Jumping back on board the digital train - part I
My first brush with a micro as programmer was as a schoolboy - my maths teacher had a National Semiconductor SC/MP board with LEDs and toggle switches. It could be programmed in binary. I was hooked.
About a year later Science of Cambridge came out with adverts for their Mk14, using the same SC/MP. This though had a real hex keyboard and calculator style 8 digit 7 segment display. I ordered one almost immediately I saw the ad. Trouble is, it seemed to take an age to come. Someone joked that Clive Sinclair's approach was to gather up the cheques and when he'd got a few thousand pay someone to do the design! I dialled SoC's number so many times chasing my order that its still engrained in my memory over 30 years later: 0223 311488.
When I got to uni, it was clear my room-mate was in a league above me - he'd designed a system with a Z80. That was a real man's processor, compared to my little boy SC/MP. I felt a tad threatened by his prowess. Science of Cambridge obviously saw the SC/MP had no future as their newest products also used the Z80. It was clear I'd backed the wrong horse. I was clear that I was absolutely not going to learn the Z80. I knew most of the binary codes for the SC/MP opcodes and hated the idea of going back to square one.
Fortunately, another door opened to me. My university lab work introduced me to the 6800. I did a project using the same chip a couple of years later. I sat in the library and pored over Byte magazine's series of articles about the 6809 - at last I'd found another architecture to love. Such elegance, such simplicity and power
I was even able to use the 6809 in my final year's project. There was a migration path too - upwards to 16 bits. Sat in the corner of the micro lab was a machine I could only dream of programming - a 68000. Undergrads were permitted to peer and drewl - it had no lid and was a bus-based design (S bus perhaps) so all the glorious internals could be gawped at. I resolved to get into the 68k world one day.
It wasn't long after graduating that yet another architecture came along to despise and ridicule. The IBM PC, using an 8088. This was from the same lineage as the abhorred Z80. How could IBM choose such an inelegant device? In my first real job I programmed 8748s - they were cool because they were one of the first real SoCs I'd ever come across - the memory and some peripherals like a timer, uart and GPIO were on the same chip as the CPU. The '7' in the number indicated EPROM - the whole chip could be easily reprogrammed. Mind you, the architecture was nothing special, just the self-containedness was really appealing. I built a guitar tuner using one of these which I still have somewhere - I don't know if the EPROM has retained its contents over the intervening decades.
A couple of jobs later I finally found a chance to design and program with the 68000. So many registers (I didn't write C code, only assembler) and almost orthogonal at that - I was permanently in machine code nirvana.
About a year later Science of Cambridge came out with adverts for their Mk14, using the same SC/MP. This though had a real hex keyboard and calculator style 8 digit 7 segment display. I ordered one almost immediately I saw the ad. Trouble is, it seemed to take an age to come. Someone joked that Clive Sinclair's approach was to gather up the cheques and when he'd got a few thousand pay someone to do the design! I dialled SoC's number so many times chasing my order that its still engrained in my memory over 30 years later: 0223 311488.
When I got to uni, it was clear my room-mate was in a league above me - he'd designed a system with a Z80. That was a real man's processor, compared to my little boy SC/MP. I felt a tad threatened by his prowess. Science of Cambridge obviously saw the SC/MP had no future as their newest products also used the Z80. It was clear I'd backed the wrong horse. I was clear that I was absolutely not going to learn the Z80. I knew most of the binary codes for the SC/MP opcodes and hated the idea of going back to square one.
Fortunately, another door opened to me. My university lab work introduced me to the 6800. I did a project using the same chip a couple of years later. I sat in the library and pored over Byte magazine's series of articles about the 6809 - at last I'd found another architecture to love. Such elegance, such simplicity and power
I was even able to use the 6809 in my final year's project. There was a migration path too - upwards to 16 bits. Sat in the corner of the micro lab was a machine I could only dream of programming - a 68000. Undergrads were permitted to peer and drewl - it had no lid and was a bus-based design (S bus perhaps) so all the glorious internals could be gawped at. I resolved to get into the 68k world one day.It wasn't long after graduating that yet another architecture came along to despise and ridicule. The IBM PC, using an 8088. This was from the same lineage as the abhorred Z80. How could IBM choose such an inelegant device? In my first real job I programmed 8748s - they were cool because they were one of the first real SoCs I'd ever come across - the memory and some peripherals like a timer, uart and GPIO were on the same chip as the CPU. The '7' in the number indicated EPROM - the whole chip could be easily reprogrammed. Mind you, the architecture was nothing special, just the self-containedness was really appealing. I built a guitar tuner using one of these which I still have somewhere - I don't know if the EPROM has retained its contents over the intervening decades.
A couple of jobs later I finally found a chance to design and program with the 68000. So many registers (I didn't write C code, only assembler) and almost orthogonal at that - I was permanently in machine code nirvana.
Total Comments 7
Comments
-
I love 68K assembler
It's so intuitive!! I had a go at 6502 and gave up, and I never really like X86 either. IBM mainframe assembler is just bizarre! Give me 68K any day 
Tony.Posted 26th April 2011 at 07:08 AM by wintermute
-
I still use a BULLET Z80 computer (CPM) and coupled it to an SST drive. This is surprisingly fast with 128K ram running assembler programs. Especially in timing sensitive applications.Posted 26th April 2011 at 07:18 AM by Tarasque
-
Totally! Except it really doesn't have a future, so I'm going on to talk about what's replaced 68k in my heartQuote:Originally Posted by wintermute
Give me 68K any day Posted 26th April 2011 at 09:37 AM by abraxalito
-
Yes - my biggest beef with the increase in clock speeds is that user frustration seems to have gotten worse. Computers get less responsive. My old Acorn Atom never hung like my PCs do and it ran at 1MHz (2MHz when I overclocked it).Quote:Originally Posted by Tarasque
This is surprisingly fast with 128K ram running assembler programs.Posted 26th April 2011 at 09:40 AM by abraxalito
-
Simply amazing! I was with Wave Mate when we built this board. Too bad I no longer have technical info on it.Quote:Originally Posted by Tarasque
I still use a BULLET Z80 computer (CPM) and coupled it to an SST drive. This is surprisingly fast with 128K ram running assembler programs. Especially in timing sensitive applications.
Are you using CP/M 2.2 or 3.0? We licensed 3.0 and sold quite a few boards with it. It was so much better than 2.2 but by the time it came out the IBM PC was way down the road.Posted 27th April 2011 at 03:34 AM by Dennis4SB
-
I wrote a LOT of 6800 code, not the 68K, and it's not the best architecture. When the 6809 came out I loved it but we moved on to Z80 and CP/M.
The 6805 was a nice embedded controller. Much better CPU architecture than the 6800. I forget what was embedded in the chip, timers and UARTs at least. We built a special terminal with the 6805.Posted 27th April 2011 at 03:53 AM by Dennis4SB
-
I agree with all this! Always had the feeling with Z80-style processors that I never really had a handle on all the tricks you could do with them, using the obscure, seemingly arbitrary, combinations of registers they'd built into it.Posted 8th May 2011 at 11:29 AM by CopperTop
