2 Mb Expandable memory card
Posted: Wed Jul 07, 2021 5:37 pm
This project was some what of a personal challenge. The journey began when I decided that I wanted to expand my memory in my Adam, and instead of buying a card I decided to download and build one I found online, just to find out that this card had major issues. While attempting to diagnose that card I began researching how the Adam addresses and accesses extended memory. When I finally decided that card was beyond help, I decided to design my own.
The design is for a 2Mb base module controlled with programmable Gal chips. The base board includes a 60 pin header and a pair of 2 pin power headers for extra power and stability. Onto this board one should be able to add more 2Mb expansion boards increasing the total capacity of the system. I estimate that in an Adam memory module one should be able to fit 2 expansion boards allowing for 6Mb total expanded memory. If one has an expansion module 3 with some minor cable management I think the total amount of memory is to be determined by how far the address and data lines will travel. My personal guess is 4 to 5 cards but until I can start experimenting I will not truly know.
This project was challenging, in that I had a very hard time getting bread board circuits to work consistently. My theory is that due to the delicate timing required on these boards the combination of jumpers and questionable breadboard connections limited my testing abilities. I ended up testing it in phases first getting a basic memory bank to work, and them separately testing the programmable control circuit. The big issue with this is I still have never been able to test the circuit as a hole.
Now with the first prototype board built I eagerly look to begin testing, unfortunately the programming clip I ordered online, that was advertised for SOIC packages was actually for DIP packages making it useless. I have ordered the correct clip and am now patiently waiting on it to be delivered.
I have already found a couple of missing traces, and the SOP footprint for the memory chips ended up being too small. I have completed a re-design of the board, but am waiting to send it off until I get the programmer and make sure there are no more issues.
I would have to say that one of my biggest concerns on this project is that I can't truly trust the memory checkers that are currently available to proof a prototype board, and don't want to sell a board until I feel comfortable that it is truly working as designed. I will discuss this more tomorrow when I do the write up on my first attempt at Adam coding, A memory checker for questionable boards.
The design is for a 2Mb base module controlled with programmable Gal chips. The base board includes a 60 pin header and a pair of 2 pin power headers for extra power and stability. Onto this board one should be able to add more 2Mb expansion boards increasing the total capacity of the system. I estimate that in an Adam memory module one should be able to fit 2 expansion boards allowing for 6Mb total expanded memory. If one has an expansion module 3 with some minor cable management I think the total amount of memory is to be determined by how far the address and data lines will travel. My personal guess is 4 to 5 cards but until I can start experimenting I will not truly know.
This project was challenging, in that I had a very hard time getting bread board circuits to work consistently. My theory is that due to the delicate timing required on these boards the combination of jumpers and questionable breadboard connections limited my testing abilities. I ended up testing it in phases first getting a basic memory bank to work, and them separately testing the programmable control circuit. The big issue with this is I still have never been able to test the circuit as a hole.
Now with the first prototype board built I eagerly look to begin testing, unfortunately the programming clip I ordered online, that was advertised for SOIC packages was actually for DIP packages making it useless. I have ordered the correct clip and am now patiently waiting on it to be delivered.
I have already found a couple of missing traces, and the SOP footprint for the memory chips ended up being too small. I have completed a re-design of the board, but am waiting to send it off until I get the programmer and make sure there are no more issues.
I would have to say that one of my biggest concerns on this project is that I can't truly trust the memory checkers that are currently available to proof a prototype board, and don't want to sell a board until I feel comfortable that it is truly working as designed. I will discuss this more tomorrow when I do the write up on my first attempt at Adam coding, A memory checker for questionable boards.