MiniMig EAGLE schematic & flying board

[AJCopland]

Hi everyone,

Got a couple of questions.
1. Does anyone have an EAGLE v4.16 / v5.10 schematic and/or board layout for the MiniMig. I have a schematic file... but I have no idea where from and nothing I have will open it :-? Could it be from a gEDA version or something... who knows.

2. I have a desire to build a small 'flying' board. What I mean by that I that it is a seperate thing like the "Gayle chip adapter" or "IDE-Fix Express Buffered IDE" that clips over another chip... However, there's no chip for it to clip over and I'd rather not put one there just for it to act as a mechanical bridge!
Instead I'd like to use the bare pads that are meant for the chip which would be moved onto the flying board.
I just need some way of doing the physical connection itself from the flying board to the original pads. Does anyone have any ideas?

I'm not going to say what I'm hoping to build because we all know what happens when anyone actually says they're going to do anything :-D Oh but I would hasten to say taht it will never be anything as useful as either of the two example I gave!!! :lol:

Andy

[AJCopland]

Ah well might as well put out any ideas I have had.

1. Mount the current chip in question, get another one and a socket and do it the usual way. This somewhat simplifies things... it's also quite wasteful and dovetails horribly with the idea of prototyping the design this way before making changes and ordering new pcbs.

2. Take a small cube of... wood, or something. Run metal pins up the side and put it in place as a dummy or standin "chip", then get a socket for the underside of the flying-board and proceed as planned. This is better in that it's less wasteful, and the design won't need to cope with disabling or ignoring whatever chip it's clipped over.

3. erm... I dunno that's why I'm here :-D

Andy

[Floid]

I don't really know what you're trying to do, but if it's a DIP and the boards are drilled, use a wire-wrap socket placed through drilled pads in the riser to provide pins into the pads on the host board?  Those milled legs can be pretty sturdy.

Looking at DIP IDC connectors might give you other ideas; this company got creative with single-row pin headers, which seems to work.  Of course if you need to clip over a PLCC or something, you'd need something ridiculous like this.

[rkauer]

A bearded man from another forum have board you want! It is a gender changer (68k PLCC to DIL) It clips over the original 68k chip (or the other way), allowing you to use a IDE controller over the second board.

 Yes, the Minimig become a three boards "sandwich". :-D

 Hmm... Try EAB. ;-)

[AJCopland]

I've seen that one, its Zetr0 design isn't it.

I think I will try EAB as well, thanks rkauer.

[boing4000]

Nice idea but don´t forget that the Minimig has an asynchronous external bus. Insinde the FPGA the RAM and CPU are handled synchronous to the "Amiga chipset". An external Gayle chip would run at far more speed then the internal Amiga (zorro) bus. Even the CPU do not hold by signal as the real 68000 in any Amiga is doing. Also the S-RAM is working at aprox. 50-60ns (depending on the chips and core settings), that is much faster then in a real Amiga.

Best would be to code an Gayle chip inside the FPGA and let it control the timing. There are a few unused lines available on the chip. Or construct a new board with special IDE connector and hardwired line to the FPGA as the DEx Boards do.

Also its possible to simulate a harddrive via hardfile on SD-Card. Unfortunately I dont have any luck in compiling the PIC C-sources. If I would, even adf write support could work just now.

[AJCopland]

@boing4000
Ah you may have misunderstood my use of those examples ;-) I simply listed them as "things which clip over existing chips" because I'd like to avoid saying what I actually want to do :-D

Would like to avoid this being one of those threads that sails off in 15 different directions with everyones different dreams of "what if xyz?!?!?!" and questions of "Why would you do that it's a stupid idea!" etc :-(

Just hoped for an answer or some ideas and then to get on with seeing if it can be done in peace.

It has been suggested via PMail that i could use ribbon cable which is currently the best sounding suggestion I've had.

Andy

[wolfchild]

@boing4000

Nice idea but don´t forget that the Minimig has an asynchronous external bus. Insinde the FPGA the RAM and CPU are handled synchronous to the "Amiga chipset". An external Gayle chip would run at far more speed then the internal Amiga (zorro) bus. Even the CPU do not hold by signal as the real 68000 in any Amiga is doing. Also the S-RAM is working at aprox. 50-60ns (depending on the chips and core settings), that is much faster then in a real Amiga.

So, this means that one cannot add memory mapped peripherals directly to the 68k?  You know, the kind that piggy backed to the socketed CPU on the old Amigas...

Of course, I'm considering issues as 3.3V and actually connecting to the MiniMig's 68k as 'solved' for the scope of this question.

Regards,
Edwin

[CD32Freak]

@AJCopland
Something like this?

:-D

[AJCopland]

YES! Exactly like that! :-D

Do you have the file(s) and would you mind sharing?

Andy

[CD32Freak]

YES, of course! :lol:
I've used the tool gbr2scr (which is part of EAGLE PCB Power Tools V4.16, you can download a free version here) to convert the Gerber files into a Eagle SCR file.  It's a long process, because it is limited to 3000 lines of Gerber data, so I had to split the Minimig gerber files in 4 seperate SCR files. Nevermind, without further ado:
http://rapidshare.com/files/126593489/Minimig11_TOPTSK.brd.html
:-D

By the way, it's only the top layer and not really usable yet, because there is no schematic file, but you could use it as a template :-)

[AJCopland]

Cheers for that, I'm quite surprised there's no gschema to Eagle converter actually since gschema is open source and the ".sch" file itself is plain text.

Andy

[boing4000]

wolfchild wrote:
So, this means that one cannot add memory mapped peripherals directly to the 68k?  You know, the kind that piggy backed to the socketed CPU on the old Amigas...

Of course, I'm considering issues as 3.3V and actually connecting to the MiniMig's 68k as 'solved' for the scope of this question.

Regards,
Edwin

As in my understanding out of Denis's posts and the comments in the source code it is not possible that way as in a real Amiga. The external bus of the Spartan fpga is splitted into RAM and CPU part. The CPU is working nearly the same as in a real Amiga (7/14MHz) but as mentioned it is not in sync and the _hold line is combined with _reset line on the PCB. The CPU is always operating (even if chipram is adressed or DMA is active) and inside the fpga all signals are buffered and resynced with the actually chipset.

The S-RAM is indeed operating at very high speed and one can not simply add some D-RAM. Adding comparable S-RAM chips would work, only an additional adressing line has to be used (there are sill some free fpga lines).

So we are in an fpga environment and need to think of "hardware inside a chip" to create functions and "pluggung" additional Amiga hardware to the "programmed system".
Outside the fpga it is really another world running :-)