16MHz mod & current thoughts

[exxos]

While the Alpha needs testing with the 16MHz mod, The problem is its really needing the later SMT M MU & GLUE as they should stand better for overclocking. This was the idea of the mongrel to use SMT chips, but of course harder to solder and I would really like to keep the PLCC chips in there for now.

My thoughts a few weeks ago was to possible do a expansion type header so the SMT glue & mmu could be plugged in as a replacement to the PLCC chips, without forcing the mongrel to only use SMT.

A expansion board would also make it easier ultimately to move to FPGA for a integrated glue & mmu.

Then I thought this morning, that what I should do , is keep the PLCC glue & mmu and add a expansion header to move to the SMT chips, but also this actually opens the door to integrate the 16MHz mod onto the SMT expansion board! This will actually speed up the mongrel as the 16MHz mod can be worked on at a later date.

For example, I still need to send Phil the MMU adapter board to try, still need to the the glue lib etc etc so it would help speed things up by doing this work at a later date. Also there have been a huge number of changes on the mongrel , so I think its been complicated enough already.

We are in March already and no progress has been made since xmas. I will add in the 72pin simm socket and more IO headers for things. But I think doing any more mods is better waiting for a 3rd gen board. I've got a lot of work do do with the physical layout of the chips and holes yet. I wanted to plot how to mount ultrasatan internally, I will add some holes, but getting connectors to mash up and align is a lot of work.

So I think just adding the header with the thought of just using a ribbon cable for now will help speed things up. After all, this isn't a final board, its just fixing all the Alpha issues and adding in some more stuff to push the project forwards. I suspect we will likely ramp up to a 5th gen board before we get to a more final board. I think getting things working and functional is the main thing more than anything at this point. I think trying to do a more final board at this point is just creating a lot of work all at once.

[PhilC]

As I've said before, I'm happy to help wherever I can.

I've got an ATF programmer coming this week, ready for the ide board and whatever else we all decide to use it on.