View Single Post
Old 09 January 2017, 16:12   #29
Pat the Cat
Banned

 
Join Date: Dec 2016
Location: Nottingham, UK
Posts: 482
This is a puzzle, but I think I can explain things about the A600 and A1200 caddies.

There ARE three variants. The A600 variant fits them all. The A1200 variant (from the factory) only fits an A1200 without bending it. The Deformed variant can be also be either, or an A1200 which has been "straightened Deformed" so that it is a little bit too tall for the case, as well as the "Bent Deformed" for a better fit. Akira I think, and maybe others with same problem, do your caddies have perfectly straight backs? That might be the cause of the bad fit.

The caddy's all seem to have come off the production line with a uniform curve on the back. That makes the back much springier and gives you some tolerances. If somebody once thought "I'd better straighten that, its bent", you might have problems with a different case and keyboard, or a combination. Even if any original fits, you lose the slight tolerance you had before. Even with that, at the very least the paint will rub off, which is ALSO a design "feature". Caddies were designed very carefully but very quickly manufactured.

Anyway, the back is a UNIFORM curve, so you could replicate it on an original piece that has become deformed. Put the tolerances back, it's a neat little metal bending exercise but it's doable by some people, to some extent. Panel beaters and such.

Next posts will be

1) Github link, to photos to show original colour and preservation status, Original multiple raw .ply data from at least two different angles, hopefully more, for fidelity of all surfaces correctly.

2) After that, stl model of original postulated from multiple scans (never done that before. Could take me some time). Also the same original, but shortened to fit an A600. That should let people tinker once the data and hypothesis that the back was curved at the factory has been validated by other people as a fact, not just a hypothesis. Best orientation for ideal printing will also be determined and released. The orientation of printing is VERY important on an FDM or "Reprap" style printer - try it angled someways and gravity makes the process difficult if not impossible
They are the most common and cheapest type to print with.

3) One final printable model for A1200 SUITABLE FOR PLASTIC, Also one similar but say 1mm lower ( on the keyboard slope, give the tolerances back on a tight case), The printed parts don't need to be curved, they might be stronger if printed straight. . Other types can cope better, and print much faster. But FDM is most common.

The "Suitable for plastic" bit is important, some non critical changes happen to the model that don't affect the fit of everything, you end up with a much stronger finished part that is also a fit for the Amiga. The material is different, steel not plastic. The thinnest parts will still be 1mm thickness or so of ABS or PLA plastic and that's pretty strong. But it's better being a little thicker and stronger some places, obviously not where it slots on.

Might be an issue with some A1200 motherboards. I think Fillipino and maybe Brazil motherboards, they seem to be a little bit awkward sometimes. Caddies can be made to fit these boards, but, the holes in the board are soldered solid sometimes. It doesn't slide seem to slide in. You have to remove the blockage, somehow, soldering iron, solder sucker and maybe solder braid too. I guess that can happen on any A1200? Or was it also done at the factory some places, sometimes?

The caddy is very well preserved, I'm convinced the curved back was part of the production process, and maybe that's where people have gone wrong in some cases. The initial data should be able to confirm or deny that idea.

It's not a big job to have the jusr the bottom part as well, no slope for the keyboard. Depends if you want this case or another one, I suppose. Rackmount, whatever.

Last edited by Pat the Cat; 09 January 2017 at 16:24.
Pat the Cat is offline  
 
Page generated in 0.08418 seconds with 9 queries