Hardware hackers of a certain age likely got started with microcontrollers via the RCA 1802 — a relatively easy-to-use processor that was the subject of several excellent articles in Popular Electronics magazine back in the late 1970s. [Al’s Geek Lab] has an interview with [Hugh Anderson], who saw the articles and eventually designed the HUG1802, which may be the first microcontroller kit designed and sold in New Zealand.
The 1802 was very attractive at the time since it was inexpensive, static, didn’t require exotic voltages, and had a DMA system that allowed you to load software without complex ROMs. He initially marketed a kit unsuccessfully until an Australian company convinced him to create a proper PC board — the resulting kit was sold to about 100 customers.
The HUG1802 reminded us somewhat of the Quest Super Elf since it had a keypad, a cassette interface, and even a TV output. The 1802 had a DMA-enabled chip that made crude memory-mapped video output. The computer eventually morphed into the ETI 660, which they talk about at the end of the interview.
A lot of people built 1802 computers back in those days. If you don’t have an 1802, but you have an Arduino… ell, there’s always emulation.
Wern’t the articles in Popular Electronics written by someone at RCA, maybe the 1802 designer?
Yep Joseph Weisbecker
“the RCA 1802”.
Sandia Labs negotiated an agreement to produce the ‘Intel’ 8051 as a rad hard part.
At the time Sandia/Intel concluded the agreement, the 8051 was on revision number ~528, I was told by a Sandia hardware engineer.
This engineer quit Sandia and went to work for Intel.
This engineer also told me that Intel got the 8051 design from RCA through a trade of a Intel-designed part.
Also learned from this engineer [deceased] that the design revision is printed on the bottom of the 8051 chip.
Intel 8051 had i/o ports prone to failure, I know from experience. AMD 8051 parts i/0 ports did not fail, I also know.
Some flash memory controllers continue to use the 8501 but most are using current technology controllers, I heard.
Like the ARM M series?
I built an 1802 kit in 1979 which also had 256 bytes RAM. I expanded the RAM to 8 kB, added a cassette interface, added a teletype interface. Then I acquired a working surplus KSR35 Teletype from the phone come and BASIC on a cassette tape.
With that setup I was able to write useful BASIC programs and operate using the teletype. I didn’t add the video chip though.
It had no ROM so every time I turned it on I had to enter a machine code bootstrap program in Hex on the Hex keypad (I knew it by heart after not long) to load BASIC from the cassette. In 4th year EE we were able to play with a small attache case-sized PDP11 which used a paper tape reader to bootstrap.
Still kept the computer for sentimental reasons, although when I moved overseas I had to give away the KSR35. Donated it to UCLA for their internet museum. It was a tank! Our basement flooded twice and each time after letting the KSR35 dry out, it just went on fine!
I’m jealous! I only had 4K and a KSR28. They were tanks though.
Wirewrapping those memory boards was certainly a pain!
The electric powered tools certainly helped.
https://hackaday.com/2014/12/18/wire-wrap-101/
“1802 fig forth c-h ting”.
“1802 fpga”.
1802 used in weapons system by Sandia Labs.
8085 rad hard too. But failed because of too many peripheral chips. At cost of ~100s of million dollars.
8051 rad hard project succeed?
Embedded controller forth for the 8051 succeeded. :)
“8051 fpga”.
2023 portable c [x86, ARM M and A, RISC-V, …] <$50 nanocomputer SBCs the future?
Along with 64, 32, 16-bit VM Intel MCS BASIC-52 too, of course.
There are not many books about the good old 1802.
I republished and cleaned up the book recently:
BMP802 – Brussels Microprocessor Publication – in 1980 the 2nd 1802 book
https://www.amazon.co.uk/gp/product/B09GWCVHGG/ref=dbs_a_def_rwt_bibl_vppi_i5
This processor is just very easy to understand.
And there is a very active COSMAC group.
One of the recent projects there was a ROM made of diodes.
Just for fun I built the 4 byte version: 7A 7B 30 00
2x cd4011 for address decoding
about 20 diodes
on a breadboard.
I could convince a friend to do a (reduced) version of a CDP1802 for FPGA – and running a Forth
https://www.amazon.co.uk/gp/product/B01N42VLJE/ref=dbs_a_def_rwt_bibl_vppi_i21
If I remember correctly, microForth was the first Forth for a microprocessor then.
I still own a Microtutor – just have to switch it on again.
“The 1802 was very attractive at the time since it was inexpensive, static, didn’t require exotic voltages, and had a DMA system that allowed you to load software without complex ROMs.”
In short, for a CPU, the 1802 was very humble, kind and forgiving.💙
On a side effect, these quality made it ideal for harsh environments (space or terrains with varying temperature/humidity), as well as countries with lower technological standards (unstable mains, poor power supplies, low quantity of non-defective ROM chips).
Love my elf, even with only 256 bytes of memory and individual LED outputs, which was all I could afford at the time. Originally built in 1979-80; rebuilt/re-wirewraped in 2014.
7B will be etched in my memory forever.
7B 7A 30 00
Good stuff, but I challenge the suggestion that this was the first New Zealand machine! I today dug mine out of storage and took photos, this project was started about 1972 before microprocessors were available in NZ (hence the 24 switches & lamps on the front panel, I was going to use the Texas Inst. 4-bit ALU slices) but subsequently realism set in and I converted to use the NS SC/MP 8-bit processor. I managed to acquire some parts from an old ICL mainframe and cut the connectors off, aralditing these to veroboard to plug into a backplane, and hard wired with self fluxing wire. I re-wired the keyboard diode matrix for ASCII, and my first bootstrap ROM was actually made from germanium diodes, later replaced with the early fusible-link chips. When I first powered it up I couldn’t see what was happening, so used a pair of 8-bit DAC’s to produce a 2-D program execution map on an oscilloscope- which immediately told me that the CPU outputs were inverted! Unbelievably, inverting these produced a working computer. You could slow the NS chip right down, I could single step to a led display. Output was later to an old Baudot teleprinter, then eventually (later!) to a TV set via encoder (this was an EA design for which I had PCB’s made, the only non-original element) and composite video converter. I had a working 8-bit paper tape reader (made from an old teleprinter part), an EPROM programmer and lots more – up to 20 circuit boards. I added to the NIBL language to produce almost an operating system. Lots of EPROMS and static R/W memory, then I imported some dynamic memory chips and got them going. My flatmate refused to accept that my machinet was a computer unless it could do Pi, so when it came out I interfaced the NI floating point processor via tricky level switching. Dates are hard to pin down, but I did present a significantly later paper at the NZ Nelcon conference in 1977, this was on a cross assembler for the SC/MP programmed in BASIC, so by 1975 my machine must have been well working, but development continued until about 1978 before I changed to 6809’s. Don’t think I can add photos here? Regards, Graeme O’Brien.