Apple Lisa 3d Case Project

[phunguss]

Some guy did just recently 3d print a full scale Lisa

Here is that Pi case at 325% and thinned out. It will probably be easier to modify this model than recreate from the 3D scanned version.

 

[Trash80toG4]

Agreed. Some parts could be seamed with minimal visual impact, but the front panel is still an issue. There is room to shrink it further to fit inside a 300x300x400mm build volume of an old CR-10 or 350x350x350 K2 build volume. Further shrinkage may make the 9.7 screen appear disproportionate, but an original 12" to 9.7" conversion should be 80.83 percent.

Percentage may be off:
- Specs for LCD and CRT are not the same. Tube diagonal is not equivalent to LCD dagonal.
- Need to consider how much larger a 9.7" screen would be than an 80 percent opening for a 12"diagonal CRT
- Display area of an LCD is the tuge diagonal vs. the display area of a 12" CRT.
- The front bezel covers a portion of the CRT.
- The diagonal of the "12inch" bezel opening is only 11 5/8 inch.
- The "pixel field" diagonal is less than that.
- "Squared up" LCD opening dropped back from surface of the front bezel a tad more than 1/4" yields a diagonal of 11 7/8"
- So your baseline for scaling would be something like 11.875" -> 9.7" yielding a scaling factor of at least 81.68 percent
- Still wouldn't look quite right with no Start/Stop/Front Porch/Back Porch blanking areas surrounding the Lisa image
- Assuming my caffeine deprived brain is up to snuff for this.

All measurements are from my Lisa2 bezel with pencil scribed setbacks. A 12" LCD won't work in it for same reasons.

 

[phunguss]

Percentage may be off:
- Specs for LCD and CRT are not the same. Tube diagonal is not equivalent to LCD dagonal.
- Need to consider how much larger a 9.7" screen would be than an 80 percent opening for a 12"diagonal CRT
- Display area of an LCD is the tuge diagonal vs. the display area of a 12" CRT.
- The front bezel covers a portion of the CRT.
- The diagonal of the "12inch" bezel opening is only 11 5/8 inch.
- The "pixel field" diagonal is less than that.
- "Squared up" LCD opening dropped back from surface of the front bezel a tad more than 1/4" yields a diagonal of 11 7/8"
- So your baseline for scaling would be something like 11.875" -> 9.7" yielding a scaling factor of at least 81.68 percent
- Still wouldn't look quite right with no Start/Stop/Front Porch/Back Porch blanking areas surrounding the Lisa image
- Assuming my caffeine deprived brain is up to snuff for this.

All measurements are from my Lisa2 bezel with pencil scribed setbacks. A 12" LCD won't work in it for same reasons.

Thanks for those measurements. I think we still need to wait for word if the LisaFPGA can even output to 1024x768 over HDMI, as the documentation seems to show that it outputs full 1080p. Not sure if an iPad control board scales higher HDMI signals down to 1024x768.

 

[Trash80toG4]

Angles for setback of display are different. Sides are steep, Top is more shallow and bottom is very shallow.

Are you taking that into consideration? Matching the multi-angled framing of the LCD would enhance the look, being more natural.

Note scribed lines in pencil. Strange bulge on bottom curve screws up adaptation of stock bezel to LCD. Would probably need an 11" 3:4 panel.

 

[wottle]

Thanks for those measurements. I think we still need to wait for word if the LisaFPGA can even output to 1024x768 over HDMI, as the documentation seems to show that it outputs full 1080p. Not sure if an iPad control board scales higher HDMI signals down to 1024x768.

I believe it does, but my LisaFPGA arrives Friday so I'll confirm and we can proceed from there with the result. But I forgot the conversion of CRT size to viewable. It definitely seems like even with a smaller 9.7" LCD, it will be difficult to get something printable on a 350mm x 350mm bed.

Possibly if we use the Pie model, and print the pieces vertically at a diagonal on the bed. Anyone print a really large thin piece standing on edge on a large format printer? @jcm-1, I believe you have some large bed 3d printers. How do you think it would handle something like this?

 

[Trash80toG4]

I believe it does, but my LisaFPGA arrives Friday so I'll confirm and we can proceed from there with the result. But I forgot the conversion of CRT size to viewable. It definitely seems like even with a smaller 9.7" LCD, it will be difficult to get something printable on a 350mm x 350mm bed.

Glad to have pointed out that discrepancy, it drove me to distraction playing with my "Little Bit of Lisa" before the LCD/Controller revolution induced by the Raspberry Pi movement. The solution I came up with was the 12" RGB PizzaTopper. Fits like a glove, but not at all practical unless a 640x480 hack can be pulled off. The targeted 6400/6500 boards can drive the smaller resolution, but why bother?

One single piece print done diagonally makes no sense to me whatsoever. All that filament gone to wa$te in one bad shot? The Lisa case is nicely divided into sections, unlike the unitary Macintosh bucket. I say leverage that sectional flexibility to reduce the price impact of a failed print?

 

[phunguss]

Well, if an iPad 1/2/3 at 9.7” is too large for a smaller print, we could try an iPad Mini 1/2/3/5 at 7.9”, but working controller boards are harder to find. I think this may be the only one. eBay for AbuseMark.

 

[Trash80toG4]

The front bezel (only Little Bit of Lisa in my collection) is your print limit factor. If you angle just that one part to max size in you printer, might it fit that cubic closer to life size?

edit: I can knock back the return at the bottom of my bezel a hair over 3/4" at the mold break line to gain a few extra degrees of slant for printing.

 

[wottle]

The front bezel (only Little Bit of Lisa in my collection) is your print limit factor. If you angle just that one part to max size in you printer, might it fit that cubic closer to life size?

edit: I can knock back the return at the bottom of my bezel a hair over 3/4" at the mold break line to gain a few extra degrees of slant for printing.

I must be misunderstanding what you are saying. The front piece is ~18" wide The full width of the Lisa assembled is 18.75". The difference between a part that can print a 457mm wide print and one that can print one that is 476mm wide is not going to rule out any 3d printers. Even scaled down to ~80% (which is not going to accommodate a 9.7" LCD, that difference is 365mm vs 380. Again, at that point you'd need a 400mmx400mm print bed printer and you'd be better off printing together.

There are advantages to printing exposed surfaces face down on a bed (can generally get better surface finish), but ultimately, we're likely going to need to sacrifice to make this producible by more than one person with a very expensive massive bed 3d printer. It might be that we find that multiple parts and seams are going to be needed. I still would rather waste a good bit of filament on supports to avoid the time, materials, repainting, and look of seamed prints.

I may also just not be understanding what you are saying.

As for the modeling, I really wish we could get access to the models that were created for this project / book:

There would need to be objious adjustments due to the lack of the internal metal frames, but having it assemble much like the original would be cool. And it would certainly not be worth the effort, but adding a switch that cuts power when the front faceplate was removed would be a fun easter eggg.

 

[Trash80toG4]

if printer cubic is a 350x350x400 it looks like I could print my bezel at full size?

"The hypotenuse is the side opposite the 90° angle. Using the Pythagorean theorem, √(350²+400²) ≈ 531.4 mm."

Yep! The bezel is a bit over 450mm wide. NoPro!

"If the right triangle’s legs are both 350 mm (an isosceles right triangle), the hypotenuse is found by the Pythagorean theorem: c=3502+3502=3502c=3502+3502=3502. That equals about 495 mm."

Dunno, why can't it be printed for 9.7" screen?

 

[wottle]

if printer cubic is a 350x350x400 it looks like I could print my bezel at full size?

"The hypotenuse is the side opposite the 90° angle. Using the Pythagorean theorem, √(350²+400²) ≈ 531.4 mm."

Yep! The bezel is a bit over 450mm wide. NoPro!

If the right triangle’s legs are both 350 mm (an isosceles right triangle), the hypotenuse is found by the Pythagorean theorem: c=3502+3502=3502c=3502+3502=3502. That equals about 495 mm.

Dunno, why can't it be printed for 9.7" screen?

I thought you were saying you don’t want to post it standing on end. And you wanted to print it flat on the surface. Are you saying we should print it like this with the shorter edge on the print bed angled up to use the 400mm of height diagonally to take advantage of the 530mm or diagonal space going from the bottom right of the bed to the top leftmost on the x axis:


that would have even more supports wasted then printing it on edge. Printing the front surface at an angle also creates pretty bad layer lines. Again, I think I’m missing the point you’re making but maybe @phungus is following and he may be doing more of the modeling and can take that into account.

Let’s wait and see when my board arrives tomorrow. If the 9.7” display doesn’t work, we’ll need to decide on another LCD and adjust all the measurement.

 

[Trash80toG4]

That's exactly what I was suggesting to get the largest possible parts printed. Front bezel and top are the major problems. I was looking at this printing method as a huge waste of filament with a full case print gone bad:

This one is great if you can get the size you need:

Filament for the extra supports for front/top printing as you illustrated pales in comparison to a lost print of the whole case, no?

OK, wait and see time!

 

[joevt]

This one is great if you can get the size you need:

View attachment 29557

If the base is 350x350 then that's 495 mm diagonal.

Yep! The bezel is a bit over 450mm wide. NoPro!

Of course, a 495 mm long rectangle could only have a thickness of 0 in that case. So I googled this:
"thickest rectangle 450 mm in length that can fit in a 350x350 square"
it says 𝟒𝟒.𝟗𝟕 𝐦
I did not verify the math.