Z80 Computer - Part 12 Thinking about VGA

I want to start looking at the VGA signals that will be used to produce a graphical output. Being one of the more difficult aspects of the project. Taking ideas from the Ben Eater graphics card, I build up the horizontal timing circuits.
I explain an issue I had when trying to use the circuit without the inverters, which resulted in the SR latch getting very hot. So putting the inverters back into the circuit I measure the timing signals with an oscilloscope and the horizontal timing looks good.
I then explore ideas to build up the vertical timing circuits. After failing with breadboard and a soldered prototyping board I decide to make a PCB. There are further issues with the PCB, but I do finally manage to get some colours on the screen.
Ben Eater's design can be found here:
eater.net/vga

[@pevkh8359]

I love your Videos

[@KomomeRyu]

Happy to see you back in te series!

[@SteveRaynerMakes]

I've had some problems. I actually re-did this video from scratch, a shorter version, higher quality with better explanations. However I lost a lot of footage and audio. So in the end I decided that I was not going to make progress unless I released this video as is. I'm thinking of scrapping the next video as well because I go into great depth explaining a circuit which turns out doesn't work. I think the next video might just have to be a short one showing an idea I'm exploring but explain that there are problems preventing me from making progress.

[@KomomeRyu]

@@SteveRaynerMakes I think a better way to go would an FPGA. That's just one chip that can be reprogrammed for any logic. But they are a little expensive and there is learning curve

[@SteveRaynerMakes]

@@KomomeRyu I think that is the way I will end up going though. I just need to get one and have a play around to see how they work. It's a little overwhelming knowing which one to get and what additional hardware/software is required.

[@zxborg9681]

Saw this from reddit z80 sub. You could do with some decoupling caps on your breadboard. Both bulk electrolytics (47-470 uF) and ceramics 100nF range. Dunno whether you saw other options for generating timing, but you could also just run all your counter values to an EPROM and program it so that bit 0 generates HSYNC. Bit 1 decodes counter reset at max value. Or any arbitrary pulse shape. Skipping all the counter inverters and the high fan-in decoder gates. Trade off logic for memory cost and better density. Easy to create the bin file with a simple C or Python or whatever program. Or even Emacs in hexl-mode :-) Register the PROM outputs with a register to avoid glitches as the memory array accesses in the few nanoseconds after the address change. But a fun project, looking good!

[@goofyrulez7914]

The Z-80 was my second microprocessor (after my CDP-1802). I had an early TRS-80, Mod 1.

[@colonelbarker]

I ended up using 74HC590 8 bit counters

[@phils_arcade]

Can you change the counter to something like SN74LV4040A-EP which is a 12 bit counter running on 5 volt. Also has surface mounting versions if you wanted to take the layout further. Can also get 12 bit inverters as well so each line can be inverted. Other option is running it through a microcontroller to handle all your sync signals.

[@SteveRaynerMakes]

I did come across that IC when trying to look for ways to reduce the chip count. However I discounted it because it is described as asynchronous. My limited understanding is that with asynchronous counters the updated count ripples though the stages, rather than all stages updating synchronously. Am I mistaken? Would an asynchronous counter work in this application?

[@phils_arcade]

@@SteveRaynerMakes No you're right, these do ripple, also realised you are using HC as apposed to LS in your circuits. Not sure about this one: SN74HC590A 8 bit binary counter, just came across it.