[v1.0] MECB Grid 22V10 Card

[Editor]

Introducing a new family of plain Grid Cards, to join the existing Prototype Card family.

So, you now have a choice of six different Cards for your MECB custom Card build or prototyping needs.

Inspired by getting out my Wire-Wrap tools, to re-live the wire-wrap experience, and also by feedback from those who preferred a plain grid style of prototype card.

Read all about it here: Minimalist ECB (MECB) - Grid Cards for Prototyping and Wire Wrapping

This topic is for discussion related to the Grid 22V10 Card (with the more versatile 22V10 glue logic PLD).

[djrm]

When the grid cards were first announced I took the opportunity to use my PCB maker coupons and ordered some of each of the three kinds. The plain grid one arrived after just a few days but the other two including the 22v10 version got held up multiple times but they arrived together yesterday by surprise ahead of the tracking update.

I had been waiting for the 22v10 version boards to use with a couple of 6522 VIAs to tidy up my AIM-65 clone. Having a few spare 20v8 GALs I decided to use one in this design, I started with Greg's example which uses pin mappings different from those I'd been using on another prototype board. I found that pin 23 is not able to be used for output with this chip, so I borrowed pin 19 for CS0.

The reverse side wiring is to complete yet, when finished this board will provide access to the serial port and a connection to the 1541 disk interface and anything else which uses the VIAs

Thanks for the board update, greatly appreciated, David.

[Editor]

I started with Greg's example which uses pin mappings different from those I'd been using on another prototype board. I found that pin 23 is not able to be used for output with this chip, so I borrowed pin 19 for CS0.

Hi David. It’s great to hear the new Grid Cards have helped.

I do admit that the lack of 20V8 GAL backward compatibility (pin 23), is an oversight on my part.

With the ATTF22V10 being in production and relatively cheap, I didn’t even consider to review the data sheets for the previous generation devices, for backward compatibility.

If I had, I would have avoided the use of pin 23 for one of the CS output.

But I imagine having now published the MECB pin use standard, it’s possibly too late to change.
Unless, you think there are still plenty of 20V8 family chips in enthusiast’s parts drawers, to warrant reconsidering a pin standard reallocation?

After all, it is still early days, and it would be nice to ensure a solid prototyping foundation for future MECB use (which was the intention of expanding the prototyoing card family). Unless you think just focussing on the current ATF22V10 generation is sufficient?

Let me know! (Yes, I know, I do "overthink" things... but then again, I completely missed the pin 23 backward compatibility issue).

[djrm]

Hi Greg, its not a problem. there are plenty of other limitations when trying to use the 20V8. I was surprised it compiled and worked on my initial board. I tried using it in a latch application without success and had to revert to the 22v10, but thats another story.

One thing to consider with future boards and updates is to make it easy to repurpose pins by making it possible to cut traces to re-assign pins in a simple way, i.e. have pads each side of a place to cut a track if needed.

Best regards, David.

[Editor]

Hi Greg, its not a problem. there are plenty of other limitations when trying to use the 20V8. I was surprised it compiled and worked on my initial board. I tried using it in a latch application without success and had to revert to the 22v10, but thats another story.

Okay, I'll leave it as is, just supporting the 22V10 (as already noted on silkscreen). I had been thinking about making the change, but realised it would also involve re-routing 3 KiCAD projects (Prototype, Grid, and Template).

One thing to consider with future boards and updates is to make it easy to repurpose pins by making it possible to cut traces to re-assign pins in a simple way, i.e. have pads each side of a place to cut a track if needed.

Do you mean specifically the I/O pins of the PLD? All pins? Also, just the protoype cards, or on regular cards as well?

I could add some closed solder jumpers adjacent to *some* of the PLD pins (in future boards / revision updates), to make trace cuttting / re-linking easy.
But doing this for all pins might create some routing space issues (and I don't really want to relocate the PLD, taking up more PCB space).

[djrm]

Do you mean specifically the I/O pins of the PLD? All pins? Also, just the protoype cards, or on regular cards as well?

From time to time I have had to modify boards by cutting tracks, things which spring to mind are IRQ pins, PLD inputs, memory address pins.

In the case of the new 22v10 boards, some of the inputs are used only in some situations, here an option to use these pins for other purposes would have helped, here I re-purposed the otherwise unused r/w pin. I have a use in mind where more than four outputs would be needed too.

Who's to know where the scalpel will fall next ... David.

[Editor]

Do you mean specifically the I/O pins of the PLD? All pins? Also, just the protoype cards, or on regular cards as well?

Who's to know where the scalpel will fall next ... David.

Okay. I'll keep this thought in mind when making future revisions / new boards, and see what I can accomodate (within reason).

I suspect this will also be a candidate for Murphy's Law. As in, I suspect your scalpel will still fall wherever I haven't added solder jumper pads.