Really nice and interesting project! Open source Hardware and Software. 

Many people can use it to learn in-depth how computers works.

Thank you very much for sharing!

Thanks! This project has been a blast, and still is...

It turns out there is a Gigatron Service centre in Belgrade. How cool is that! https://forum.gigatron.io/viewtopic.php?p=115#p115

We have kicked off a Gigatron user forum: https://forum.gigatron.io

Hello. I have a small question for creators. This design seems damn effective and useful, but I still have to wonder: why exactly something like this wasn't built in late 60/early 70 despite TTL logic was avaiable since 1966?

Two reasons. First but not foremost is that the RISC architecture was not recognised as commercially viable until much later, like 1985 or so. Doing RISC drops your component count by an order of magnitude. Second is that the real hero in this design is the RAM chip. That chip allows for software to do a lot of stuff you would normally do with circuitry. Dating it: if the design were to be retrofitted for slower TV signals of the day, something like this might have existed around 1985 considering part availability. (Mind: slowest VGA speed = 4x TV speed). The CPU-part could still be made from early 1970s logic. The total system would have been expensive because 32KB SRAM (not DRAM) at that time was expensive.

I have scored some amazingly cool milspec SRAM chips from eBay that are from 1988. They *WORK* in the VGA version! In fact they are already 3 times faster than required for that (25ns while 70ns needed). They must have cost a fortune back then.  I did go through a lot of old BYTE magazine's component ads: you can't find such amazingly fast specs in the commercial offerings at the time... Nothing even close.

Thank you very much for your answer, Marcel. So, it was virtually impossible to build in 70s, wasn't it? May be use magnetic core memory instead?

About TV signal - may be it's possible to DIY assemble a simple display device with Gigatron resolution 160*120 (even monochrome) on the 80's technologies? This would solve a TV problem.

And the second question. How hard is it to code BASIC interpreter on this machine? It does have it's own assembler, AFAIK.

The same question applies to OS.

"First but not foremost is that the RISC architecture was not recognised as commercially viable until much later, like 1985 or so."  --- I suspect you're right about this, despite the fact that the clearly *were* people in the industry who knew this was the best approach.  Seymour Cray had been building designs that were effectively RISC systems and which dominated the high performance computing world from the mid 60s right up to the end of the 80s, but nobody else really seemed to understand what lesson to learn from that.

"Second is that the real hero in this design is the RAM chip.  [...] The total system would have been expensive because 32KB SRAM (not DRAM) at that time was expensive."

For reference of how expensive: I found an advert from Dec 1981 (which would probably have been about the last date a machine like this could have launched and been commercially successful -- any later and it'd have to compete with the C64) that priced 70ns SRAM at $30/4kilobits, so you'd be looking at about $2000 worth of RAM to build this machine at the time.  That said, I wonder how well it would work with, say, 8KB of fast SRAM for critical parts of the system and 16-64KB of slow DRAM for bulk storage, which would put the build cost at approximately the same price as the retail cost of a C64 at launch.

With pure 1970s components the CPU section could certainly be built using 74LS parts, but something else has to be done around the RAM. At least make it DRAM, slow down the clock to match (which might be ok for slower TV signals). It is difficult to really go back to that era, because RAM chips of those days are practically unavailable now. So it will look a bit different, but either way still be expensive for the 32KB RAM capacity (even if it were DRAM).

About BASIC: it is just a lot of work. Microsoft could become a company precisely because they had done a BASIC and the microcomputers of the day needed it. This project is a bit in the same state. Unfortunately, Bill Gates is doing different things nowadays! There is a proper assembler already written by user at67 on GitHub. It does both the 8-bit native instruction set as the 16-bit vCPU instruction set.

You want several steps:

- Direct or indirect keyboard hookup. Indirect is easy, but a bit cheating. Direct hookup is a harder nut to crack (at least: if you want to stay with the minimalistic spirit of the rest of the system. If not, just use the Arduino).

- BASIC parser/interpreter. Several weeks of full time effort for an "integer basic".

- Some kind of screen editor

If you go all the way, you need to add:

- Floating point package

- Program save capability. (Load is already there.).

Thanks again. So is it possible to code some network drivers to connect this thing to Internet? At least it seems capable to support some kind of relatively easy protocols like IRC or whatever.

@Julian : I'm interested in that 1981 reference to 70 ns SRAM, because it helps date this project. Do you have a URL or can you scan a copy for me if it is from a magazine?

Kit worked great, placing the chips in position on the foam like that simplified things hugely and the build manual was very clear - my only soldering error was leaving one pin unsoldered which resulted in a strange offset on the VGA signal, till I spotted it.  I'd be keen to learn more about exactly why it is that for instance this clump of 74xx chips are the ALU and that clump are a register, etc - once that kind of info is made available it could be a great teaching resource along the lines of the excellent Elements of Computer Systems / From NAND to Tetris (https://www.amazon.co.uk/Elements-Computing-Systems-Building-Principles/dp/0262640686) which @monsonite once pointed me towards.

Might I suggest Google Groups as a quick & easy way to get a forum community up?

Hi Bob, sorry I missed your update, but I think I saw it on Twitter :-)

We recognise the usefulness for a forum, but we're super busy so we're really in need of helping hands here, otherwise it isn't going to happen anytime soon. If somebody starts one and I like it, I will join threads I'm interested in. I have to say that I find Google Groups disappointing from user point of view, and for this project there are builders and tweakers/hackers and those threads shouldn't mix. The bbCode type of forums are so much nicer in presentation, and they can support two sections under one roof: Assembly and Hacking (no need to divide it up further, I would be against that).

But then also beggars can't be choosers. Now some programming discussions happen on GitHub for example.

Google Goups seems to work fine for other similar kits / kit makers e.g Oscar Vermeulen's awesome PiDP-8/i is here https://groups.google.com/forum/#!forum/pidp-8 and soon to be joined by PiDP-11/70 here https://groups.google.com/forum/#!forum/pidp-11, Spencer Owen's excellent Z80 kit RC2014 is here https://groups.google.com/forum/#!forum/rc2014-z80, Chris Davis' Altair-Duino is here https://groups.google.com/forum/#!forum/altair-duino, Julz' Forth-based Fignition is here https://groups.google.com/forum/#!forum/fignition, etc - I imagine a number of your customers / readers are, like me, already denizens of these places.

I'm aware. Let me say I have just have no real positive experiences with Google Groups. The groups I know and sometimes check out feel disconnected to me and the overview and content presentation is poor, as if it is still running on Usenet. Also no inline images, no markup. I find it frustrating to use for technical content, and therefore groups is not something I would personally frequently checkout and become like a home.

We could be looking into bbPress as forum plugin for on the main site... Any experiences with that? Good or bad?

We're looking at phpBB now... Either self-hosted, or we're going for a rental scheme.

Interesting video from 8bit guy.  It's sort of extraordinary that the gigatron outperforms a VIC 20 using only discrete logic gates for a CPU, but the trick is using modern memory chips & cost.  If it just drove a serial port instead of driving a CRT, it could run much more powerful demos.  8 bit guy was hard on it because he only saw the default firmware spending all its clockcycles driving the CRT.  

We absolutely love his episode! We don't think he is hard on it at all... He presents the project in a superb manner. As he mentions at the end, the programmability is a matter of software, and it is the next phase of the project. The Apple-1 didn't come with a built-in BASIC either, for the same reason: it takes time to build a software environment.

Such a cool project and also very extensive!
Ordered the kit and built it this week: https://twitter.com/BrruuD/status/978355534033227777/photo/1

Now for diving into the mechanics and figuring out how everything works exactly :-)

So cool to see it work!

I'm guessing the tail end of "On The Turning Away" heard in the bouncing ball video is an artefact of something playing elsewhere in the background and not being played by the computer itself?

Still, damn good effort for purely TTL.

Thanks :-) It can make proper bleeping sounds though. It has 4-bits sound output (16 levels) and the default loop squeezes 4 software channels through that. They can synthesise pulse, triangle, sawtooth, etc. It sounds crappy but charming, just right!

Given a lot of music was constructed with those four primitives… that's probably just fine.  The SID chip does similar with 3 voices.

My only concern here is look out for the DMCA trolls, while it's clear to us (and anyone with intelligence) that the music is just co-incidentally playing in the background … some lawyer might have a momentary lapse of reason and declare the entire video a derivative work. :-)

Anyone attending 34C3 in Leipzig this week: we will be strolling around, and of course we'll bring Gigatron prototypes!

Hi Marcel,

Great work!  Faster, simpler and cheaper than the original IBM PC - love it!

Is there a schematic available for your design?  I'm interested in the final approach you took for your ALU.

Interested to hear some more about the virtual machine you hinted upon.

I have shared your work with anycpu.org

Well Done

Ken

The full schematic might appear in some later stage as it is quite a mess in need of serious cleanup and I'm working on software the upcoming weeks. But the ALU is described adequately elsewhere: the "Logic" stage comes from here  http://www.6502.org/users/dieter/a1/a1_4.htm  (bottom of the page) and the ALU uses two of those with two 4-bit adders, as described here: http://www.6502.org/users/dieter/a1/a1_6.htm  except that my design does without the 3rd stage that only provides right-shifting. So in total it is 8x'153 and 2x'283 with 5 control signals going in. Dieter's pages suggest that you need 8 signals, but most of them have the same value for the common operations. The carry-out goes back into the condition decoder where it acts as an allzero-indicator during jumps. There is no flags register as it costs too many chips for too little added value. I do right-shifting in software using the "single-instruction subroutine" technique described in the log on pipelining. This computer doesn't need shifting that much because the graphics layout is 1 byte = 1 pixel.

Hi Marcel,

I'm suitably impressed with the elegance of your design that I have taken time to transcribe it to EagleCAD, so that I can get a better understanding of the control and data paths.

After pondering your Control Unit schematic for an hour yesterday lunchtime, and how it decodes the ALU instructions and generates the various addressing modes and jumps, I was happy that I could figure from the bus and dataflow diagram how most of the rest of it worked. 

Now that the bulk of the design is captured, I'm figuring out a way of using an ATmega1284 as an easy to update ROMulator device, running code from faster RAM and a way of extending by adding more VRAM.

When you consider that the first Sinclair ZX80 machine took 18 TTL chips, plus Z80, ROM and RAM to create a monochrome TV picture of 32 x 24 8x8 characters, for just another 18 chips they could have replaced the Z80 and had something that produced colour graphics and ran at 4 or 5 times the speed.

regards

Ken

I don't think schematics are ever really "finished", but the schematic that the layout was made from would be most helpful even if it is a mess.  I would like to see them because I would like to try to recreate it in an FPGA.  You could post them now, and then replacement them with a cleaned up version once you have it finished.

Amazinger :-D

I like it! 

Every homebrew computer should have video output, even made with TTL!

Have you seen my project? It is quite similar (custom CPU architecture, video output) even if I used an FPGA  instead of TTL components.

https://hackaday.io/project/18206-a2z-computer

It's fully documented and open source. 

And welcome on the "homebrew CPU Webring" !

Nice! FPGA's are a step up for me. Clearly the TTL and the RAM size are limiting my video resolution, but it is also cute in its own way.

Consider one or two AMD2901 or a single CY7C115/116/117 -> More computer, fewer chips.

Thanks. I can find the first. I do have ALU chips available, I just don't feel like using them in this build. Do you have data sheets for the second series? I have trouble finding out what it is. Some memory? I plan on using 62256 for RAM and 2x AT28C256 for ROM, with a possibility to extend the memory using the AT27C1024.

Good luck Marcel.