OSI Challenger 4P Reproduction

I am making a full sized Challenger 4P reproduction, a 6502 based personal computer from the late 1970s.

Similar projects worth following
The Challenger 4P was a very capable personal computer introduced in 1979. Here are the specifications:

- 6502 microprocessor running at 2 MHz.
- A full 53-key ASCII-like keyboard with full upper/lower-case capability.
- NTSC video display 64x32 characters with scrolling and character editing capability.
- Cassette storage using the Kansas City audio cassette standard.
- Microsoft 8K BASIC in ROM.
- 2K Machine code monitor in ROM.
- A minimum of 8K RAM expandable on board to 32K.

Like all vintage machines of the era, especially ones not made by Apple, Commodore, etc., these machines have become rare and hard to find. That's my cue. For me reproduction is the sincerest form of flattery.


When we think about the dawn of the personal computer era, names like Apple, Commodore, Radio Shack, and Sinclair come to mind.  That's because these companies were the big winners in this new market, selling millions of personal computers to first time buyers. There were of course many other companies competing for market share at the same time, with capable and innovative products, that didn't quite make the cut. Ohio Scientific Inc. (OSI) was one of these companies. 

Formed in 1975, Ohio Scientific sold a wide range of computer products from the low end Model 500 single-board computer to high end Challenger III disk based business systems selling for $13,000 or more (in 1970s dollars). One of their most popular systems was the Challenger 1P, based on the upgraded single-board computer the Superboard II. The Challenger 1P was relatively inexpensive, selling for around $350 (US) in 1978. 


I started looking more closely at the Challenger 1P as a potential reproduction candidate when I was working on my MCM/70 reproduction.  It turns out that the keyboard that I'm going to use for the MCM/70 is an OSI reproduction keyboard from Dave at since they were very similar. I used Dave's Sol-20 reproduction keyboard for my Sol-20 Reproduction and it worked out great.  If you check out Dave's Unified Retrocomputer Keyboard Project on GitHub you will see that he also offers keyboards for Apple I, Apple II/II+, Generic ADM/3A teletypes, and most importantly to me OSI computers.  This is what OSI keyboards looked like.

This got me looking seriously at the Ohio Scientific line of computers. I liked what I saw. I was an Apple II nerd back in the early 80s so a reproduction project that was based on a 6502 processor was very appealing. 

So I have ordered an extra OSI keyboard PCB and stabilizer, OSI keycaps, and an encoder from Dave for my new Challenger 1P reproduction project. 


There is a very active and robust OSI community out there.  Here are a few links to get you started:

As a result, making a reproduction of many OSI computers is probably easier than it would be with other lesser known offerings from the era. For instance reproduction motherboard PCBs for many models including the Challenger 1P have been created based on 4000 DPI scans of the originals. I can see these on eBay for as little as $45.  I'm not sure how much parts would be. These boards have somewhere in the range of 10-30 TTL chips per board, and you'd probably want machined pin sockets, so parts could add up. While this is something I might consider doing if I was making this reproduction just for myself, I'm not sure this approach would be for everyone. For all of my reproductions I try to design them so that anyone could make one for themselves. When a project is finished I always create an Instructable with the step-by-step directions I used to make my reproduction. 

So my Challenger 1P reproduction will be based on a software emulator. While not as "pure" as a hardware implementation, this will greatly simplify the build for anyone attempting it.

A Challenger 1P Emulator

There is an existing emulator for OSI based computers. WinOSI is fantastic and can emulate many different OSI models and hardware options. If all you want to do is "kick the tires" on the various OSI computers, this is the way to go. Unfortunately because WinOSI is Windows only, it's not suitable for my purposes. I need an emulator that will run on a small single-board computer, probably a Raspberry Pi 4. Since there is no "portable" Challenger emulator I will have to create one. 

Based on what I learned creating my Sol-20 emulator, the first step towards creating a Challenger 1P emulator is to find an emulator for the CPU being used, in...

Read more »

  • Case Complete

    Michael Gardi12/15/2022 at 15:33 2 comments

    The case parts printed well.  There were a couple of places where the print pulled up a tiny little bit in the corners, despite a 10 brim, but not very noticeably.  I joined the pieces to together with M4 x 12 mm bolts and nuts.

    Then I fitted "skin" onto the frame. With the skin on I was able to adjust the keyboard to fit inside of the cutout and screw it down. To Steve's credit (the measurer) and Dave's (the keyboard maker) it fit perfectly.  All I had to do then is 3D print a logo and voila!

    Even though not necessary, for authenticity I thought about adding the 14 or so screws that can be seen on the top of the original's case, but in the end decided that I liked the cleaner look of what I have. 

    So this marks a milestone in my project. With the case complete I was able to change the primary photo for this project post to be my reproduction's, replacing the original's photo that I was using as a place holder. 

  • Building the Frame

    Michael Gardi12/12/2022 at 02:04 0 comments

    So I used the models based on Steve's measurements to create a DXF file to cut out the side panels.

    The inside panel has holes defined that will be used to join the inside panels to the cross-bars that will separate them at the correct distance apart, and there are also holes to connect the outside panels to the inside panels. Using 1/8 inch plywood I laser cut one piece exactly as above with the inside panel nested within the outside (thanks kwartzlab), then I cut three more inside panels only. 

    I glued pairs of  inside panels together so I had two 1/4 inch inside panels. I then used the outside cut panel as a template to layout the two outside panels on a nice plank of 1/2 inch walnut. These I cut out with skill and jig saws and used a belt sander to smooth the edges. 

    Finally I cut out some cross-pieces 364 mm long. One was 1/2 x 4 inch pine, plus two 5/8 inch square dowels. In addition I cut four 130 mm lengths of 1/2 inch dowel.

    So here is was I ended up with.

    I attached the cross-pieces to the insides with #6 - 1 inch wood screws making sure the the heads were flush with the panel.  The shorter 1/2 inch dowels were doubled up attached to the inside of the side panels at 20 mm below the slanted edge to support the keyboard. 

    Finally I used the outside template to properly align the inside pieces to the outside and joined them with 5/8 inch wood screws.

    Also notice that I cut a 10 x 10 mm notch into the back of the 1/2 x 4 inch cross-bar to accommodate a support beam on the case. 

    I've started printing the case pieces. I will have to print six separate pieces that are small enough to fit on my print bed. To facilitate the reconstruction I have added support beams underneath with holes so I can bolt the separate pieces back together again.

    All together there is about 50 hours of printing ahead of me.

  • Modelling the Case

    Michael Gardi12/08/2022 at 20:28 0 comments

    When you don't have an original to work from, it is often difficult to create a reproduction. There are usually lots of photos available online, but they don't give any indication of the case dimensions. Often I am helped by enthusiasts that own an original and are kind enough to share some measurements and photos with me. 

    Fortunately for me there is a robust Ohio Scientific community out there. A great place to start is Dave's OSI Page. This was an invaluable resource when I was creating the 1P (4P) emulator. For the case, Dave himself pointed me at a thread in the OSI Discussion Forum where a participant name Steve Gray posted all of the very measurements I was looking for.  Using Steve's measurements I modeled some case parts.

    I will probably laser cut the "inside" side pieces, CNC the "outside" side pieces, and 3D print the case skin.

  • Keyboard Assembly

    Michael Gardi11/03/2022 at 09:35 0 comments

    I've kind of been in a holding pattern waiting for the keyboard "kit" to arrive but no more as it was delivered last week. This is a classic retro-style ASCII keyboard, modeled after the ADM-3A keyboard, which can also be populated to fit perfectly in an Apple II/II+, or an Ohio Scientific Inc. computer. For details check out the Unified Retro Keyboard project.

    I was able to purchase the keyboard PCB,  a set of Fubata MD-4PCS keys, a stabilizer panel for the Fubata keys, and a set of reproduction OSI keycaps from Dave at I also ordered an OSI specific encoder which is on backorder.

    So I went ahead and assembled the keyboard.  I'm not going to detail the process here since it was mostly identical to the work I did on my Sol-20 Reproduction. For details check out:

    After adding the OSI Challenger style keycaps this is what I have. 

    Now I just need a case to put it in.

  • To 1P or 4P, That Is the Question

    Michael Gardi10/31/2022 at 17:27 0 comments

    Over the course of their six year run from 1975 to 1981, Ohio Scientific sold a wide range of computer products from the low end Model 500 single-board computer to high end Challenger III disk based business systems selling for $13,000 or more (in 1970s dollars).

    The lowest cost OSI offering was the  Superboard II, a compact and simple single-board computer. This entry-level system featured a keyboard, video display, audio cassette, BASIC-in-ROM, and up to 8K RAM for around $300. The Challenger 1P was basically a Superboard II with a case and power supply. Both were part of the Ohio Scientific Challenger 1 Series.

    In 1977 Ohio Scientific launched a number of systems in their Challenger 2 Series, the smallest of which was the Challenger 4P. Similar to the C-1P but on a standard OSI 48-pin bus, this system is composed of separate boards including a 6502 CPU board, a 540 video card, and a keyboard all mounted in a nice case. With a newer graphics card than older systems, featuring a whopping 2K of video RAM, a 64-by-32 character display was supported.

    My original intent was to do a Challenger 1P reproduction, however I am finding that the 32x32 character display (24,24 actual) is a bit limiting. Since a bare bones Challenger 4P system is similar to the 1P but with a 64x32 character display I feel this might be a better choice. To that end I have updated my emulator to support the 4P in addition to the 1P. You can switch back and forth by selecting the appropriate monitor rom at startup (cwmhigh.hex for the 4P or synmon.hex/cegmon.hex for the 1P). 

    There is another reason for choosing the Challenger 4P model for reproduction. The cases for both systems are quite similar but not quite identical.

    The Challenger 1P on the right has a base and sides made out of a single U shaped piece of metal. IMHO this does not compare to the beautiful wooden sides of the Challenger 4P. I actually have a nice piece of walnut that would work really well. So the case will be built as a 4P with walnut sides, but of course will be selectable for 1P or 4P emulation.

    All of the above changes have been posted to this projects GitHub repository.

    Updated 12/15/2022:  I have changed the project name and description to reflect the change in direction from a Challenger 1P to a Challenger 4P reproduction.

  • Better Load and Save Screens

    Michael Gardi08/30/2022 at 22:32 0 comments

    While the Python/tkinter based load and save panels work OK, the windowed popups are disconcerting and would not be suitable when my Challenger 1P reproduction is working with a full screen text only display.  So I created some "integrated" replacements. Now when you press F1 or F2, my emulator saves the current screen, "takes over" the display and keyboard, and uses them to present the user with Load and Save dialogs.  When you exit the dialogs by pressing the Return key, the original screen is restored with the appropriate file names set behind the scene just as with the tkinter dialogs. Loading and saving has not changed and is implemented through the Challenger 1P cassette interface.

    So as not to create too much work for myself I tried to keep the screens as minimal as possible. I think that the aesthetic is in keeping with the other period screens.

    The load screen looks like this:

    It supports a scrolling list of any number of file names. The files listed must be located in the TAPEs folder and only files with a .bas extension will be shown. Similarly the Save screen looks like:

    Simply type the file name in and press Return. If the user does not add a .bas extension one will be added on their behalf since only BASIC files can be loaded or saved.  There is an additional check to see if the file name entered already exists. If it does:

    Only pressing the 'Y' or "y" key will allow overwriting. Any other key and the user will have to start over. 

    All of the above changes have been posted to this projects GitHub repository.

View all 6 project logs

Enjoy this project?



John Croudy wrote 12/23/2022 at 02:02 point

This is really interesting and it looks great. I had a Superboard II in 1981 and used it for a few years before I got into the Atari 400. I managed to interface it to a really huge TTL-based video terminal that I got from a surplus store and also an old ASR-33 teletype that I acquired from a decommissioned ICL 1900 system. I learned so much about computers from that machine. Amazing times that I still wish I could revisit.

  Are you sure? yes | no

Michael Gardi wrote 12/23/2022 at 02:29 point

Hey John. I had a similar experience with the SYM-1, a 6502 based single board computer.  Connected to it with Volker-Craig terminals that were popular in Waterloo in the late 70s early 80s (home town company).  Interfaced the SYM-1 to a TMS9918 based video card. Many fond memories. 

  Are you sure? yes | no

Martin C. Foster wrote 12/22/2022 at 22:11 point

Back when I had a C1P, I envied the C4P. By the time I could afford one they were gone. Your project looks great!

  Are you sure? yes | no

Michael Gardi wrote 12/23/2022 at 02:31 point

Thanks Martin. The Ohio Scientific line of computers was pretty strong. It's too bad they didn't "make the cut",

  Are you sure? yes | no

brian.cockburn.1959 wrote 12/22/2022 at 02:02 point

Perhaps you meant aesthetic rather than ascetic ??

  Are you sure? yes | no

Michael Gardi wrote 12/22/2022 at 02:28 point

I had to look up ascetic , but ya that's not at all what I meant. ;-)  

Thank you.

  Are you sure? yes | no

Michael Gardi wrote 12/13/2022 at 10:51 point

Thanks Tom.

  Are you sure? yes | no

Tom Nardi wrote 12/12/2022 at 19:24 point

Always fascinating to read about your process to bring these old machines back to life.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates