Vintage Computer reproduction - Ohio Scientific Superboard II

[jbtech]

Hi all,

- this was a little side excursion into early computing. It's funny how nostalgia will transport you all the way back to the very beginning of your story.

I bought my first computer in 1980 when I was still in high school, with money saved up from weekend work. It was an Ohio Scientific Superboard II, model OSI 600. For me this was a defining moment.

1980 was the first year schools in my region offered any class directly related to computing. It was called 'an Introduction to Computer Science' and mainly concentrated on programming in BASIC. With just a single computer for the entire class to type in and test run their program assignments the pace was limited by access to this mysterious device. It was an unknown model with some early 8 bit processor possibly an 8080 or maybe even a 2650, BASIC in ROM and 8k Bytes of RAM.

Probably the result of a government tender for supply of computer equipment for education, it looked like it had been cobbled together by some local engineering firm rather than mass produced. At least it would have been locally made. Program storage was by audio cassette and there was a tiny printer for program listings. That was a sort of electrostatic device which printed onto rolls of metallised paper about the width of a modern cash register docket. The printed text looked like blackened dots on a silver - grey paper background.

As limited as this now seems it opened my eyes to the world of digital electronics, logic and programming. I was already interested in electronics and had built a number of projects, mostly from kits. I was also fascinated by the technology behind early video arcade games such as Space Invaders which really did seem space age to me at the time. Needless to say I took to the computer class with great enthusiasm and very soon needed a system which I could learn on in my own time.

Somehow I found an advertisement (possibly in the local newspaper) by a newsagent and stationer who had branched out into selling home computers, for the Ohio Superboard II. The Superboard was significant as it was a complete system on a single board requiring only a black and white TV monitor and standard cassette recorder for program storage - and a 5V power supply which was not included.

The fact that it was just a printed circuit board with no case or power supply didn't worry me in the least having built a number of small PCB projects and even making cases for some of them in woodwork class. At a slightly higher price the OSI 600 could be bought as a finished product in a case with internal power supply, known as the Challenger 1 but the Superboard at around A$300 was just within my budget at the time.

Besides in those days computers such as this were for experimenters and engineers alike, being sold with complete circuit diagrams and technical information encouraging and even leaving unused spaces on the PCB for customisation - unlike the pre-packaged and proprietary consumer products of today.

The Superboard II had a 6502 CPU which was clocked at just under 1 MHz, 8k Bytes of BASIC in ROM by some company called Micro Soft and in standard form was loaded with 4k Bytes of RAM with space on board to expand that to 8k Bytes. Having only 4k of program memory to work with was a definite learning experience; finding ways to write programs which were economical on storage and ran efficiently with minimal resources. Modern programmers would do well to take a step back and re learn these attributes!

In addition to the usual alpha numerics the Superboard's character set included many graphical figures such as tanks, spaceships as well as geometric shapes. I soon learned how to place these on screen and move them around using the BASIC POKE command, writing a number of games with titles such as Road Race, Tank Battle and inevitably the one which was my final project for the school computer subject, a tribute to the current arcade sensation - Space Invaders.

Admittedly the graphics didn't scroll smoothly like the arcade original, the figures just jumping one character column or row at a time but it was immediately recognisable and quite playable. I wrote it for the Ohio character set and screen map initially, using every trick I had learned to save space such as combining multiple commands onto a single line (a colon using fewer bytes than an additional line number) etc. then adapted it to run on the school computer.

Having taken every imaginable shortcut to optimise the game speed on my computer such as moving each invader one at a time before jumping back to update laser base and missile positions it played quite smoothly. Unfortunately the school computer turned out to be much slower so the game play was a bit jerky on that machine but it was well received all the same. There is no question that I learned so much with the benefit of my own computer at home to experiment with, something that most people take for granted today.

I never did expand the RAM to 8k Byte or increase the clock speed to 2MHz (another popular mod and easy to achieve on the Superboard) but it did gain a case (in fact two different cases) along the way. The first case was timber with plywood panel cut out for the keyboard and covered in gloss white laminate. The Superboard being totally flat defined the shape of the case which looked much like the Challenger 1. The power supply was still external, housed in a metal box.

Eventually I replaced that case with a metal one which had originally housed a keyboard for an electro mechanical paper tape punch. These were being phased out at the time and surplus ones dismantled for spare parts so I was able to have the empty case for free. It was heavy gauge steel, I made the keyboard panel from aluminium and painted it white and grey rather than its original industrial shade of blue.

This scan from a photo taken around 1988 is the only one I can find of my original Superboard, seen in its final shape and form. To align the keyboard with the slope of the case the entire PCB was mounted inside on an incline and the power supply built in to the base at the back. The case is obviously quite a bit wider than the Superboard PCB as you can see I added a numeric keypad repurposed from a desktop calculator, wired in parallel to the numeric keys.

Just visible in the photo is a pair of joysticks which I had begun installing into the front panel just above the keyboard, giving the machine a slightly Dalek-like appearance. They were analog ones with potentiometers but the intention was to convert them to switch closures so they could also be wired in to the existing keyboard and operate in the same way as key presses. That's as far as I got with the Superboard, eventually it was sidelined as IBM compatible machines with MS-DOS, floppy and hard disk drives became commonplace.

Sadly, the Superboard stopped working at some point and was eventually discarded to free up space for newer machines and projects, its large and heavy case became its downfall. Technology was advancing so rapidly at the time that previous generation machines seemed irrelevant, their incredible significance unappreciated until recent times. With a renewed interest in earlier technology I have searched online for surviving examples of the Superboard from time to time but never came across another one which was within reach.

Looking again just recently I appear to have missed any chance at finding an original one for sale. Very few seem to have come to Australia to begin with, the survivors having mostly found their way into museums or already snapped up by collectors. Thanks to the efforts of some OSI and Vintage Computer enthusiasts however there is another possibility, to assemble a reproduction of the original Ohio Superboard II using a brand new PCB and mostly original components.

Ohio Scientific were trailblazers in the micro computer field in the late '70s to early '80s and the definitive resource for information on the systems they produced is Dave's OSI Repository at www.osiweb.org. Thanks to its many contributors the site hosts an archive of documentation for several systems produced by Ohio Scientific during their brief corporate history. One contributor of note, Grant Klyball ( klyball.com ) has faithfully recreated the OSI 600 Superboard II PCB layout as well as the 610 expansion board and kindly shared those on the osiweb site.

My sincere thanks to Grant for his outstanding feat in recreating the Superboard PCB so perfectly and Dave for keeping the history of Ohio Scientific alive for enthusiasts around the world like myself. So having finally abandoned hope of picking up an original Superboard I've resolved to build a replica using the Klyball 600 PCB design and sourcing the components to complete the build. This takes me back to my very early days experimenting in electronics and building kits.

To begin with I've sent the PCB files off to have a small batch produced. That turned out to be a learning experience in itself, like many late '70s products the OSI PCB had no solder mask originally, the entire copper area being tinned. This does require care when assembling to avoid shorted tracks and also caused some concern by the PCB manufacturer who are accustomed to producing modern, high volume PCBs to the latest precision standards.

Initially the PCB files bounced in pre production audit - the solder mask layers are missing, please correct. I responded there was no intended solder mask so they asked if I just wanted the copper completely covered by HASL? Doing a bit of reading, HASL stands for Hot Air Solder Levelling; after solder mask is usually applied any still exposed copper areas are tinned in a sort of wave soldering process then the boards are hit with hot air to smooth the soldered area. Yes, thanks that's exactly what I would like was my reply.

Then came the caution, there will be some risk of shorted tracks - I assume they mean during production not only during assembly - was that OK? Yes, I assured them I would take that chance. Then they came back and asked wouldn't I prefer an ENIG surface treatment? More reading and ENIG stands for Electroless Nickel Immersion Gold and instead of tinning the copper tracks a fine layer of nickel is deposited followed by a dip in gold which is ideal for surface mount and precision assembly and would also reduce the likelihood of shorted tracks.

Imagine the cost of nickel coating the entire copper area on these boards (which are over a foot square, in the old money and double sided) then dipping them in gold! Besides, that would be nothing like the original '70s finish. No thanks, I replied, HASL will be fine.

Now we were getting close but the Klyball design did include a silk screen layer with some component outlines and that was another issue. Apparently the normal procedure is to etch the tracks then apply solder mask then silk screen then HASL so without any solder mask the silk screen would be applied directly over the copper and destroyed during the HASL process - so that was not possible. I have seen older PCBs with no solder mask, these were usually silk screened after being tinned but that apparently doesn't fit in with the current production process.

In any case the original Ohio Superboard PCB had no silk screening at all so the simple answer was Yes, please just leave the silk screening off. Having made those choices the PCBs sailed through production and arrived very soon afterwards and they look fantastic. Just the look of the unpopulated PCBs and their outline is bringing back memories of my original Superboard.

There is no urgency to complete this project and I am sure it will take some time to source the components as some are obsolete and can only be bought as old stock. Thankfully Grant addressed the issue of non available original keyboard switches in his design, adapting the layout to accept modern day Cherry MX key switches commonly used in mechanical, gaming keyboards. Likewise he incorporated a minor adjustment to the pinout of the 24 pin sockets to accept 2716 EPROMs in place of the original ROMs.

I believe at one point a small batch of reproduction key caps were commissioned by someone, to suit the new key switches with the original style lettering and layout but suspect these were all spoken for some time ago. I may just begin with an 'off the shelf' set of keycaps if a full custom set is beyond the budget, in fact I've been considering the option of using a compatible gaming keyboard as a donor for key switches and caps as a possible way to reduce cost and keep the project more in line with the original intent of a low price computer for experimenters...

12/10/23

I've begun assembling a reproduction of the first computer I owned, the Ohio Superboard II from 1980. This reminds me of previous projects which I've assembled from kits except in this case I'm having to source all of the components separately including the PCB itself. Fortunately there is a complete set of assembly instructions available.

Though my original Superboard was bought as a complete PCB needing only a power supply and TV monitor to operate, the model was eventually sold in kit form so I'm following the assembly manual which was issued in 1982 after Ohio Scientific was bought out by M/A Com.

A very similar computer known as the Compukit UK101 was also released in kit form, this was apparently a non authorised clone of the Ohio Superboard which originated in the U.K and featured some enhancements. Although similar to the Superboard the two computers are not identical, some UK101 kits may have made it to Australia at the time but I have never seen one.

The assembly procedure begins with IC sockets followed by small components such as diodes, resistors and capacitors. I have many of these on hand to begin with and can source the few remaining passive components locally. Meanwhile I have some of the TTL ICs as well as the CPU and EPROMs but will need to order the remainder including a number of obsolete types and wait for them to arrive.

12/11/23

The Ohio Superboard II reproduction is coming along, I now have almost all of the major components with the exception of the key switches and caps - and a couple of hard to source obsolete ICs which are on order but yet to arrive from overseas suppliers. All of the IC sockets which are normally present on the Rev. D PCB have now been installed and the full complement of TTL ICs fitted.

The discrete semiconductors and passive components are also installed with the exception of a single 47uF axial electrolytic capacitor which remains outstanding from a previous order. If that item doesn't arrive soon I will substitute a 100uF axial which I have to hand and will work equally well in this application.

Also on hand but yet to be fitted are the 6502A CPU, 6 x 2716 EPROMs and 10 x 2114 RAM ICs (4k Bytes of program memory plus 1k Byte of Video RAM). Yet to be received are the 6850 ACIA port and 2 x 8T28 buffer ICs.

Here I'm using my PC controlled EPROM programmer to install the firmware ROM files into the 2716 EPROMs. The Klyball PCB reproduction incorporates a minor modification to support 2716 EPROMs in place of the 9316 type ROMs which would have been used in the original.

The files I'm programming into the EPROMs are; the original Syn600 monitor ROM which contains the bootup routine and machine code editor program, Basic ROMs 1 to 4, each 2kByte to contain the 8k OSI Basic in ROM - in this instance Basic ROM 3 is the 'patched' version which corrects a known bug in the original so I'm labelling that one 'Basic 3b' - and the Character Generator ROM.

I've done a small 'hack' to the Character Generator ROM file, editing some of the content to include a set of 'Space Invaders' inspired characters. These differ slightly from the Taito originals, being only 8 x 8 pixels to fit within a single character space while the Taito graphics each occupied two bytes wide by 8 high.

30/11/23

A little more progress with the Ohio Superboard II reproduction, the 8T28 buffer ICs have arrived but the 6850 ACIA ICs which I ordered some time ago appear to be stuck in the postal system. I've installed the buffers, EPROMs and CPU to their respective sockets on the PCB.

Meanwhile I conceded the effort to desolder and remove 53 key switches and caps from a donor keyboard would be too great and the results less than satisfactory as some of the Ohio keycap sizes and legends don't match up with modern keyboard layouts.

Instead I've ordered, received and fitted new Cherry MX key switches, these ones are the basic 'black' type MX1A-11NW with a linear, not clicky action. I also managed to source a levelling mechanism for the spacebar with the correct mounting arrangement and spacing, that one has type number G99-0744.

I'll try to obtain a set of reproduction keycaps for the Superboard if possible. It may also be necessary to substitute a higher force key switch for the spacebar to compensate for its extra weight and friction of the levelling mechanism, we'll wait and see.

23/12/23

In this update we'll get the Ohio Superboard II reproduction up and running, sort out any issues with the build and even add a couple of minor modifications. 'Hacking' or adding hardware modifications was common practice back in the day and the Superboard, being simple and well documented lent itself to experimentation.

In this instance the intention is to recreate the original Ohio Superboard experience and not to 'hot rod' the board with increased capacity or performance, I'm just adapting the external connections slightly to fit in with the more modern peripherals I have to hand. More on that later.

Apart from not having any key caps as yet the only missing item was the 6850 ACIA (Asynchronous Communications Interface Adapter = serial port) IC which never arrived from order and appears to be lost in the postal system.

It wasn't an expensive part and I could reorder but the lead time would be an issue so, hunting through my old spare PCBs I've found one which I can desolder and reuse. Unfortunately in my haste I've broken one of its pins (should have taken more care) so I'm just repairing that by soldering another leg cut from a faulty IC onto the stub of the broken leg.

Eventually I'll replace it with a new part but it should be fine for testing so, cleaning up the traces of solder from the IC legs and placing it in the waiting 24 pin IC socket the Superboard is ready to power on.

Initially I'm using my bench power supply which has an adjustable current limit with meters to show the Voltage and current drawn, also connecting the Superboard's monochrome composite Video output to the 14" NEC colour TV video monitor which I use on the bench for testing arcade PCB repairs. Powering on I see a faint image with the letters 'H / D / M ?' near the top of screen.

That's not quite as expected, the familiar startup prompt should read 'D / C / W / M ?' and be located near the bottom of screen. Also there is no response when any of the keys are pressed. Initially thinking there may be some memory error causing the incorrect text and placement, I soon realise the monitor ROM file which I've used was not the correct SYN600 version for the Challenger 1 / Superboard II.

That would also explain the lack of keyboard response as some other Ohio Scientific products used a serial rather than polled keyboard. Locating another copy of the SYN600 ROM file from a different source, erasing and reprogramming the 2716 EPROM gets us back on track.

Powering on again and seeing the correct prompt - still a little faint on screen but we'll worry about that later - finding the C key and then pressing RETURN for 'Memory Size' and 'Terminal Width' prompts, the familiar startup message displays but the available memory seems incorrect.

I've fitted 4k Bytes of RAM and the message should read '3327 Bytes Free' but I'm only seeing 1285. The OSI RAM test appears to just check each byte of RAM in sequence, stopping when it does not detect a correct result. So the first 2k of RAM should be working but there is an issue in the third pair of 2114 RAM ICs.

The main RAM can be added or removed in 1k Byte increments, each pair of 1k x 4bit ICs being 1k Byte. Only the first 1k is required to boot but after the reserved memory only 255 bytes of free RAM will remain. A second pair of ICs will increase that by 1024 Bytes to 1279 and a third pair will result in 2303. Swapping the 3rd and 4th pair on this board gives me 2309 so the fault is now in the 4th and final pair.

I only have a couple of spare 2114 ICs left so substituting one at a time, for the ICs in the 4th row identifies the faulty IC and fixes the problem. I now have the correct '3327 Bytes Free' message showing on screen.

So it's time to check the basic operation with some meaningless program, finding the unmarked keys is a bit tricky so we'll keep it brief:

10FORI=1TO24
20?"HELLO WORLD"
30NEXT

Old habits quickly resurface and it seems as if I'm transported back to 1980 - omitting spaces as they waste valuable RAM - did I mention this computer has less than 4k free to work with?

That runs as expected so now having a short program in memory the next step is to test the all important cassette interface. I don't have any pre recorded program tapes to load, my old tapes and hand written programs sadly having gone with my original computer years ago so I will begin by saving the current test program to a blank tape.

I have a suitably old mono cassette recorder to use with the Superboard, which I picked up from a pile of junk left out for kerbside collection. With a clean and a new drive belt it seems to play OK but to begin with I'll use a higher quality Onkyo cassette deck which has manual level control, metering and monitor output to ensure an optimal recording.

The cassette is the only one I had which was still wrapped and unused, I've since used it to save a couple of Commodore 64 programs to tape so for the Ohio I'll turn it around to side B and rewind the tape to start with an unrecorded and unworn section. The Onkyo cassette deck is stereo of course so I'll just record the program onto the left channel and leave the right channel blank.

Using the SAVE command then LIST and starting the recording before pressing return, the program seems to record well and sounds fine on the headphone monitor. Now using the BREAK key to reset the computer, after startup prompts typing LOAD and replaying the tape - nothing happens.

If there was an issue with the recording having marginal level or noise I would expect to see some random characters appear during the LOAD operation so I'm guessing there is some hardware problem either with the 6850 or the input circuit around the CA3130 comparator IC.

I have my oscilloscope handy so will begin with the signal arriving at the comparator. At this point there is a pair of diodes back to back to square up the signal and limit its amplitude so there does seem to be sufficient input level for that to occur. There is however no signal on the output pin 6 of the CA3130 comparator IC.

That part could well be faulty or may even have incorrect markings - at the time I was ordering parts for this project my usual suppliers had no stock of this item so I ordered a small quantity from an overseas seller online. Testing each of those yields the same result so I'll try to find some locally available stock to replace this component.

- With a bit of searching I have managed to find some local stocks of this item so, picking up where we left off the LOAD operation now works perfectly and the program which I saved earlier now loads, lists and runs without error.

Next trying a SAVE and LOAD operation with the Digitor cassette recorder, unfortunately does not work and reveals an issue with its recording levels. There appears to be a problem with the input amplifier or possibly a lack of bias to the record head so the playback level is too faint to be useable. No matter, I'll put that aside for the moment and should be able to repair it at a later date.

Moving on to the faint composite video image on my TV monitor, I believe that issue is more to do with the TV not being suited to the Superboard's video standard. The TV is a PAL standard, colour model which can display the 60Hz monochrome signal but its composite video channel would have a filter to suppress the chroma signals which centre around 4.43 MHz for PAL as opposed to 3.58MHz for NTSC.

The Superboard's dot clock frequency being 3.93 MHz may be a bit too close to the PAL chroma frequency, causing a reduced response - in any case a black and white TV or monitor would be more suitable but I unfortunately don't have one to try.

The next best thing would be to use the NEC TV in RGB mode with separate sync, this will give a better response having no chroma filter in the video channel and I can just use the green input which will give a clearer display than white text on the colour screen, with no convergence issues.

Getting separate video and sync from the Superboard is quite simple as these signals are just combined at the final output stage to provide the standard composite video. Looking at the connector J2, pins 7 to 12 are used for the cassette I/O and composite video output but pins 1 to 6 are not used, only connecting to a part of the PCB which is not populated.

It appears those pins were reserved for additional serial data and clock signals, maybe for synchronous communication but as they are not implemented in the standard configuration I'll use pins 4, 5, and 6 for Video, Sync and Ground.

Adding a ground point is simple, linking pin 6 to 8 and sync (pin 5) is just linked directly to pin 12 (composite video) then the video level reduced by turning R58 anti clockwise. Finally the video signal from IC70, pin 6 is linked to J2, pin 4.

And here is the result, a nice looking 'green screen' - even though it is really just a colour TV monitor. I'll keep an eye out for a dedicated Black and White monitor or TV to use with the Superboard but meanwhile this will do just fine.

Edit:- With hindsight, the faint text on the composite video output may have been due to a component substitution I made. The original Superboard II used a mix of original 74 series TTL ICs and lower power 74LS components. The original series are by now very obsolete and often quite difficult to source so in this build I've attempted to standardise and use 74LS versions throughout.

There are often issues arising from mixing these types especially when the lower power 74LS is required to drive a number of standard 74 series inputs but generally using 74LS series components throughout sidesteps this issue. In the case of IC70 which would have been a 7403 originally, this device drives an external load so substituting the lower power 74LS version may have caused a reduction in the video output levels.

When setting the Superboard IIs output for separate video and sync I did notice that when R58 was turned anticlockwise to reduce the video level on the composite output, the level at the output of IC70 actually increased due to the reduced load. When running separate video and sync the 74LS03 appears to be performing correctly but for composite video this part should probably be returned to the original 7403 type or alternatively the higher power and speed 74S03 version.

3/1/24

The Ohio Scientific Superboard II reproduction appears to be fully functional with the video monitor now showing a clear image but there's one little extra which I'd like to get working with this computer. My original Superboard had no means of producing a hard copy program listing, only being able to save or load programs to and from audio cassette.

I'm pretty sure that computer was a Rev. B whereas the Klyball reproduction is of the Rev. D PCB. The Superboard had provision for an RS232 interface which was not populated on the Rev. B but the required components are fitted in the Rev. D version. A 12 way Molex 'shorting plug' configures the single serial port for either cassette, RS232 printer or modem / terminal connection.

Apparently the Challenger model had a rear panel mounted 3 position rotary switch to select the serial destination. I'd like the ability to select between the cassette and printer interface so that a program can be loaded, edited and the program listing printed out before saving again, without the need to remove and swap plugs. This can be achieved with the addition of a 3 pole, 2way toggle switch.

Now all we need is a compact printer with an RS232 interface, capable of printing ASCII text and at least 24 characters per line - and this is what I've found:

OK, so it's really just intended as a receipt printer but why not use it with the Superboard? It is very compact, has RS232 interface and its default font will print more than 24 characters per line so should be fine for program listings. It requires no ink and uses standard cash register style thermal paper rolls. The only immediate issue is its 9600 Baud rate default.

The Superboard does have Baud rate selectable via a link on connector J2 but the two default choices are 300 Baud for the cassette or modem and 1200 Baud for RS232 printer. Fortunately by utilising the undivided clock frequency CLK signal the highest available, 9600 Baud rate can be obtained. Thanks to Mark Csele for mentioning that in his technical description of the Superboard II (link below).

To make the Baud rate switchable between 300 (for cassette) and 9600 (printer) I need to link the CLK signal to J2. Pin 11 is listed as unused but is actually connected to +5V so I'm cutting the track from pin 11 just near the point where it joins the +5V rail, instead adding a link to the CLK signal at the nearest point, U59 pin 2.

Here's the J2 socket with 3 pole toggle switch to select cassette or printer operation. I'm not planning to put this new Superboard into a case, just keeping it as a standalone PCB this time so I've wired the switch directly to the socket pins rather than extending the wires to suit panel mounting. At this stage I don't envisage using the Superboard with a serial terminal but could make up a separate socket and just swap them over to change the configuration if required.

That reminds me; the first 'printer' I ever bought in about 1984 was a second hand Digital LA30 DECWriter in working condition, complete with full keyboard and stand (see link below). I wish I'd had the space to keep all of the historic equipment I picked up over the years, those were the days when real computers and peripherals required floor space rather than desk space.

Meanwhile at the other end of the scale it's time to test the little thermal printer with the Superboard, powering it up with a 5V, 2A supply and connecting the RS232 printer port. Once the SAVE command is entered all of the output from the Superboard is duplicated to the serial port including commands typed, program listings and print statements. To discontinue printing the LOAD command is entered followed by space and enter keys.

So, typing in SAVE followed by LIST and RUN commands, the little printer buzzes away and here is the result:

Terrific! I would have loved to have had a little printer for my original Superboard back in the day, to print out listings of programs I had written rather than copying them out by hand. The irony here is this modern receipt printer probably has way more embedded processing power than the computer which now controls it.

The narrow printout is very reminiscent of the ones produced by the original school computer printer except that one printed onto metallised rather than thermal paper in a process known as 'spark printing'.

So having proven the concept I'll fit the thermal printer module into a case of some sort - it is designed to be panel mounted inside some larger equipment, cash register etc. Originally I planned to use a standard 'Jiffy' type enclosure, cutting openings for the printer top panel, power and interface connectors but none of the readily available housings were a good size match so instead I decided to design and 3D print a custom enclosure to suit the exact dimensions of this device.

Here's the design for a simple enclosure made in one piece which the printer module can drop into, with a little extra space at the rear for connections. I've chosen a Molex style 2 pin connector for 5V power and a 3 pin Mini Molex type for the RS-232 interface, aligning the rear panel cutouts vertically to minimise their unsupported overhangs. I've also added a tapered support for the top surface so the part can be printed with no additional supports and minimal infill.

printbox.123dx 3D drawing
    

printbox.stl 3D object
    

And here's the finished product, printed in ABS. The cutouts for the connectors printed almost perfectly, I had made them a snug fit so just used a small file to finish them off. The connectors themselves are readily available from Jaycar, Part Numbers PP2020 and PP2023. Each comes as a complete set so there is no need to buy pins or mating connectors separately.

One little trap I've noticed; if the interface is set for printer operation but the printer is not connected and ready the screen will not update as the computer waits for a /Clear To Send signal. This appears to be normal behaviour and not a fault so the best practice seems to be to leave the interface set for cassette operation unless actually using the printer.

So, with the reproduction Superboard II now working and the RS-232 printer complete and tested all that remains is to source a set of key caps and to repair the portable cassette recorder which I had planned to use for program storage, when time permits.

 

Ohio Scientific Superboard II model OSI600 - specs.

Made        From 1978
CPU          MOS Technology 6502 @ 0.98304 MHz
RAM          4 kByte expandable to 8 kByte, plus 1 kByte video RAM
FDD           Required 610 Expansion board
Ports          Cassette / Serial, Expansion (40 pin DIP socket)
O/S            OSI 6502 BASIC V1.0
Display       Monochrome, 24 x 24 characters displayed
Sound         None apart from 8 bit DAC which can be used to generate sounds
Monitor      Monochrome Composite Video
Keyboard    On board, 53 keys
Speakers     Not included but amplified speaker can be connected to DAC output

Web Resources (External Links) -

Dave's OSI Repository - osiweb.org

KLyBall's Vintage Computers - klyball.com

Mark's Ohio Scientific/Compukit UK101 Archive

Ohio Scientific Superboard II - Mark Csele

Ohio Scientific Superboard II Replica - Jeff Tranter, YouTube

DEC and Centronics Introduce the First Dot Matrix Printers - History of Information

Spark Printing - Wikipedia

 

Anyway having enjoyed that little detour down memory lane it's time to get back to some arcade machine projects and PCB repairs,

More updates to follow, regards, John.

[needlebat]

What a wonderful project - and a pleasure to read. I built a Microbee when they were available as a kit, and a MSC-ICO - Mountainside Computer Intelligent Console Object - a CPM 3.0 machine with full source code for the O/S - what exciting days, and so much to learn, like how to build a modem because you can't afford the fancy new ones, but you absolutly have to have the 1200/75 bandwidth! Thank you for the great read!

[furballx]

Nice John.

Around the same time at my school we had a Microbee and two Dick Smith System 80's.

It feels like a long time ago!

Dave

 

[jbtech]

Great stuff! I think a lot of schools eventually went with the microbee which I guess was a sort of clone of the BBC Micro and good value for education.

Whatever happened to all of those Dick Smith computers, I'd just about forgotten about those altogether. I guess they were similar to the Tandy TRS-80 at the time (we called those TRASH 80's for some reason)

I used to read the DSE catalogues from cover to cover back in those days... https://archive.org/details/dick_smith_electronics_catalogue_1981-82/page/2/mode/2up?view=theater

I seem to recall hearing from a school mate who was a year behind me that my school replaced their original clunker with an Apple - or maybe one of those 'Apricot ' clones which were also around at the time. So many advances In the space of a couple of years!

[furballx]

Yep, Dick Smith was ripped off or rebadged TRS-80 as far as I am aware.

They did sweet F ALL! 🙂

[jbtech]

3 hours ago, furballx said:

Yep, Dick Smith was ripped off or rebadged TRS-80 as far as I am aware.

They did sweet F ALL! 🙂

Any computer with wood grain sides must be good 🙂

[Arcade King]

Love it!

[Sideburn1]

A couple years ago a friend gave me his Superboard II and I just got it out of the closet last night to dive into repairing it again.  When I power it up I get the random characters but hitting break would not exit to the console. I put a scope on it and saw no signals on any of the rom or cpu address lines but have a clock. 
 

I then pulled the chips and cleaned with deoxit etc. I put the cpu into and asteroids board and confirmed it is working. I then tested all of the 2114 rams and they test good. 
 

eventually I found I could hit break enough times and the cpu should come to live and I would see activity on all the data and address lines but still only random characters on the display. 
 

I also realized I could remove all of the roms and the cpu and still have the characters on the crt so the character generator is doing its own thing independently. 
 

the two mc6889 chipping very warm. 
 

Anyway good news is the cpu, ram (8k), video ram, character generator and I think the roms ? Are good. Just need to figure out how to get the thing to boot up consistently and get my prompt on the screen. Now hitting break a bunch of times isn’t getting me activity anymore. 
 

would any of you be up for helping me fix this? Or is there a forum somewhere where I can get some support?

 

oh there’s also a bunch of bodge wires on the back of the board and not sure if it’s required or some mod. It is a rev B board. 

Here’s a little video clip of it I just got ot to fire up again after banging on the break key like a monkey: https://youtu.be/80U1U5QqdBo?si=BotRwM755Y8uA9Vi

[jbtech]

Hi, it's nice to see any original Superboard II out there, that one appears to have had a lot of modification!

Either that or maybe someone has attempted to fix it by bypassing corroded tracks with wire. Hard to imagine there would be so many bad tracks though.

It would be worthwhile checking to see if tracks have been cut and modified connections made with wire or whether the wires follow the original circuit connections.

There are originally no additional wires on the Rev. B PCB. Circuit diagrams for that and other revisions can be found at Dave's OSI Repository:

https://osiweb.org/ and there are discussion forums there at https://osiweb.org/osiforum/

 

Troubleshooting the Superboard II would be much like an arcade PCB of the era, beginning with supply Voltage, CPU clock and reset signals then address and data activity, address decoding and ROM / RAM select lines etc.

Have you checked to see if the BREAK key is stuck on (short circuit) or whether the reset line goes low when Break is pressed then high again?

The 6502 will begin by reading near its highest address to find its boot location so that should lead it to the monitor ROM which is furthest from the CPU, next to the 6850 ACIA, so there should be some address activity immediately after reset.

Your monitor ROM appears to have been replaced with an EPROM (there are other versions of monitor software available) but depending what EPROM type there may be a couple of minor track modifications required as the 2716 for example is not fully pin compatible with the original ROMs - something to check.

At least there is sync and the character generator appears to be reading random data from the RAM and outputting valid graphic characters.

But the CPU is apparently not writing anything to video RAM at all. The data buffers running warm could be a clue to a data bus issue but they do tend to run a bit warm anyway especially if they are stuck in one state rather than having normal activity. Is there any data activity on either side of those buffers?

Good luck with it, check out those OSI forums as there could be known issues or modifications that I haven't come across.

Regards, John.

Edit: I should have noticed earlier that there are waaay more characters on your screen than as per the standard 24 x 24 characters displayed so assume that is a substantial mod and not just mis-timing. It seems to be running 48 x 32 characters visible approx which would require additional video RAM to the 1kbyte original...

 

Edited February 3 by jbtech

[Sideburn1]

There's no corrosion on the board and yes there are cut traces so it is mods.

Yes prior to my first post I had already checked the clock (good), tested the CPU on my Asteroids board (good), tested all of the 2114 rams (good), supplying 5v / 3A to the board from a bench supply, and there was zero activity on the bus but i could hit break (reset) over rand over and sometimes get lucky and there would be some activity on the BUS...

I have made some more progress. I was able to read the BASIC ROMs #1,2, and 4 but #3 is erroring.

Same with reading the Monitor ROM, so I burned a new SYNMON1 onto a 2716 EPROM (there are bodge wires under the board at the location so i think it may be modded   but the monitor eprom it came with is a 2708 and i just burned a 2716) and now when I hit break i get locked in activity on the data bus and the screen goes form the random characters to a blank screen.

Do the BASIC ROMS need to be installed in order to get to the monitor page? I think my #3 BASIC ROM may be bad.

What EPROMS can I use to burn 4 new BASIC ROMs? Will 2716's not work without modifications to the board? I just ordered a batch of them.
 

[jbtech]

Hi, yes I realised later that your board must be modded to display more characters on screen than the original 24 x 24 - and could have other modifications as well...

And I did read you had already checked the clock signal, I was just describing a general checking process - CPU reset being the next thing I would check after power supply and CPU clock.

That's unusual if there was a 2708 in the monitor ROM position as I'm pretty sure all 2708 ICs require three power supply rails +5. -5 and +12V - maybe it was just put there to make the board look complete at some point?

the 2716 has the correct capacity (2k x 8bit) and will work in place of the original ROMs with only slight modification to the PCB. There is a section about modifying the Rev. B PCBs to use with 2716 EPROMs on this page: https://markcsele.ca/hobbies/my-vintage-computers/ohio-scientific-superboard-ii/

I think SYNMON1 is not the correct monitor ROM version for the Superboard II, the version which would have originally been fitted was SYN600, as found on this page:

http://osi.marks-lab.com/software/roms.html   There are other alternatives such as CEGMON, the Challenger 1 (C1) version of that would also suit the Superboard II.

I'm pretty sure the BASIC ROMs don't need to be installed in order for the monitor page to work, though I don't think I have ever tried it.

I don't think any EPROMS are exactly interchangeable with the original 2316B ROMs so the 2716 with a small PCB modification is probably still the best bet.

The original BASIC ROM 3 had a bug which sometimes affected memory free space when using character strings so if you're replacing it you might as well use the 'fixed' version of the BASIC ROM 3 file, also found on the page link above.

Regards, John.

 

 

 

 

[Sideburn1]

Oh my bad I had programmed SYN600 into it (not the SYNMON1) and it seems to be working.

Wow if it does not need the BASIC ROMs then I need to keep digging elsewhere in the video output area or something.

I have done some more probing and here's what I have found:

With the new monitor in the 2716 and after hitting break, the screen goes blank and I have activity across all data lines and address lines AND pin 20 (output enable) is cycling on THAT EPROM and there's activity on all banks of RAM so it looks like the monitor is running.

I then checked U17 where the chip selects are going to pin 20 of each of the BASIC ROMs and I setup a single shot capture on the o'scope, cycled power and hit BREAK and every single time I am getting a chip select on U17 pin 4 which is BS3 (chip select for BASIC ROM 4). And that ROM must be installed or the board is not coming alive when I hit break. there is only a single shot to BS3 after that none of the BASIC ROMS have CS activity.

Also, none of the other ones are triggering on the single shot capture when I hit BREAK, only ROM #4 (U12)

I am able to read all but #3 on  my eprom programmer but I am selected a 2716 for the EPROM model... But #3 is giving me an error on read every time so that makes me suspicious.

So, if it is true that I do not need the BASIC ROMs installed then there must be something going on somewhere else.

Every ROM i find online (here: https://github.com/sjgray/OSI-monitor-roms and here: http://osi.marks-lab.com/software/roms.html) is not verifying agains my ROMS.

I am reading in completely different data so I will not be able to put a new set of BASIC ROMs in until my additional set of 2716s get here.

If you have a working board would you be able to pull your BASIC ROMs out and see if you can get into the monitor still?

Thanks for helping me out. It's a pretty sparse subject online and hard to get info. Are these boards really rare and valuable? I don't know much about them but I am learning fast.

** I had skimmed over markcsele's page a couple days ago. I will look at it in more depth
 
-Tavis

 

Edited February 4 by Sideburn1

[jbtech]

 

 

3 hours ago, Sideburn1 said:

If you have a working board would you be able to pull your BASIC ROMs out and see if you can get into the monitor still?

Thanks for helping me out. It's a pretty sparse subject online and hard to get info. Are these boards really rare and valuable? I don't know much about them but I am learning fast.

Hi, I tried that on the WinOSI emulator (that is also available from marks-lab link above), substituting a different ROM file for the default 8k BASIC (just selected an 8k arcade ROM, completely unrelated) and it crashed on reset with a blank screen.

So it appears the monitor and startup prompt won't run without BASIC ROMs installed for some reason. Sorry - my guess that it should run without them was wrong.

Compared to a lot of arcade PCBs and even other computers I believe there is actually a lot of detailed information available for the Superboard II and OSI computers in general with complete manuals, circuit diagrams, technical descriptions and software available thanks to a group of dedicated enthusiasts and collectors.

For a pretty obscure brand with a short life span during a period of such rapid advancement I think it is pretty amazing there is anything still available at all.

I'd say original Superboard II computers are fairly scarce, rare here in Australia and sought after. I wouldn't know what the U.S. market value would be though.

Regards, John.

P.S. I nearly forgot to mention; make sure you have the latching shift lock key ON otherwise the keyboard won't function as expected, it's not the same as the shift keys.

Also, just found this at the very top of the troubleshooting section of the Superboard II kit building instructions - seems BASIC 4 is critical to startup:

 

[Sideburn1]

Good point I guess there is a good amount of info concidering how rare it is.

The schematics and manual really don’t cover much technical detail on how each section of the board works. I’m struggling trying to map out the video section and what all the edge connectors on the board does. I know the one where you tap the video in is also for cassette/ data I/0 bit can’t figure out what the others are for. 
 

yeah I’ve had shift lock down. It makes no difference at the moment.  
 

So it’s looking more and more like a bad basic rom. Unfortunately I’m going to have to do the mod to get the 2716 roms to work. I wonder if I can do the mod for just the basic #3 rom and try since I have one eprom available. Maybe since the original roms are masked is why my reads aren’t matching up when I verify in the programmer. 🤷‍♂️ 

in the meantime I have been combing through the schematics trying to figure out how the video works. I know the character generator chip is independently putting the random characters on the screen at power up but I cannot yet figure out what is making the screen go blank on power up once the cpu and monitor boot up upon reset. 

Trying to sort out how the cpu / rom section of the board controls the character generator and video. 

And trying to figure out what U22 - U23 are doing and what their purpose is.  

inspected the bodge wires some more and most of them are wires running from u22 -u23 to the PROTO section where there are socketed chips that have been added. Also I think there’s a lot of rework done in order to get 40+ columns of screen real estate which is why there’s two more rams piggybacked onto U39 and U40

other wires are jumpering over to some of the pins on some of the J connectors. If I get the board running then I can go though carefully and undo some of that stuff and remove the extra chips in the PROTO areas. 
 

** I’m going to see if I can use a socket to do the 2716 mods in a non destructive way for those BASIC roms.  Right now I want to try to avoid cutting traces on the board. Looks like I just need to reroute pins 20 and 18 if I am reading that post correctly since I have a rev B board…

Edited February 4 by Sideburn1

[Sideburn1]

I will look into the emulator. I found this page : https://osi.marks-lab.com/reference/files/C1P-ServiceManual.pdf

 

and under the rom monitor and operation check it mentions to check rom #4 only and that is interesting because as I mentioned when I hit break only rom #4’s OE pin goes high/low so that makes me wonder now if the system would run if only rom 4 was in it. 
 

could you try in the emulator again with just a valid rom #4 ? 
 

I will look into setting up that emulator here so I can experiment. 

[jbtech]

10 hours ago, Sideburn1 said:

The schematics and manual really don’t cover much technical detail on how each section of the board works. I’m struggling trying to map out the video section and what all the edge connectors on the board does. I know the one where you tap the video in is also for cassette/ data I/0 bit can’t figure out what the others are for. 
 

Hi, the J2 connector which has the video and cassette connections also has some pins which are unused, connected to a couple of IC locations which are unpopulated but they are shown in the circuit diagrams and seem to be for serial data and clock signals so I'm guessing a provision for synchronous serial communication.

Back in the day a lot of dumb terminals and other devices used synchronous serial interfaces which required clock signals to accompany the data for synchronisation. That standard gradually faded away in favour of the simpler asynchronous serial interface so was probably just left off from production models.

The J3 connector has pins for standard, asynchronous RS232 connection but the input and output devices are also unpopulated on the Rev. B PCB. They can be added if you want to interface to printers or terminals (or may have been already added in your case as your machine has been heavily modified) Mark Csele's site mentioned above describes the serial interface in detail.

The J4 connector just has the polled keyboard rows and columns on it for external connection. I recall seeing some mention of accessory joysticks and numeric keypads which may have been available to buy and plug into that port, I did something similar on my original computer using a keypad from a desktop calculator. On the Rev. D PCB there is just a 2 pin connector there with the DAC output which is driven by the same port as the keyboard polling. When the keyboard is being read it just has varying tones according to key presses.

10 hours ago, Sideburn1 said:

So it’s looking more and more like a bad basic rom. Unfortunately I’m going to have to do the mod to get the 2716 roms to work. I wonder if I can do the mod for just the basic #3 rom and try since I have one eprom available. Maybe since the original roms are masked is why my reads aren’t matching up when I verify in the programmer. 🤷‍♂️ 

 

You should be able to modify individual ROM positions or not in order to use mixed original ROMs and 2716 EPROMs.

I'd be surprised if your original ROMs read correctly using your EPROM programmer as I believe the enable signals are the inverse of the nearest equivalent EPROM. An adaptor which corrected the enable pins would usually be required to read them with an EPROM programmer. Maybe that's why none of them would verify against known ROM files.

10 hours ago, Sideburn1 said:

in the meantime I have been combing through the schematics trying to figure out how the video works. I know the character generator chip is independently putting the random characters on the screen at power up but I cannot yet figure out what is making the screen go blank on power up once the cpu and monitor boot up upon reset. 

Trying to sort out how the cpu / rom section of the board controls the character generator and video. 

 

Like most computers and arcade machines with shared video RAM I think the CPU / Program just has to place characters into the video RAM area, a dedicated hardware section reads that out, converts the characters to pixels in the character generator and outputs it to the monitor as video.

If the CPU doesn't run or overwrite the video RAM, the random characters which are in there from power on are clocked out continuously. When the CPU boots it clears the video RAM before placing characters.

11 hours ago, Sideburn1 said:

And trying to figure out what U22 - U23 are doing and what their purpose is.  

U22 and U23 appear to be address decoders, depending on address bits 13 to 15 one of eight select lines from U23 becomes active, if all are low then U22 is selected and that creates eight select lines depending on A10 to A12. For example RS0 should be active low when A10 to A15 are all low and /02 clock is also low.

11 hours ago, Sideburn1 said:

inspected the bodge wires some more and most of them are wires running from u22 -u23 to the PROTO section where there are socketed chips that have been added. Also I think there’s a lot of rework done in order to get 40+ columns of screen real estate which is why there’s two more rams piggybacked onto U39 and U40

other wires are jumpering over to some of the pins on some of the J connectors. If I get the board running then I can go though carefully and undo some of that stuff and remove the extra chips in the PROTO areas. 

A lot of the known mods are documented in the Tips, Tricks and Notes section on Dave's site https://osiweb.org/#Tips

10 hours ago, Sideburn1 said:

could you try in the emulator again with just a valid rom #4 ? 

With BASIC ROM 4 included and 1 to 3 either blank or all zeros the emulator booted to the D / C /W / M ? prompt and 'M' opened the monitor program.

'W' brought up a crash notice in the emulator and 'C'' or 'D' seemed to put the CPU into a loop of some sort, no response to keyboard after that.

But at least the monitor did seem to load even if only Basic ROM 4 was present. Maybe the startup prompt is actually generated there...

Hope that helps, regards John.

 

[Sideburn1]

45 minutes ago, jbtech said:

Hi, the J2 connector which has the video and cassette connections also has some pins which are unused, connected to a couple of IC locations which are unpopulated but they are shown in the circuit diagrams and seem to be for serial data and clock signals so I'm guessing a provision for synchronous serial communication.

Back in the day a lot of dumb terminals and other devices used synchronous serial interfaces which required clock signals to accompany the data for synchronisation. That standard gradually faded away in favour of the simpler asynchronous serial interface so was probably just left off from production models.

The J3 connector has pins for standard, asynchronous RS232 connection but the input and output devices are also unpopulated on the Rev. B PCB. They can be added if you want to interface to printers or terminals (or may have been already added in your case as your machine has been heavily modified) Mark Csele's site mentioned above describes the serial interface in detail.

The J4 connector just has the polled keyboard rows and columns on it for external connection. I recall seeing some mention of accessory joysticks and numeric keypads which may have been available to buy and plug into that port, I did something similar on my original computer using a keypad from a desktop calculator. On the Rev. D PCB there is just a 2 pin connector there with the DAC output which is driven by the same port as the keyboard polling. When the keyboard is being read it just has varying tones according to key presses.

You should be able to modify individual ROM positions or not in order to use mixed original ROMs and 2716 EPROMs.

I'd be surprised if your original ROMs read correctly using your EPROM programmer as I believe the enable signals are the inverse of the nearest equivalent EPROM. An adaptor which corrected the enable pins would usually be required to read them with an EPROM programmer. Maybe that's why none of them would verify against known ROM files.

Like most computers and arcade machines with shared video RAM I think the CPU / Program just has to place characters into the video RAM area, a dedicated hardware section reads that out, converts the characters to pixels in the character generator and outputs it to the monitor as video.

If the CPU doesn't run or overwrite the video RAM, the random characters which are in there from power on are clocked out continuously. When the CPU boots it clears the video RAM before placing characters.

U22 and U23 appear to be address decoders, depending on address bits 13 to 15 one of eight select lines from U23 becomes active, if all are low then U22 is selected and that creates eight select lines depending on A10 to A12. For example RS0 should be active low when A10 to A15 are all low and /02 clock is also low.

A lot of the known mods are documented in the Tips, Tricks and Notes section on Dave's site https://osiweb.org/#Tips

With BASIC ROM 4 included and 1 to 3 either blank or all zeros the emulator booted to the D / C /W / M ? prompt and 'M' opened the monitor program.

'W' brought up a crash notice in the emulator and 'C'' or 'D' seemed to put the CPU into a loop of some sort, no response to keyboard after that.

But at least the monitor did seem to load even if only Basic ROM 4 was present. Maybe the startup prompt is actually generated there...

Hope that helps, regards John.

 

Thanks for the info and wow so all I need is BASIC ROM 4 and I should get something on the screen. That is very helpful. 

I did try burning a 2716 with BASIC 3 on it and did the pinout changes but no changes.... blank screen.

I have attached photos of my pinout changes based on what is described on markcsele's page about doing the BASIC roms on a rev B board. Does this look correct to you?

With the new info now I will try just a BASIC 4 rom and see what happens.

Herses my pinout changes:

With a socket I bent out pin 21, 20 and 18:

I then ran pin 21 to +5, pin 20 to pin 5 of U17, and pin 18 to GND:

 

Edited February 5 by Sideburn1

[Sideburn1]

The plot thickens.... Just poking around I pulled U65 and powered up and the video was scrambled and rolling diagonally but when I hit the break key the "C/W/M ?" showed up so now I know for sure the monitor is indeed running. Hitting a key does nothing though and i see no chip selects occurring on any of the BASIC roms. 

So now I know I must have some sort of video issue and well as BASIC not running.

 

Edited February 5 by Sideburn1

[Sideburn1]

We have success!!

The ROM on board was that 2708 EPROM. I removed all of the re-routing on the board for that chip and then pulled pin 21 and 19 high and connected 18 to U18 pin 3 and bam! got the prompt.

Here's a video link of it getting into the monitor: https://youtu.be/Iikr4KRpWhM

I then got into BASIC and wrote a hello world test! --> https://youtube.com/shorts/DQwJLW4sQwE?feature=share

The only issue now is its split across the screen... some kind off wrapping going on..

Edited February 6 by Sideburn1

[Sideburn1]

User bxdanny in the OSIWeb.org forum made me a modified monitor rom to support the 48 column mod my board has and now it's fixed. So awesome 😃

Edited February 6 by Sideburn1

[Bob63]

 Jbtech,

 

    I have been trying to get a thermal printer to work with my Ohio Scientific Model 600 CPU, 1978, REV B, Challenger 1P as you described in you fantastic write-up.  Can you detail how you connected your switch to increase the baud rate to 9600?   The location of your switch looks like the J3 socket on my board.  I have tried hooking it up several ways to no avail.

 

The printer I have is a DFRobot Embedded Thermal Printer Model DFR0503 (receipt printer).   (https://wiki.dfrobot.com/Embedded%20Thermal%20Printer%20-%20TTL%20Serial%20SKU%3A%20DFR0503-EN)

 The printer does work with an arduino nano.

I can load programs from a cassette player/mp3 player.

The RS-232 out is sending a signal as show in the attached.  This was taken while a program was running.

 

The connections were

J3 pin 4 (Rx DATA) to J3 pin 6 (RxDATA 2) jumper

J3 pin 1 (GND) to ground of printer

J3 Pin 2 (RS-232 out) to TXD of printer

J3 pin 3 (RS-232 in) to RXD of printer

 

W10 has been cut (on the backside) and a rotary switch has been installed (the switch was not used for this work) per a Microcomputing, April 1981 by Edward C. Jones article titled, "Energize Those OSI Peripheral Ports".

 

https://ia802907.us.archive.org/12/items/kilobaudmagazine-1981-04/Kilobaud_Microcomputing_1981_April.pdf

 

This W10 trace cut is the same trace cut as referenced on Professor Mark Csele's Webpage.

 

Any help would be appreciated.

 

Thanks

Bob

 

Edited March 21 by Bob63

[jbtech]

Hi Bob,

I've been reading that article "Energize Those OSI Peripheral Ports" in the magazine link you added, and thanks for including that - I'll have to read the rest of that mag. soon, it looks like a real trip down memory lane.

I've never seen the additional serial outputs on the Superboard II PCB explained in detail before and it does help to identify some differences between the Rev. B and Rev. D approach.

In the Rev. B version, once the empty spaces are populated it uses duplicated serial outputs and only the serial receive lines are switched between sources. A separate switch enables /CTS flow control and the baud rate appears to be set at 300, requiring a change of connections on board to select a higher rate.

On the Rev. D PCB the serial input and output on J3 are normally populated and the secondary serial outputs on J2 are not populated or used. Some links on J3  (or a rotary switch connected via J3) select both input and output between tape, printer or modem with baud rate selection of either 300 or 1200 and /CTS also selected via pins on J3.

To get your thermal printer working it should first be necessary to set the baud rate by linking pin 2 of U57 to the (CLK) signal at U59 pin 2 INSTEAD of the (C4) signal at U59 pin 14 where it originally connects. To switch between 300 baud tape and 9600 baud printer one pole of a double throw switch can be used to select either of those signals to U57 pin 2.

The other requirement is to have /CTS input active and connected to the printer for flow control. There is no need to connect the printer Tx to the Superboard Receive input. So only 3 wires are required, Tx data to the printer Rx, Ground and /Ready from the printer to the Superboard IIs /CTS input.

If you would like to reconfigure the connections to J3 slightly and do all of the switching as per the Rev. D board I'll add those details below along with my 3 pole, 2 way switch version which selects between cassette and printer without a 3rd position for modem.

On the Rev. B PCB J3 pin 8 is shown as not used but it would be best to double check the traces in case it is connected to something else. Then it can be connected to U57 pin 2.

J3 Pin 9 is /CTS and comes from the ACIA via link W3 - if not already changed that needs to be set so that /CTS at the 6850 connects through to J3 pin 9, not ground. See below.

J3 pin 10 on the Rev. B PCB is /CTS2 so that trace would need to be cut as below and pin 10 can then connect to U59 pin 14. The original trace from U57 pin 2 to U59 pin 14 should be cut so that signal can be switchable instead of fixed at 300 Baud.

J3 pin 11 on my Rev. D PCB was shown as not used but was actually connected to +5V so I cut that trace and instead connected pin 11 to the 3.93MHz (CLK) signal from U59 pin 2. That frequency when linked to U57 pin 2 will enable 9600 Baud rate.

J3 pin 12 on the Rev. D PCB is 499.2kHz for 1200 Baud - we don't need it so that pin can just be left as is.

If you do change those connections you should end up with J3 wired as per the Rev. D pinout above and the switching would be possible as below.

Hope that helps, Regards John.

[Bob63]

John,

    Thank you for your reply to my first post.  I haven’t got the printer working but I have been reading some interesting stuff. 

First, these are the set-up connections I made based on your write-up to try and get the printer to print:

 a.  U62 pin 8 has been removed from circuit to account for the W11 J3 pin 10 cut.  I don’t want to cut a trace until I know what works.  I removed pin 8 from a socket, put the chip in that and put that socket into the board socket.

b.  U57 pin 2 has been also been removed from the circuit . I removed pin 2 from a socket, put the chip in that and put that socket into the board socket.

c.  U57 pin 2 has been connected to U59 pin 2

d.  J3 pin 6 to J3 pin 9 connected with a jumper.

e.  J3 pin 4 to ground.

f.  J3 pin 1 to printer ground.

g.  J3 pin 2 to the printer RxD.

h.  J3 pin 3 to the printer DTR.

If you see anything wrong please let me know.

Second,  a link to an article that details how to increase the baud rate to 4800 via poke statements from Compute! Magazine, August 1981, Issue 15,  by Paul Lilly, titled, "OSI RS232 Port And The High Speed Printer Interface".

Start at page 101

https://ia600804.us.archive.org/28/items/1981-08-compute-magazine/Compute_Issue_015_1981_Aug.pdf

Third, attached instructions detailing how to wire up a Radio Shack Quick Printer 2 written by DEE Products.

Fourth, an attached pdf file showing the  locations of W3 and W11 on the board

 

Finally,  an attached jpg file of the SAMS C1P Service Manual (https://osiweb.org/manuals/C1P_Manual.pdf) schematic.  I aligned the 8 pages as best I could, labeled it and scanned the full schematic into one file.

5/8/24 inserted a cleaner copy of schematic.  It'is blurry on the website.   The website is converting a 5.5 mb file to 138.25 kb file.

5/10/24  Found out the dimension limit is 1280x1280 200kb.  No point in attaching stitched together schematic as it is unreadable and thus unusable. 

Bob

 

 

Edited May 10 by Bob63
Fixed missing attachements problem. Removed unreadable schematic.

[jbtech]

Hi Bob,

Thanks for that information. I've just noticed a couple of details which may represent a large part of the problem so far:

The printer I used in my writeup above was found as a search result for an 'RS232 printer'. It has 3 connection modes; USB, TTL and 'RS232'

 but the DFRobot printer you mentioned above apparently only has USB or TTL.

A TTL interface over a short distance would be fine and the RS-232 standard is well obsolete anyway being descended from earlier teletype communications.

Apart from different Voltage levels the RS-232 standard has inverted signals to TTL with a 'mark' or logic '1 being the more negative Voltage and a 'space' or logic 0 being the positive Voltage.

So virtually all of the details above including the addition of Q1 and Q2 to the Rev. B PCB are to connect to a printer with RS232 interface.

The printer I purchased was very inexpensive and readily available so it would be possible to continue, using a different printer.

Otherwise it should also be possible to communicate with the TTL printer, bypassing the Q1 output stage on the Superboard II PCB.

Tx data could be taken directly from U62 pin 12, lifting resistors R92 and R63 to isolate the inverting Q1 output stage.

DTR from the printer I presume is a logic high when ready so that should be OK to connect to the RS232 in at J3 pin 3, the inverted Rx data 2 appearing on J3 pin 6 then linked to the /CTS input at J3 pin 9.

To enable flow control it would still be necessary to check the W3 link to ensure that the /CTS input at U14 pin 24 is connected through to J3 pin 9 and not just linked directly to ground - as shown on the circuit diagram extract in my previous reply.

also:

Lifting U57 pin 2 and connecting to U59 pin 2 ( instead of U59 pin 14 as per original cct ) should be fine to set the Baud rate to 9600.

Lifting U62 pin 8 probably isn't necessary unless you are planning to use J3 pin 10 for Baud rate selection as per the Rev. D diagram, to allow switching from 300 Baud tape to 9600 Baud printer.

Hope that helps, good luck with it. Regards John.

[Bob63]

John,

  Thanks again for the reply.   I ordered an EM5820 off eBay.  Will keep you posted.

Bob