ElEctric_EyE wrote:
I'm all ears for a better type of decoding though.
Code:
Broken external image link
http://www.laughtonelectronics.com/decoderimage.png
Here's your diagram again, but modified to take Phase 2 out of the RBA14 signal. I'm not terribly surprised you were able to make the old version work, but it does "break the rules", and is asking for trouble. From the SRAM's point of view, all addresses should be stable before CE goes true.
Quote:
So the idea of "Locking" 16K blocks, which is really bank switching, is not a problem? Even in the middle of a program?
Just be sure that when the Bank Select bits get altered, the program that does the alteration doesn't reside in the area being remapped. In other words, don't pull the rug out from under your own feet!
-- Jeff