So Many Replicas!
 

I compiled a short list of replica projects. Unlike Grant's remarkable kits, most of these are using new hardware to mimic the look and feel of the old times:

PDP-1 replica:
http://www.vintage.org/gallery.php?grouptag=PDP1REPLICA
PDP-8 replica:
http://www.vintage.org/gallery.php?grouptag=PDP-8
PDP-8 replica single board computer and front panel replica:
http://www.sparetimegizmos.com/Hardware/SBC6120-2.htm
Apollo Guidance Computer replica:
http://agcreplica.outel.org/
Apple 1, Kim 1 and Altair front panel replicas:
http://www.brielcomputers.com/
Apple 1 replica:
http://www.achatz.nl/catalog/product...&products_id=5
Apple 1 replica:
http://www.bytecollector.com/apple1/
Apple II motherboard replica:
http://www.willegal.net/appleii/appleii-first_page.htm
Apple II on an FPGA:
http://www.mirrow.com/FPGApple/
Altair emulator and front panel replica:
http://www.altair32.com/Altair32FrontPanel.htm
IMSAI front panel version 2:
http://www.imsai.net/products/imsai_series_two.htm
ZALTAIR Z-80 single board computer running CP/M http://www.zaltair.net/:
http://www.ez80sbc.com/index.htm

I'm sure there are many more out there.

The PDP-1 replica kit was going to be for sale by Sellam. He sells the front panel, switches, PCB, and some linux software that boots off of a CD.

The Apollo guidance computer is pretty cool.

I've never seen the Apple II motherboard replica until now! That's some nice work!

The Zaltair is inside of one of my cases. :) The Zaltair and the Altair 32 emulator front panel are pretty much the same thing, literally. I'm not sure when it will be finalized, but I have pledged to help make the rear panel of the Altair case compatible with an ATX motherboard and power supply.

Thanks for sharing those links!

Pseudo replicas
 

Most of the replicas mentioned are really pseudo replicas as they replicate original features with modern parts in more compact and simpler configurations. None of them can truly compete with Grant's work as being (almost) true replicas, although Vince Briel's Apple I and Micro-KIM boards come very close.

I'd love to have a real pdp-11/70 machine, but I can't afford the space or the electricity for a two rack, 240 VAC current sucking monster. So a small, single board pseudo replica would be acceptable if it had the same front panel and the same functionality.

Don't discount Mike Willegal's beautiful Apple II replica. Although it doesn't include a case or power supply, the board itself is almost indistinguishable from the original.

Geoff.

Full disclosure: the pictures of the original Rev 0 board on Mike's site are of my machine :).

I have a real Apple IIe I bought via a Usenet ad many years ago. Alas, it no longer boots and only displays gibberish on the screen. I never used it much, so it looks like I never get around to diagnosing and repairing it.

My own little project is breadboarding a W65C02S system with a fully functional front panel. Unlike Grant's projects, it isn't a replica of anything; rather, it's just plain retrocomputing. The W65C02S is a fixed up version of the original 6502; it's implemented in CMOS and can be run or stepped at any speed. Well, any speed below 14 MHz. My idea for the front panel is to have it be run by an Atmel AVR (only $14) and include full bus display and 6502 register access. But unlike any front panel machine I've ever seen, it will have a serial link (supplied by the AVR) that will provide a complete remote access facility. Each switch on the panel will be implemented with a momentary contact and so can be operated in "soft mode". The panel could even have an IR photodiode so I can use a handheld IR remote control.

A program load can originate with a typical computer connected to a USB/RS-232 converter that connects to the front panel AVR. At 19,200 bps the typical computer can tell the AVR to load 56 KB of RAM in about 30 seconds with a nice light show as a side benefit.

My idea is to use 64 KB of static RAM but map the upper 8 KB of the memory space to a boot PROM and various I/O chips. (The 6502 uses memory mapped I/O exclusively.) I'm also considering adding a second AVR accessible by the 6502 for I/O, timers, etc. This AVR could also be used to implement a TCP/IP stack and a physical Ethernet connection. Another idea is to have the second AVR run a compact flash card emulation of a disk subsystem.

What else? A non volatile real time clock and calendar chip. Maybe a speaker, and maybe several piezoelectric buzzers of different frequencies. Or an X10 control node. A cassette tape interface, maybe one that controls my little solid state voice recorder and its sixty hours of indexed audio storage. A modem with caller ID support so the machine can hang up on annoying solicitors. Possibly an MP3 player, although I think this would be a bit taxing of the 6502 capabilities. And then there's the obvious possibility of a retro video and keyboard interface. A D/A converter that runs an analog front panel meter that indicates processor utilization. A papertape reader, if only I could find an inexpensive papertape punch.

It may be that the initial breadboard realization won't run at a full 14 MHz, so I'm considering a wirewrap version the second time around. (The great mystery: why do wirewrap chip sockets cost more than some of the chips themselves?) I could eventually produce a two board design (panel on one PCB, 6502 system on the other) although I might need a little help on this. A case and power supply shouldn't cost too much if appropriate off the shelf products can be had cheaply.

More kits....
 

I originally posted this on the Vintage Computer Fourm while asking if anyone was constructing kits. I copy it here for those interested:

Everyone:

Andrew Lynch e-mailed me links to Grant Stockly site (again, outstanding work Grant): www.altairkit.com, Tod Fischer's IMSAI Series 2, www.IMSAI.net, and Mr. Roganti S-100 board reproductions at http://ragooman.home.comcast.net/rog_s-100.html. Thanks again Andrew.

It hasn't been a easy search for hardware and peripherals and hence why I ask. A google search hasn't been very fruitful and if it wasn't for an a thread in Applefritter two years ago, I wouldn't have know about Vince Briel's Replica 1 (which I enjoyed so much I sold it and built another one). Has anyone seen or heard of any European computer kits? BBC Micro or Apricot?

So... with the collective subject authority we have in this forum, please post what you know about available kits here.

COSMAC Elf Reproduction: www.sparetimegizmos.com

Apple-1 and KIM-1 Replicas: http://www.brielcomputers.com/

ZX-81 Kits from England: www.ZX81kit.com

ZX-81 Kits in New York: www.ZebraSystems.com

CF Drives for Apple-1 and Apple II: www.dreher.net

A-One, Apple-1 kit with accessories, Europe: www.achatz.nl

Semi-Virtual Disk (SVD) drive interface, various 80's computers: www.theSVD.com,
(appears as he has stopped shipping units but unclear if he is producing a newer version)

I have one of Vince's Micro-KIM boards and it works as advertised. I'll probably get one of his 32 KB expansion boards to go with it.

----

I'm continuing with my design for a 6502 (W65C02S specifically) system with a classical front panel. I'd probably not bother with this if there were some way of getting a new (kit or built) IMSAI 8080 or Altair 8800b. At the moment I'm waiting for parts. Eventually I'll make the schematic and BOM available.

David Griffith has a fairly nice IMSAI up on eBay right now. No disks, though.

Geoff.

What about doing something with the 65C816? There is an affordable 512kbyte SRAM DIP too. You could use the $6 flash chips I used with the 680 to give it 512kbyte of eeprom style flash.

Yes, the W65C816S was considered. And it still could be used in a modified design. But it just doesn't seem sufficiently retro!

I've been thinking about building an SBC out of all DIP parts that could run linux. Using a 64pin 68000, 32pin 512kb SRAM, etc. An inexpensive kit that has something in common with modern tools and software would be fun. A true "modern" computer kit. I wanted to limit it to DIP/PLCC/through hole parts since even coarse pitch SOIC parts are hard for some people.

I have had this idea since before the Altair kit ever came. Now I'm thinking it would be neat to have a 68000 and some memory on an S-100 card so that the Altair can run linux. :)

I know of at least one case where uCLinux has run on a 68000, but the FPGA company who built it for their core doesn't want to share the kernel patches.

I only suggest the 65816 because there are already a few 6502 kits out there.

I've been thinking about selling my Altair 8800 kit prototype. Its taking up space and I don't think I need to keep it for "memories". If you or anyone are interested, let me know.

One thing I would like to make in the future is a 6502 CPU board that is 100% compatible with the original Altair. This would allow a modified version of the Wozniak Apple 1 monitor to run on the Altair. If you prototyped your S-100 card in an Altair you would help accelerate me to that goal.

I would probably have a 6502 CPU chip, buffer chips, and a spot for a medium sized FLASH chip AND a 1702. The woz monitor would be
modified to use the 2SIO.

There IS a version of "CP/M" for the 6502. It would also be neat to run that!

Too many ideas and not enough time! :(

This is the prototype that I would be interested in getting rid of...

http://www.altairkit.com/images/0610...rview_1554.jpg

68000 DIPs
 

It's Freescale that owns the 68000 IP, and they don't make DIPs of any kind. Old 68000 DIPs can be had, possibly salvaged from old Macs, Amigas, etc., for about US$20 each.

Anyway, a 68000 really needs a 68551 MMU to do Linux, and even that's a little clunky. I wouldn't try it with less than a 68030 as the base chip, and that was never made in DIP form.

S-100 and the 6502
 

The S-100 bus had too many problems physically, electronically, and logically. That's why it died out so quickly. Designed for the 8080, and not very well at that, it is a piling of kludge upon kludge to make the S-100 bus work for a 6502, 65856, or 68000.

Failure to share in this case is very likely illegal due to GPL conditions.

For uCLinux you wouldn't need an MMU. That means its just an application host and not a development environment, but could still be fun. Possibly more useful than an 8080???

PLCC versions of the 68LC000 are available from DigiKey, and they are 100% compatible with the old Macintoshes. I even built a carrier board and ran a Macintosh on one.

The DIP versions are available from second source places like Unicorn.

http://www.unicornelectronics.com/IC/68000.html

I buy ALL of the ICs for my kits from Unicorn.

I would rather buy an old DIP 68000 than a new PLCC 68LC000...because a 68 pin DIP is just TOO cool... :D

If they sell it they are required to make the soruce code available to their customers, but as far as I could tell they just did a report and listed the fact their core booted and ran a full blown uCLinux as a test of their cores compatibility.

Here is the page.

http://www.dcd.pl/ashow.php?page=uclinux_d68000

If you can help them to give us the code, that would be great! :)

Its already enough of a stretch to throw a 68000 on an S-100 card and say "Altair runs Linux". :) It has to at least get power from the S-100 bus and make the lights blink!

Imagine it, with Linux the Altair would finally be Slashdot worthy. ;)

I think it may be a 64 pin DIP, which is still better than a 68 pin PLCC. :)

The Unicorn Electronics link is inviting, but I checked and all of their catalog pages are at least a year old. So maybe the stock counts are not really up to date.

If someone uses GPL code but never releases the product, then they're in the clear.

Until the advent of the fairly well specified PCI bus, all earlier microcomputer buses were either too slow, too big (VME), proprietary (IBM micro-channel), or just plain crap (S-100, ISA). The only exception is the MIT NuBus that was used in the early slot days of Macintosh computers, and even that bus was a little slow (33 MHz).

It would be possible to run Unix/Linux on a 4 KB 8080 with enough mass storage; simply write a small interpreter that's powerful enough to run an interpreter for a 68K or some other process, and have that interpreter run the OS and applications. Maybe it would be better to have 16 KB 8080 and only run one interpreter. I've heard of one Mac enthusiast who ran a PowerPC interpreter on an ancient 68K Mac and it boots OS/X. It runs a couple hundred times slower than the real thing.

The thought of Vista on an IBM PC or Mac OS X on a Lisa has crossed my mind. :)

Unicorn maintains a stock of any item on their website. They will sell you anything on there at that price. If they don't have it, they get it. They then will sell you one part.

They do not list 8008 microprocessors or Kenbak memory, but I have bought both. If you want a part they do not stock, then you have to buy a minimum of 25-100 parts.

An early version of the Mac OS ran on a lightly modified Lisa; the product was called an Macintosh XL and included a hardware mod that changed the original oblong pixels to square pixels. Its 68000 ran at a blistering 5 MHz.

If you're serious about running a 68000 on an Altair bus, you might want to go with the somewhat castrated eight bit version of the chip, the 68008. Hard to find, though. Its address bus was cut from 24 to 20 bits while the data bus was cut in half from the original eight bits. I have a circuit for this, and let me tell you that you're going to have to hijack a few unused S-100 lines for getting the 68000 bus grant and DTACK scheme up and running. Also (and I may be wrong here, it's been a while), the 68008 was an NMOS design and had a minimum operating frequency requirement. Hopefully it is less than the 2 MHz upper limit of the rest of a typical S-1000 system.

I always had a fondness for the 68000 family, even though it really wasn't fully debugged until the 68030 (Integrated MMU) plus 68882 (FPU) combination. You see, I had learned Unix on a pdp-11/70 and the 68000 family is the logical 32 bit extension of the good parts of the pdp-11 CPU design. (Compare with the VAX-11 CPU that suffered a design by committee approach.)

http://vt100.net/mirror/harte/Cromem...ual%201983.pdf

There are a few 68000 boards out there for the S-100 bus.

I would probably bring the excess memory addresses and specific processor timing out to a header to connect to a memory or ROM board.

The RAM seen to the MITS bus would either be a window or simply 0-64k. I/O addresses would have to be decoded some how...

Its not too much of a serious project, but its nice to think about. :)

Cromemco was probably the last S-100 hold out. I remember them having significant advertising space in Byte magazine right up to the end of the S-100 era. The were some high hopes, mostly the hope that microcomputer class hardware could be sold at near minicomputer class prices. But it was not to be. And it wasn't too many more years until most minicomputer makers themselves were just footnotes in the bigger history.

More on buses and cards
 

Here's an interesting manufacturer site: http://www.vectorelect.com/

They have hardware for sale for just about every bus for prototyping and other purposes. They even have three different S-100 prototyping boards. Alas, I have no idea about price and stocking details.

S-100 cards: http://www.vectorelect.com/Product/P...d/PBSTDBUS.htm

EuroCard selections: http://www.vectorelect.com/Product/Plugbord/PB3U.htm

These use a standard 96 pin and socket arrangement, and these seem to be popular among the European homebrew crowd. I believe this format was also used for Apple's NuBus implementation.

replicas
 

Hi Grant;
I have a cromeco s-100 card that has both a Z-80 and a 68000 on it, and it is an s-100 card...
THANKS Marty