• Please review our updated Terms and Rules here

What was the first microprocessor trainer board?

alank2

Veteran Member
Joined
Aug 3, 2016
Messages
2,256
Location
USA
If I understand clearly, these boards were put out to make it easy for someone to use and/or develop for the microprocessors they wanted to sell.

What I find fascinating about them is that they are stuck somewhere between the control panel computer (no ROM) and a computer running a terminal based machine code monitor (ROM of some type) in that they usually have to have a ROM (of some type) to allow controlling them via a 4x4 keypad and often 7 segment displays.
 
I can't say which was the first but in the UK possibly the first one which sold in quite large numbers was the SC/MP based 'Science Of Cambridge' (Later Sinclair) MK14, which was sold as a self assembly kit and mainly aimed at people who were already electronics hobbyists.

However the British MK14 was itself a 'spin' of the enhanced (with keypad and display) form of the slightly earlier 'National Introkit' from the USA and used the same 512-byte monitor initially.

As the name suggests a Microprocessor Trainer board is usually aimed at teaching the user how to program a microprocessor - typically (but not always) at machine level, directly in hex opcodes, or in the simplest of trainer systems, via binary switches and LEDs. Later more sophisticated Trainer boards might include an integral assembler.

Boards which are offered (often free or at very low cost) by the chip manufacturer and intended to let would-be customers and developers play with the capability of the chip are more usually called 'dev kits' although these were more prevalent from the Flash microcontroller age, where the code is typically developed off-board on a PC and sent to it down a serial or USB connection. Dev boards usually have some rudimentary I/O such as a row of LEDs, a row of switches, a potentiometer, possibly even an LCD display but will not usually have a self-contained interface by which the user can enter program code, ie, they won't usually have a full keypad and display and monitor program like a Microprocessor Trainer does.
 
I thought the idea of a dev kit was to train the engineer to use the durned microprocessor so it can be evaluated. Else, why purchase one?

A reasonable question.

A dev kit / dev board effectively gives you the circuit you are planning on making half-built already, typically it will have onboard power supply, the clock, some kind of comms on and off the board, some indicators, some switches, maybe a potentiometer or two so you don't have to start with a bare chip and add all of those things to it before you can even get started. The developer may already be familiar with other parts in this 'family' of devices so the main aim of a dev kit really is to evaluate the specific device on the board for the purpose that the developer is considering using it for. If the developer has not written for that particular micro architecture before then of course there will be a degree of learning involved. I have quite a few dev boards for things like PIC chips and every project I ever made with those chips started life as a dev board with other parts that I had added dangling off it. When I'm happy that it works I build the complete thing as a standalone item and the dev board is free to be used to develop the next project based on the same chip.

I think a microprocessor 'trainer' as referred to by the OP is more likely to be a teaching / learning aid used to teach people with no previous experience of microprocessors a little bit about how they work at low level - I would put the aforementioned MK14 and National Introkit in this category along with the 'Microprofessor' and 'Emma 2' which was a 6502 based microprocessor trainer. I'm sure you can think of a few more in this category.
 
When the SIM4-01 was introduced, there were darned few people who had any experience with microprocessors! The same could be said of the 8008 dev kits. I'd say that any early SBC qualifies as a trainer. Of course, if one wants to restrict the field to "training", many of the technical schools had their own trainers.

But if we're aiming for the earliest microprocessor trainer with integrated output and display, I guess the PE COSMAC ELF would be on my list of contenders.
 
I believe the KIM-1 was the earliest mass produced example of the single board with integrated keypad and display style of trainer.
 
In my mind, a trainer needs to have enough of a UI that it can be controlled. One needs to be able to write bytes into RAM and be able to execute them at a minimum. Something that could be used without HAVING to hook it to a terminal or some other external hardware.
 
You do realize that computer trainers pre-date microprocessors. Some don't even involve electronics. For example: the Arkay CT-650. I recall that on a visit home, my old physics teacher showed me this new acquisition. His had round knobs, but was otherwise as pictured.

Even though it's not a single-board setup, I submit that the Scelbi 8H was more of a trainer than anything. Same for the Mark-8.
 
Last edited:
I'm not surprised. I'm just wondering when the concept of moving from a real control panel system with zero ROM to using ROM as a monitor and the CPU itself to control the system occurred. Maybe there is overlap.
 
The original Altairs were shipped without ROMs--you keyed everything in with the front panel switches. The CDC 6000 mainframes used a matrix of toggle switches to get things going. Earlier systems would simply read a card or tape record into location 0 and execute it.
 
I'm not surprised. I'm just wondering when the concept of moving from a real control panel system with zero ROM to using ROM as a monitor and the CPU itself to control the system occurred. Maybe there is overlap.

Well, that's the irony there, maybe; Chuck mentioned the Intel MCS-4/8 dev kits from 1971/1972; they came with monitor ROMs and teletype interfaces, therefore predating "control panel" micros like the Altair.

Dev kits that have a hex keypad and display, ala KIM-1 and friends, are actually closer in spirit to the TTY-based kind because of course they need a monitor ROM to drive the keypad and display.
 
After I retired my Altair, I moved to an Integrand box (still have both). Z80 CPU card with 2708 EPROM. No front panel, just a reset button. The irony was that when I installed Don Tarbell's floppy controller board, the little 16-byte bipolar PROM was sufficient to get the system booted, so I disabled the 2708. On the other hand the Heathkit H8 had a keypad and numeric LED display, but I doubt that it qualifies as a "trainer" with the given criteria.
 
1976. The May 1976 issue of Byte had an early review of it. I believe the MEK6800 from Motorola was a few months later. The COSMAC ELF articles started in August 1976.
Motorola had ads for 6800 in trade mags like Electronics in mid 1974 into 75 for their early dev boards. My MES6800 dev/sales demo board is dated 10/74. MEC6800A slightly later, both with bus buffers. Mid 1975 was MEC6800C aka MEK6800D1 w/o bus buffers and chips dated early 75. All with serial I/O. MEK6800D2 was successor version with hex keypad and led display in 76. Amazed both my orig D1 and D2 still humming along.
 
The idea of what exactly constitutes a 'Trainer' or an SBC is somewhat fluid which I suspect is one reason why there has never been a dedicated 'SBC' section on this forum. Some which otherwise fit the profile very well happen not to be on a single board, for example I have a classic hex keypad / LED output Z80 system which was sold as a kit by 'Maplin' in the UK and it meets the SBC / Trainer profile in every respect except that the hex keypad and display are mounted on a separate cable connected PCB. This was intended to make it easier to put the system in an enclosure with the keypad and display in a convenient user-facing position.

On the other hand you have systems like the original National Introkit which relied on a serial terminal for user input and output, but otherwise fits the 'Trainer' concept, you can enter code byte by byte, run it and see the results.

Each generation of computing system has often taken advantage of tools or accessories which people may already have access to, so the earlier boards used serial terminals to keep the computer PCB hardware itself to a minimum. By the early eighties it was still very unlikely that a home user would have a dedicated computer monitor but most households had at least one TV and probably a second TV, so most of that generation of home computers used a standard TV as a 'monitor'.
 
Motorola had ads for 6800 in trade mags like Electronics in mid 1974 into 75 for their early dev boards. My MES6800 dev/sales demo board is dated 10/74. MEC6800A slightly later, both with bus buffers. Mid 1975 was MEC6800C aka MEK6800D1 w/o bus buffers and chips dated early 75. All with serial I/O. MEK6800D2 was successor version with hex keypad and led display in 76. Amazed both my orig D1 and D2 still humming along.

On the 6502 side of the fence the MAI Jolt sometimes gets credit as the first commercial 6502 ”single-board” computer, predating the KIM slightly. And like the early Intel MCS systems it also relied on serial for I/O, not a front panel or keypad. MOS was obviously expecting people to use the 6502 this way, given they introduced simultaneously with it support chips that included a machine language monitor in mask ROM.

(Which makes me want to get grumpy again about a certain autobiography that claims to have invented the idea of interacting with a microprocessor by typing on a keyboard… sigh.)
 
Well, that's the irony there, maybe; Chuck mentioned the Intel MCS-4/8 dev kits from 1971/1972; they came with monitor ROMs and teletype interfaces, therefore predating "control panel" micros like the Altair.

This is what I was wondering. Why would they go through all the work and expense of a real control panel _if_ there was another option on those later systems like the Altair. I'm not saying I would have wanted them to as I like the "charm" of a control panel in some respects. Maybe they didn't have the expertise to build a machine language monitor at the time? I could imagine an Altair 8800 with a keypad/sev segment on the front of it which would have also been interesting. I'm sure that statement is retrocomputing blasphemy to some.
 
Expertise was certainly there. I think the front panel thing was mostly an attempt to imitate older minicomputers., which had origins in a time when something like monolithic UV EPROM didn't exist. 1702s were expensive when they first came out in 1972. MITS did eventually come out with an EPROM board (the 88-PMC in 1976) which, with 8 1702A EPROMs installed would give you all of 2KB.

I remember showing a co-worker my Altair 8800. Her reaction was "That's not a computer--that's a toy!"

Guess I showed her. :)
 
Back
Top