, there is. ...Jan van de Snepscheut

Agree on pretty much everything said... getting kids interested in not a simple task, especially with all of the existing distractions already around them, which includes the internet. What I think is a bigger issue is one of cost. Just pricing out a handful of chips to make something useful quickly adds up. You can easily be in over $100 by time you get everything needed. Now add the tedious and error-prone task to breadboard it, then have some way to program an EEPROM and you've got an expensive project that has higher odds of failing than getting a newer smartphone.

Arduino is an option and has a lot of support, but it' boring in my view... And they look cheesy as does the overall dev setup. I think Zilog does a much more pro setup for one of their Z8 Encore dev kits which includes a pretty full dev/debug IDE and USB programming/debugging interface. However, it's also a micro controller single chip device, not something like a 6502 system.

Wish I had a better answer... Even my minimal system was expensive to put together using a PCB and socketed chips, but it's a safer build option should one manage to pop a chip while experimenting.

_________________
Regards, KM
https://github.com/floobydust

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I think both Lego Technic and Raspberry Pi are useful educational tools: if a family can't buy one, maybe the school will. After-school activities can expand a lot on what the regular class does.

As noted a day or two ago, the Raspberry Pi has the interesting characteristic of being a Linux computer running at 3.3 Volts, with serial I/O and a few general purpose I/Os. It fits well with the demise of 5V design. Having an actual standard serial port removes a lot of difficulties that come with USB adapters.

Cheers
Ed

The Raspberry Pi's primary driver is that it's straight forward to get lights to blink or a motor to spin using a scripting language. Making the software development cycle fast for young people.

No downloading, no prom burning, no compiling. Follow some instructions, and you can get the light to blink interactively from a Python listener. This is much more accessible than a longer, lower level development cycle around C or assembly.

Next, it becomes a task of getting Interesting Hardware that lets folks do Interesting Things quickly.

There's little reason to start folks at the low level fundamentals any more when they can be successful, quickly, with the pre-tested, higher level abstractions.

With SoC's, and the modern micro controllers, fussing with things like address decode logic and interface chips is passé. Why would someone waste time wiring all that stuff up, when this thing over here does it all on a single chip? Chip, usb port, and power (which may come from the USB port). Add on a LED and a switch to make blinky things you can turn on an off.

Not having tried it, but pretty sure you don't need a scope for basic bootstrapping with a modern micro controller, even if you don't get a pre-made board.

What about a 4 chip solution: CPU, RAM chip, Flash ROM chip, and CPLD (or some other thing) that does all the rest -- parallel I/O pins, serial I/O, decoding, etc. Make it brick proof (set switch, starts up in "download from serial port and save to Flash ROM"). Common, off the shelf components that don't have to be salvaged off of eBay.

At least the primary components are there to see: CPU, RAM, ROM, and Magic Happens Here chip vs a single ALL the Magic Happens Here chip.

Think you'd be hard pressed to find someone wanting to solve Karnaugh maps to get their software loaded and a LED to blink. Not today. Not anymore.

Sorry for beating my own drum here, but I think my Propeddle project would seem to fill a need here. Once it's done anyway

For those who haven't seen it before: I combined a 65C02 and a 128K SRAM chip with a Parallax Propeller (based on a project that did the same thing in a slightly different way). The Propeller bitbangs the clock and other signals to the 65C02 and can also emulate hardware such as ROM, video adapters, serial communications etc.

The 65C02's view of the world is determined by software running on the Propeller; If you want to map a video screen into the 6502 memory map: no problem. Emulating a VIA, PIA or CIA should also be fairly trivial. ROM is a matter of either mapping a Propeller memory area into the 6502 memory map, or copying it to the SRAM chip at system startup and allowing only read access. I'll also make it possible to connect real hardware via an extension bus.

As I mentioned, implementing ROM is very easy; if you would use my hardware to emulate some old system like the Commodore PET, you could probably just use a ROM dump file and use it directly. But I'm thinking, as proof of concept for those who would like to try to "bring up" their own system, I should provide a program that starts out with a blank ROM, and an on-board cross-assembler. You'll start out with a non-functional 6502 computer but you'll be able to "break out" from the keyboard, and "toggle in" a program from the Propeller, which will be executed from what the 6502 thinks is ROM. It should be just as much fun as bringing up your own system but without the "cheating" part of doing the cross-assembling on a PC, and without the tedious parts of waiting to flash and erase an [E]EPROM or download a ROM image like you would do with an Arduino or something similar: it will be "flashed" as you write it. A blank-slate computer with a self-contained development system. All you need is the hardware, a monitor or TV, and a keyboard.

Sound cool?

Of course this still needs a lot of work. First of all, I have to figure out why I can't run the main bitbang-loop at 1MHz yet (it runs one step slower, at 0.99MHz but something in the timing goes wrong at the full 1MHz speed, and I have to hook up my logic analyzer to figure out what it is). Then I need to re-design a circuit board: the previous PCB designs were based on the Propeller Platform but that's pretty much non-existent nowadays so I have to design an all-in-one PCB and I just haven't gotten that done yet.

Meanwhile, I did figure out that the Parallax Propeller Project Board (#32810) is big enough to do a wire-wrap version for those who might be too impatient to wait for the next revision of the PCB. Here's the first ever picture of a possible wire-wrap version of the Propeddle. This is a proof-of-concept, it doesn't work yet. And wire-wrap is not where I want to go: I want to make a simple PCB with through-hole parts only that has everything on board; it will probably be about twice as big as this.



EDIT: the wire-wrap version should be about $50, not including wire-wrapping and soldering tools and wire, and not including a PS/2 keyboard or monitor.

===Jac

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I think y'all are missing my point.

Computers, especially today, are a means to an end, not a means in itself. The interest is in the blinking light, not the decoding logic that enables the 6522 that lets the 6502 code blink the light.

The code is interesting, the light is interesting. Everything in between? Not so much. Especially not at first. Especially today.

Lego Mindstorms teaches nothing about computers or computer architecture or design or interfacing or even electronics. It teaches programming, motor control, sensors, gearing and mechanics.

This is why the one chip wonders are interesting. Why they took off. Artists are a core component of the Arduino marketplace, because they can easily Do Stuff with them. They don't understand them, and they don't need or desire to understand them at a low level. They just need them to work. To turn on their lights, and spin their spinners when the switch is flicked or the light beam broken or the sun goes down or whatever.

The modern chips are opaque. Power goes in, code gets downloaded, leds, motor controller modules, and relay modules are wired straight the the output pins, and magic happens. You barely have to know about polarity to get things working. Power, drivers, impedance, triacs, SCRs, etc.? Nope. None of that. That's handled in the little modules you plug in and wire together.

Their self contained nature makes them difficult to disassemble if you might be curious more about what's going on.

The 4 Chip 6502 doesn't necessarily suffer from that. I love the Propeller board idea. Yes, the Propeller chip is a HUGE crutch from a component and circuitry point of view. It's a black box of magic, but it's really enabling and empowering to the system as a whole. It potentially makes the 6502 Really Useful for a Lot of Things, "right out of the box". Distinctly different from of the more complicated starter 6502 systems. Add some prototyping area to the board, and jumpers to let the user add decoding logic later, along with the header for the I/O, of course.

The criticism that the Raspberry Pi is too high level is not properly answered by dumping the user in to the depths of discrete logic, two phase clocks, chip families, and bypass capacitors. Now you're too low level, and the level of understanding necessary to get the light to blink is a huge step up in complexity compared to the other solutions. So you need to have a middle ground. Something perhaps more knee deep than 7 foot deep.

The students can always find out more, if they find they are so inclined. They're not learning anything wrong by starting at a higher level and the being able to go deeper and get more detail. But requiring them to "know it all" just to get started is simply to much.

Indeed. Note that Mindstorms devices are controlled by I2C - there is scope for building your own peripherals. It's a self-contained system that helps with learning programming, but it doesn't stop there. The new revision runs Linux and is open source. Of course you only go the Mindstorms route if you want to do robots or software controlled mechanisms, but it is very motivating to have such real-world peripherals. Much more satisfying than a blinking LED and more tangible than a keyboard and display. It is expensive.

Many of us here like to break everything down to the atomic level, but there are plenty of stopping off points as you drill down from a black-box tablet to a big mess of wires.

It's clear that the 4-chip computer, or the 6-chip computer, has broken everything down to the traditional essentials. There remains the difficulty of talking to it. If you put a serial port on it, you need to have something which talks to the serial port.

This is where the Raspberry Pi is a good fit: in this case, don't think of it as part of the experiment, but as part of the infrastructure. The family PC, if there is one, probably doesn't have a serial port, and the family might not be sanguine about hooking it up to a breadboard. Solution A is a USB-serial converter, and solution B is a Raspberry Pi. The Pi just has to run any glass tty utility, of which at least 3 are standard. See http://elinux.org/RPi_Serial_Connection

As a completely separate question, instead of a 4-chip or 6-chip computer, you can black-box some of the complexity. If you do expose an 8 bit databus and a 16-address bus, and a clock and a R/W signal, then there's plenty of room for experiments like hooking up a CIA or building a parallel port from 74 series.

What's the black box? It could be a PCB with a 4-chip computer on it, it could be Propeller-based, whether as a dev board or on a breadboard, it could be a preprogrammed FPGA on a PCB, it could be a discrete 6502 with a CPLD or a GAL.

Cheers
Ed

Hi guys,

Also why not check out my DIY 8-bit Computer FIGnition? It's been designed for exactly this purpose: a (3-chip) computer simple enough to be built, coded and understood.

http://www.fignition.co.uk

FIGnition was launched properly on April 28, 2011 via an article on the BBC.. 3 days later the raspberry pi team contacted the Beeb to pre-announce (by 10 months) their concept!

Since 2011 I've sold over 1000 kits and they're now being distributed via RS components. It's nothing like R-pi, sales, but it's quite good for a kit computer

Note: although FIGnition uses an AVR, it's not an arduino. Instead it runs Forth programs at near 8-bit machine code speed from 8Kb of external serial RAM, programs you can type in directly via it's keypad. It has audio and outputs to PAL/NTSC, supporting 25x24 text or a proper 160x160 bitmapped frame buffer with a high-speed xor blitter.

FIGnition's firmware is fully open-sourced and is OSH compliant, you can even build it on stripboard from complete scratch!

Check out my posts on the blog OneWeekWonder for my thoughts on why FIGnition is a good tool for end-to-end computer learning or William Marshall's Design Spark review.

Sorry about the plug and the fact that FIGnition isn't 6502 based. Actually I frequently lurk on this site because it's enjoyable to read about all your home-brew computer goodness

-cheers!