Buffee... It's alive!!

[Dale423]

Originally Posted by trixster

Ah, I’m talking about buffee. I paid for a beta board, have so far received nothing

Beta boards went out like 2 and half years ago. At least that is when I got mine. Might want to hit them up on that.

[Pyromania]

I’m blaming Mayor Wilkins for this, he always was a thorn in Buffy’s side.

[wiz12]

Buffee... It's dead!!

[QuikSanz]

Nonarkitten has her plate full of a years of stuff. She has said that this one is first then others. After comes Agnus replacement and down the line. Patience Fellow Amigan.


Chris

[FOL]

Originally Posted by QuikSanz

Patience Fellow Amigan.

Money has already changed hands. That causes massive problems, as recent comments show.

[kipper2k]

its not dead, while we are waiting we are working on the other custom chips, Gary is done, working, pictures on discord

[eXeler0]

Originally Posted by kipper2k

its not dead, while we are waiting we are working on the other custom chips, Gary is done, working, pictures on discord

@kipper2k are all the new chips direct drop-in replacements or there cases where we’ll need adapter boards?

[kipper2k]

Originally Posted by eXeler0

@kipper2k are all the new chips direct drop-in replacements or there cases where we’ll need adapter boards?

They are the same size as the original chips and will be direct drop in. Agnus is the tricky one but i tested the design for continuity of all 84 pins and so far so good, (there is a little bump on the chip where their router backs off from the PCB a little too soon so i will talk to them

[nonarkitten]

Originally Posted by FOL

Money has already changed hands. That causes massive problems, as recent comments show.

The sole reason I even entertained the idea of the beta launch was to try and get some help with testing and development and out of fifty boards, I've have zero commitments to help.

The boards were sold as a beta with no promises made or implied. In fact, it was very clearly stated to everyone who bought one that they'd be on the hook for using (rather expensive) JTAG programmers and cables, and following along with development. They were shipped blank deliberately and I explained this and why.

I have the feeling that they all just went to hoarders who wanted to be "first," didn't read anything and had no desire to help with the project at all.

I saw one person joke about porting Musashi to it so that at least it "wouldn't be worse than PiStorm," and even that never materialized.

I even posted the bootloader, a precompiled binary that people could run to at least test out, and only one person bothered (and ended up frying his Buffee).

The basic explanation of our current hold up. Our CPU (AM3358) has an external memory interface called the GPMC, the "General Purpose Memory Controller." This should allow the ARM CPU to essentially see the 68K bus as a 16MB memory region, thus mitigating the need to bit-bang or have any specialized hardware to access directly. More importantly, it makes reading from the 68K bus a single ARM instruction, and that's a big reason why we can get the claimed performance.

We found the GPMC had a rather fatal flaw that has been very difficult to overcome. It's "WAIT" mechanism is a full-halt of its state machine. This is basically (but not quite) the exact opposite to the "READY" behaviour of the DTACK. Some EMI's allow either mode, but TI's does not. This is why we had to add the GreenPAK -- to add a 'translation' between DTACK and WAIT. And so far, its proven to not work.

And I've been a little more than completely out of my depth. Hence the ask for help. But I'm still working on it, though. And I will continue to do so despite the naysayers.