Actually, I wouldn't have mentioned the GALs but for the fact they're cheaper than most "ttl-like" CMOS 20-pin parts of comparable speed and drive. What's more, you can build the parts you lack ouf of them, for the most part, i.e. HC 573, 540, 541, etc, i.e. any part with inputs on the low-numbered side and outputs on the high-numbered side. Because you don't have to share gates with circuits on opposite ends of your PCB, they're easier to use in your board layout, AND they're WAY faster than any LS or even S-TTL, yet consume less power. They're also more noise-immune than most bipolar logic, and drive typically 24 mA per output, whether sinking or sourcing.

Unlike most of the people who've written to this forum, I've done all these things many times, and am making the recommendation I make because I think it will expose more people to useful techniquest and technology, while costing less. I particularly come out in opposition to the use of LSI's for parallel I/O (and for primitive serial I/O as in the 6522, simply because the device is too slow and too weak to be of use. Its functions can be replaced with no loss of board space because its outputs require buffering, without which the LSI is useless.

Moreover, I have extensive experience in system design, using both the 6502, and some really fast processors, partly of my own manufacture (via programmable devices) and mostly using devices from other sources.

As for cheap and available, I don't think you'll find a 6532 to be either, particularly if you want one to run with a 14 MHz CPU. I doubt you can find a good reason to expend time and effort on a large costly system running at 4 MHz when you could build a much more capable system which is smaller and MUCH less costly running at >12 MHz. In fact, I find that if I can get a processor to run at 14 MHz, there's no reason to run it at 13.999 unless there exists a special synergy between that frequency and some specific task I have to perform.

Just my 2 cents' worth . . .

Uli

We've had this discussion already, before you joined the forum.

In short, you're being way too 'practical'. If the point was to build a high-performance board, then we might want to start with one of the many modern multi-function chips, and take advantage of lots of other modern devices and technologies.

But that's not the point (I feel). The project is something that is cheap and easy to understand for beginners, and 'nostalgic' for people like Mike and me and you, who have been fooling with these things since the late 70s. I still have my original 8K PET and KIM-1 that I bought in about 1979. That's what got me interested in all this stuff. If I needed a high-performance micro-controller board, I'd go shopping for one that's already built, and expand on it from there. This forum (I think) isn't in the business of designing something like that.

What would probably be most useful to this group is if you would share your own existing designs with us - lots of us would be interested in seeing them - just like the many fun projects that are already presented on 6502.org. Those projects are low-tech, fun, and educational.

Also, we've already discussed a possible "2nd project" that would use a 65SC134 - it would be closer to what you're thinking about.

Pete

Frankly, I don't see anything that could be done with an old and slow 6502 as being particularly practical, enducational, or useful, though you never know about the usefulness. Today you can buy a part which has the I/O and RAM and ROM on the chip, hence you don't need to design much hardware. You just hook up a crystal and a serial port if you need it, and program its EEPROM.

What would probably be enlightening to the average hobbyist is a CPU running at greater than 12 MHz. That requires thinking out the problems of making things work when you have tight margins. With HC logic, 4-5 MHz means you have 200-300 percent margin. At more than twice that speed, you have much less and can't make any time allocation errors. It's not difficult, but it does require you understand how things work and how to allocate time. Of course, you still have to design the logic.

Someday soon I'm going to knock out a little 2.75x3.5-inch board to with a CPU and a 64K sram, and see how much I can bit bang. I've considered writing a hard disk controller chip code set for the Scenix SX. It runs at 100 MIPS, which isn't bad for a $5 part. (see <www.scenix.com>

I have a bunch of 6502 hardware I've made/bought/modified over the years, and I want a pin-compatible that runs at high speed. I ran 4-5 MHz back in '79. Now I want REAL speed. With a roll-yer-own, I believe I can push the model to around 50-75 MHz. That would be both interesting and useful.

See what I mean?

Uli

I can see where you're coming from about wanting a really modern 6502 board. Frankly, my mouth waters at the thought of having a powerful little board like you describe. I think that we should definitely explore a more advanced board. I'd love to have it for small robot projects. Why not start another thread for it, "Next Generation Board"? I'm sure many people would love to talk about it. I would.

However, this current project came about from requests from many people asking specifically for what we're designing. There are quite a number of us who have spent the past few months discussing this thing into its current form and we're looking forward to seeing it through to a finished product. I've even gotten e-mails from people I've never seen in the forum saying things like "save two for me". So there's definitely interest in it.

Let's stop arguing about this. We're talking about apples and oranges, radically different animals, and we can have both.

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

I've been monitoring the forum for a while, and thought I would put down my thoughts.

The hi-tech 6502 based supercomputer board sounds great, but it is way over my head, and probably my requirements. Having come down the VIC-20, C64, BBC route I am happiest in my

known territory - plain 6502, 6522 64K RAM etc.

As well as that, I have loads of old C64s, and vintage salvage 65XX chips lying around in my garage, which if I can do something useful with, would stop my wife moaning about 'all that junk in the way'.

I started with modifying an old VIC-20 for a robot, which proved to be too unwieldy, so I was inspired by Chris Ward to build my own computer. At that time 6502.org came up with the idea for a PCB of its own, so I deferred to it, and I am definately in the 'save two for me' camp.

Am I the only one who feels this? If so I'll shutup !

Mad Wombat

>As well as that, I have loads of old C64s, and vintage >salvage 65XX chips lying around in my garage,

If you want to dispose of one or two C64's cheap (or trade),
let me know. I'm at saipan59@uswest.net .

>Am I the only one who feels this? If so I'll shutup !

Not at all!
I would guess that your situation is 'typical'.

It's too early to tell what the 'higher-performance' project will turn into. For me, the requirement is not high performance, but rather small size and low power. Very, very few of us would get any 'real value' out of a 14 Mhz 6502 based on speed alone - it will have to have other useful features.
Apps that *do* require high speed, such as talking to a disk drive directly, are complicated projects. How many of us would actually follow through on such a project? I'm guessing 1 or 2 folks at most.
If we can get a certain amount of speed for 'free' from the 65SC134, then great, but it's not what I would call a "P-0" requirement.

Pete

Hi guys,

In trying to define the remain pins on the second expansion connector of our project I have run into a problem. A 40-pin connector will not allow us to include all of these signals and also the future 65816 signals. Pete has suggested using a 50-pinner for the second connector instead which should give us enough room. I think this is a good idea. Does anyone have any problems with this? Any other comments?

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

Hello everyone,

An anonymous user left me this e-mail which has some good points and ideas. Feel free to respond to them, he does monitor the forum. He's also suggested the first name for our printed circuit board project: "Iteration One".

(Here's his letter:)

PURPOSE OF AN SBC - MY OPINIONS

There is no way that an SBC board will EVER compete with the big guns like the INTEL and DEC boards. If this is true, and it is my opinion that it is, then why spend our valuable time designing, debugging, and putting into use, an SBC? The answer can't be money because an INTEL board can be purchased
for under $100.00! A home designed and built SBC will exceed that dollar amount with little effort! Even a BASIC STAMP or DOMINO series system will exceed $100.00 when all is added up! So, then, why an SBC?

As we design our SBC's should we try to emulate the big guns by incorporating memory managers with large memory arrays, sophisticated memory schemes, I/O processors, complex hard drive systems, and various other "resources"? If this is the case, perhaps we should just purchase an INTEL board to save our valuable time. Once again, I can not, nor do I want to,
compete with INTEL or MS systems!

There must be a deeper reason for an SBC!

It is my opinion that the answer is "CONTROL"!

I AM in control of my home built SBC. When it crashes it is MY fault and not an operating system "glitch". If my project requires sharing system resources, I AM in control of that sharing! If I WANT to hang some I/O at $A600, I WILL design the decoding to allow the I/O to appear at $A600. If I
want a speech synthesizer register at $8000, then I will put one there. If I want a second RS-232 port with full handshaking to communicate commands to ham radio repeater, then I will program one!

I have tried to do these things with an WINTEL system. I have found that I am held back by their design. These systems are not designed with the above in mind. Oh, sure, I can purchase a I/O board for my WINTEL system to assist in my designs, but if I do, I am once again at the mercy of the I/O board design. I have lost a certain amount of CONTROL!

I am not criticizing WINTEL systems! They are just not designed with the above in mind.

The question can be asked, "Where is the power of a computing system?" Some would answer "the CPU", and some would answer "the memory". It is true that these two units are NOTHING without the other! But the real power of a computing system comes from its "program". The more CONTROL I have of the
PROGRAM the better system I can design.

I can use a $3000.00 laptop to control my radio systems or I can use a $100.00 controller that was designed and tested using an SBC. For me, the answer is simple.

As for the design of the www.6502.org "ITERATION ONE" 6502 SBC I would recommend the following.

1. 64K of on board memory directly accessible to the CPU. The first iteration should not contain bank swapping. Once the design is solid, bank swapping can be added, perhaps with ITERATION ONE as an option, but, then, maybe ITERATION TWO should contain this.

2. Address decoding down to the 2K level (32 lines). One 1K block decoded down to the 32 byte level. That sounds like a lot of decoding. But when one is controlling something you do not want to run out of decoding lines. These lines should be accessible on an applications connector.

3. No less than 32 lines of user I/O and be available on a user applications connector.

4. Hardware debugging that can be turn off and on via software and via hardware.

5. At least one and possibly two serial communications ports. Video out is not needed at this time. System communications can be via one of these ports to a terminal program on a WINTEL system. If one is savvy, serial communications need not require a chip but can be handled entirely in software. (versatile and easy to manipulate.)

6. Clock speed should be as fast as is feasible but not so fast as to compromise a simple design. Using a 1Mhz 6502 I can refresh an entire 80 character LCD display in 4ms. Is that too slow? (I am not recommending a 1Mhz system)

7. INPUT and OUTPUT should be vectored. This makes changing the I/O device simple.

These are just my opionins and are not cast in stone. I believe the goal is a simple and easy to use system, especially in ITERATION ONE. ITERATION TWO can get more complex as can ITERATION THREE...

(end)

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

Hmmm, I guess I don't follow:
On the one hand, he suggests that Iteration 1 should not be too complicated (I agree!).
Then, he suggests a bunch of additions to the design...

My comments are below each section below:

1. 64K of on board memory directly accessible to the CPU. The first iteration should not contain bank swapping. Once the design is solid, bank swapping can be added, perhaps with ITERATION ONE as an option, but, then, maybe ITERATION TWO should contain this.

Bank swapping was not being considered, was it?

2. Address decoding down to the 2K level (32 lines). One 1K block decoded down to the 32 byte level. That sounds like a lot of decoding. But when one is controlling something you do not want to run out of decoding lines. These lines should be accessible on an applications connector.

I agree that it "would be nice", but it's not necessary. It's easy to expand the decoding off-board.

3. No less than 32 lines of user I/O and be available on a user applications connector.

Already got that!

4. Hardware debugging that can be turn off and on via software and via hardware.

Please elaborate. I don't think we want to design any software at this point - I'm thinking that software is up to the user (except for a basic monitor).

5. At least one and possibly two serial communications ports. Video out is not needed at this time.

One serial is already planned. Video is not.

6. Clock speed should be as fast as is feasible but not so fast as to compromise a simple design. Using a 1Mhz 6502 I can refresh an entire 80 character LCD display in 4ms. Is that too slow? (I am not recommending a 1Mhz system).

Yes, we're figuring that we'll support 4 Mhz.

7. INPUT and OUTPUT should be vectored. This makes changing the I/O device simple.

Please elaborate.

If it were all "up to me", I'd suggest that our design is too complicated already...

Pete

I want to apologize for being "cranky" in .39.

This person's suggestions, like those of everyone else, are valid.

But at some point, I think we should just say that we've gone "far enough" with the functionality, and move on to the implementation. We are not designing a for-profit product; rather it's just something to be fun and educational. If we include one person's idea for a "cool feature", then we are leaving out someone else's. I suggest that nearly everything beyond the basics is an off-board expansion.
What we've already got goes well beyond a KIM or SYM - I'd say that's good enough for a first project.

In short, I'd say that what's described on 6502.org is fine - let's move on!

Maybe another idea for a follow-on project would be to make a multi-function expansion board that plugs into this one: include things like a keyboard interface, a little LCD display, a storage interface of some sort, etc.

Pete

Does anyone know of confirmed sources of 65C22's that are spec'd at 4 Mhz or above?

I went searching on the web yesterday, and found very little. There's only a couple of small-quantity vendors that have them at all, it appears.

For example, Jameco has 65C22's, but they seem to be 2 Mhz (?) parts. They are VL65C22V's, which is by a maker that I can't remember at the moment (anyone know?).
California Micro Devices now owns the Rockwell line, right? They call it a G65SC22, and it's supposedly available up to 4 Mhz, but I didn't find anybody selling them (other than in 'quantity').

If we can't get 4 Mhz parts, then our 4 Mhz clock becomes somewhat irrelevant, eh?

Pete

Hi All,

Here's an update on the design project and new chips. I realize that the forum has been moving kind of slowly the past few days, but expect it to pick up quickly. The PCB design is rapidly nearing finalization so expect reports from me in the next few days so we can hammer out the remaining bits and get to work on the layout and prototyping.
Tomorrow I will be having a telephone conversation with my contacts at The Western Design Center. No more fooling around with e-mails. With any luck I will be able to get our questions answered about the availability of new chips, the possibility of us doing a bulk order for chips for our printed circuit board project, and related matters.

Thanks to everyone for their continued hard work and support.

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

Hi Gang,

I recently had a lengthy phone conversation with my contact at The Western Design Center. I'm sure everyone is anxious to hear the results. (drum roll please)

Overall I am very pleased with the outcome of the call. Many issues have been cleared up. I can write about all of it later (and in a more appropriate thread), but for now let me talk about the availability of chips for our printed circuit board project.

WDC is still supporting (and selling) its DIP package 65Cxx parts and has no intention of stopping, especially so now that Rockwell has dropped out of the game. It was expressed that they are not very thrilled about the idea of selling parts in quantities of one or two parts. However, I have made arrangements to make group orders, so getting parts for our project (and future ones) will not be a problem. I am sorry that I do not have the pricing information at this time, it will be worked out soon. They have assured me the prices will be very favorable.

I will start assembling orders for chips for our project, and anyone else that is interested, very soon.

WDC may also be helping to promote and supply information to the www.6502.org website. At very least we will have their datasheets which have been requested from time to time.

I'm sorry that I have been so busy lately. I will provide a more complete write up of my talk with WDC (all the "particulars") very soon. In the next few days I will try and post all the last details of the PCB Project so we can get a working schematic done. Now that we know we'll have the fast chips available (65C02 and 65C22 rated at least at 10 MHz in DIP or PLCC) it's full speed ahead.
Comments and questions are always welcome. Stay tuned, this is just starting to get good.

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

Hi All,

We have reached a point in the design project where there isn't that much left to talk about, we need to make a real schematic and then the artwork for a prototype.

At the request of several users we will be producing the schematic and artwork using the Cadsoft Eagle program. Eagle is available in a freeware version from this site:

http://www.cadsoftusa.com/freeware.htm

Douglas Beattie is very familiar with the software (he also has the commercial version) and has created a library for the 65XX components. He is going to send me the library and get me up to speed on using the Eagle program.

While you're waiting for this, take some time to download the program for yourself and get acquainted with it if you'd like to participate in this phase of the project. Once the design is completed in Eagle we will be porting it to other programs, including ExpressPCB.

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

Sorry I came to this so late in the game.
4_Mhz chips? yes 1,2,3,4 MHz.. 65C02, 65C22 and 65C51

G65SCxx manufactured by CMD, http://www.calmicro.com/

CMD Distributors:
Arrow Bell Components
2201 E. El Segundo Blvd.
El Segundo, CA 90245-4068
Tel (310) 563-2348
Fax (310) 563-2510
URL www.arrow.com

Jaco Electronics
145 Oser Avenue
Hauppauge, NY 11788
Tel (800) 541-9371
Tel (516) 273-5500
Fax (516) 273-5799
URL www.jacoelectronics.com

** ** ** ** ** ** ** ** ** ** ** **
I was primarily looking at 2_MHz parts,
which are a good medium ground to start...

Arrow.com search,
http://www.arrow.com/www/find_buy/search_from_home.html?p_company_name=ARROW%2FAMERICAS&p_country_code=US&p_supplier_name=allsupp&p_num_rows=10&p_base=&p_query_by=part_no&p_part_nd=G65SC02
note: "*"=available for 'order online'
G65SC02P-4 $5.69 *
G65SC02PI-1 $4.55
G65SC02PI-2 $5.15 * ............................<< this one

G65SC22P-1 $3.43 *
G65SC22P-2 $3.98
G65SC22P-3 $5.05 *
G65SC22P-4 $5.15
G65SC22PI-1 $3,86 *
G65SC22PI-2 $3.70 * -- cheaper, same speed......<< this one
G65SC22PI-4 $5.33 *

G65SC51P-1 $3.19 *
G65SC51P-2 $3.88 *
G65SC51P-4 $4.78 *
G65SC51PI-1 $2.99 * -- cheaper, same speed.
G65SC51PI-2 $3.52 * -- cheaper, same speed......<< this one
G65SC51PI-4 $5.35

** ** ** ** ** ** ** ** ** ** ** **
--
Douglas Beattie Jr. http://www2.whidbey.net/~beattidp/