I can see from photos earlier in the thread that you're using 74AHCT logic chips for this keyboard circuit. When used on a breadboard the 'A' version of the logic chips can be problematic due to rapid rise and fall times of the outputs for this logic family.

Indeed I struggled with reliability of this same keyboard circuit on a breadboard using 74AHC logic chips. The same circuit worked perfectly after switching to 74HC logic.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I support Garth's statement above.

Also, what power supply are you using? I noticed in the schematics that there is no larger (eg 47u or 100u) bypass (are they called this?) Cap between gnd and vcc. I have them on every board where the supply hits the board and sometimes even some across the board, to decouple power fluctuations between boards and the supply.
(And now I'm looking for comments of our resident experts if that actually makes sense...)

_________________
Author of the GeckOS multitasking operating system, the usb65 stack, designer of the Micro-PET and many more 6502 content: http://6502.org/users/andre/

I don't mind reading material; I like books and collect vintage computer books. I have been eyeing High Speed Digital Design: A Handbook of Black Magic but have been holding off because of the high price. I like your second suggestion too. Here are the two ways I was thinking of hooking up power and ground to the ICs:

_________________
"The key is not to let the hardware sense any fear." - Radical Brad

Ah, more replies came in while I was typing!



That is good to hear. I thought that something like that might be the case after talking with Chad upthread. I don't have any 74HC, but I ordered some to try; they should be here tomorrow.



That picture is illuminating. I read the AC Performance Problems page about 3 times this week, thanks to other threads where people are discussing similar issues to the one I'm having. I always attach GND to GND and VCC to VCC at both ends of breadboard, but it still never occurred to me to attach both rails at each IC like that.



I have a 9 volt wall-wart; it plugs into a power regulator circuit on the backplane that has a 10uf and a 100uf electrolytic capacitor. It's a pretty standard thing, I think, using an LM7805 to generate 5 volts. That's the only place I have larger electrolytic capacitors; all the bypass capacitors are .1uf or .01uf.

_________________
"The key is not to let the hardware sense any fear." - Radical Brad

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I'll have to watch that. How about through-hole caps with leads < 1mm going into ground/voltage planes? This must come very close to or exceed SMT performance.

_________________
Bill

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Through-hole caps have been improving over time. The left most I used to use in the late 70s to mid 80s, then going toward the right most the ones I use to day. Lead spacing is only 2.54mm and they sit pretty close to teh board. Using planes in a 4-layer board you can get away with very sort leads (~< 1mm). Perhaps they will not beat smal SMT on a 4-layr board, but they are a lot better than what we had in the past.

_________________
Bill

I had to take a few days away from the project, because my cat got sick over the weekend. He is doing much better now, and no longer requires constant supervision, so I got back to it this afternoon. Replacing the 74AHCT logic with 74HC did make my old messy circuit work. But, not being one to take the easy way out, I rebuilt the keyboard circuit *yet again* on a different type of protoboard, using a ground return scheme like the one Garth described earlier. This did the trick, and not only is it working great, it's working with the 74AHCT logic, so Blue April is back to running at 8Mhz! Wheeee!

_________________
"The key is not to let the hardware sense any fear." - Radical Brad

Excellent news on both fronts!

I suppose this shouldn't feel like such an accomplishment, seeing as how this is the 3rd or 4th time I've built the thing, but...



Blue April now (finally!) has an I/O board. LCD output and keyboard input are provided by something other than breadboards! Here are a couple of side / back views. I kind of like the one with all the lens flare - I didn't do that on purpose!




Now I can get back to my regularly schedule project of making a video card and importing code from 1970s issues of Dr. Dobb's Journal!

_________________
"The key is not to let the hardware sense any fear." - Radical Brad

Great job! Glad it's working well. Did you end up going with the 74HC or 74AHCT parts in the end?

So, you have 3 backplane boards, and to connect them you made a little staircase!

I see you have gotten wire-wrap figured out. That's neat to see, glad that is working for you.

Any starting places for the video? As some have told me, "it is best to start with a working model". And they were right. I was having major difficulties, but when I used George's (gfoot) design, everything worked out nicely afterwards.

Great job again! Thanks for the updates and pictures.

Chad

Hi Chad,

Thanks for the kind words!



I did switch to 74HC while troubleshooting the old circuit, but with this rebuild all the ICs are back to 74AHCT. (Actually, if you look REEEEELY closely at the photo, one of the 595s is still an HC - I must have reached into the wrong parts compartment! - but I have since noticed and swapped it out, so they really are all AHCT now.



Yep. TEBL made a kind of bridge connector for extending the backplane, but it took up two bus slots and only provided one. I figured rather than making another PCBWAY order, I could just make a third backplane and solder on some upside-down male headers. Sacrifice two bus slots to gain three. Also the front backplane module can be swapped out for the RC Project Platform board - that's the one with the half-breadboard and 6522 on it.



It seems to be working OK; I followed Garth's instructions in the primer. There are still a few solder connections on this version, but all the bus signals are wire wrap to wire wrap.



Not like having a working Ben Eater keyboard circuit to copy. This will me first time striking out "on my own," as it were. I think composite is a pretty simple signal to generate; I have Gerry Kane's "CRT Controller Handbook," and also a vintage Motorolla data book that has a data-sheet for the 6845 along with some reference implementations to study. I do have a nice Sony PVM with a composite input, so I should be able to test as I go.

There is a VDU module for the RC6502, but it uses the 6847, which I don't particularly like (there's one in my Coco2). I also think I'll probably learn more using something a little more general like the 6854.

_________________
"The key is not to let the hardware sense any fear." - Radical Brad