(not forgetting the 6809E's ability to share the address bus with another bus master by tri-stating its own) If you're looking for a good assembler for the 6809, then there is a very good one here: www.sbprojects.net/sbasm/ :-)

Thanks for the encouraging comment! Makes it worth making up some more in the future :-)

I guess it all depends what one was 'brought up on'. I started out on micros' when the 8080 was already out there, along with the Fairchild F8 (remember that??) then the product line changed over to the Z80. (no more separate clock chip and system controllers) Hence I was working with these two when I saw my first example of a Z80 board in 1978. Whilst I can say that each different 'type' of microprocessor series definitely has advantages over the others, we all have our favourites! I only worked for a short while on the 6809, so in reality have had very little experience experimenting with it.

Yes, they were all very different. I started on 8080, then Z80, then 6502 (a real shock after those first two!), then did some work on the 6809 which I would class as the best "teaching" architecture of the time. Now it's AVRs and Cortex Ms. It's been quite a ride!

A nice variety to have worked with :-) If you're interested in the more unusual types that came and went, you might be glance at our video on the Philips 8400 series of microcontrollers. A very strange beast, but actually something really useful for experiments, because of it's 28 pin piggy - back socket. We used one in a 'Radar Room' morse code transmitter project because there was so llittle space left under the chassis...

Speed was not the problem, it was not so successful because of its high price. I agree with the joy to program. To me the programming model + addressing modes made it the most beautiful 8/16 bit MPU ever.

Good question! If one does an overview of the most popular 8 bit micros of the time, (all using 40 pin packages), then one can easily see the differences between what features and facilities are considered to be the priority for that particular cpu. I believe that the main issue was probably the limitation of just having 40 pins to play with. Drop back a year or two into the middle of the '70s and we saw the Fairchild F8 with its system of several 40 pin packages for each 'cpu system', then there was the 8080 in a 40 pin package which also needed an 8224 clock chip and an 8288 system controller in order to make a working cpu. The Z80 being the logical step forwards for this particular one because of the instruction set compatibility. (8080 to Z80 but not Z80 to 8080 unless one only used the compatible 8080 instructions...) An interesting history lesson I guess.

Sorry - my error it's a 8228 system controller for the 8080

The 6809 had a companion paged MMU, the 6829, which expanded the address range to a whopping (at the time) 2 MB. Pages were 4 kB, so 16 pages were mapped to 512 physical pages. (Never seen the 6829 in the wild though; only datasheet I've ever seen was "Advance information")

Thanks :-)