Making a CoCo 3 'ish Case - Largely in Grasshopper

[Brenda]

So, a bit of a retro-computing person. I have this c1987 Tandy Color Computer 3 in the other room. It’s so well taken care of that it even smells new, but where is the next one going to come from? Well there are emulators, but I thought it would be fun to brush up on my Grasshopper by making a case.

The problem is: with emulators, perhaps someday a new board, the case a moving target, so I am making much of the case in Grasshopper, a single-point drawing, so all the figures of merit, as they say are adjustable.
(The wood texture is not mine. It’s nice.)

This is the start I have so, far:

The case is drawn from a loft outline, capped, and made into a solid.

The recess were drawn from planes that were based from planar surfaces. I created a tool, and differenced the tool from the housing.

The a copy of the section was interferenced with the housing, the only time I have ever used interference to create anything. Then the chopped section was rescaled and submerged into the housing, before a union was done.

Then after finding all the edges, I created a list for filleting the edges of the recesses.

I 3D printed some radius gauges to measure and dial-in the case. Oddly, this is the only thing I downloaded that I didn’t remake. They just worked well.

Parametric (Customizable) Radius Gauge by darealmoo

There is a little rendering defect caused by a too coarse mesh and differing radii. This should be fixed now.

[Brenda]

Generally, Grasshopper is fun. It’s a great part of Rhino.

This is my component nest for the top shown.

[Brenda]

Well, it’s getting there. With the help of Pufferfish’s Extrude-Tapered, which I feel should be part of Grasshopper, most of the top is slotted.

Though, because Rhino’s newer? Offset works better/differently than Grasshopper’s, the subtractions in Rhino, using Baked tools. Otherwise this would be a single-point Grasshopper sketch.

(Edit rendering)

Like the original, slots within slots.

An original CoCo 3, for which the blocked off the power supply vents to keep kids from sticking things into the mains:

I feel that the original like the 1000 HX, were nice bits of design. (Mournful cat not included. Warmth-gravitating cat offer void were prohibited. Cat application may slow Brownian motion of molecules.)

[Brenda]

The case bottom has been roughed-out. I am working on a dual-SD holder for the case. It needs to be wider. The logo is just for fun, actually, the whole thing is.

So, I can make the case solid shell in Grasshopper, and then Bake a set of parametric subtraction tools to do the other operations after ofsetting, BUT, unfortunately, I am aware of no practical way to shell/offset the case in Grasshopper. The offset command in Rhino still works better than anything else. Also, in Rhino, I can delete the bottom before I shell it, and it doesn’t need to be retrimmed before offsetting, but in Grasshopper, I can only deconstruct to grab the surfaces.

Even in Rhino the offset solid’s edges are not flat, and need to be trimmed.

So, for things to work perfectly, a offset tool would have to also use a plane, in order to make an truncated offset.

Still, the devil is in the details. The seam is tricky. There’s an inset, which is easy, but there needs to be a flange whereas part of the (say) top fits overlaps the bottom–all with draft angles, like the rest.

I don’t want to keep respining cases. So, at some point, the case will have a removable motherboard mounting as well as interchangeable back panels.

I am tempted to do the bottom of the case with hexagon vents. Purists are already offended.

[Brenda]

The feet, and vent-o-matic are both done for the bottom. The vents are adjustable for size and number of groups, as are the top, and nearly everything else–except for the badge size.

If it weren’t for having to offset it in Rhino, the whole thing could be done in Grasshopper, single point. So far, only the SD and one point are drawn in Rhino. Grasshopper makes the Boolean tools to do the subtractions, though.

One thing I like about doing things in Grasshopper: You can make a change before filleting. I still get stressed every time I have to fillet something in Rhino, and the scary thing is: the problem might have occurred downstream, like my pet peeve: unnecessary surface splits during operations. If you have a diagonal split on a surface you want to fillet, it’s already over before you begin. I liked the thread whereas people challenge the programmers make sure that difficult objects are filleted. Though, objects have to have sound construction. The fillet tool will find errors, you hadn’t noticed.

Anyway, I am having fun with Grasshopper. I really like this bench texture I found.

I really need to work on the seam. It’s tricky. The first part is easy, making a inset extruded polysurface/brep to separate the case halves. The tricky part: there needs to be an extruded ring that aligns the two case halves. It must have a draft angle that works on both sides. But there needs to be a tolerance gap, too.

I forgot the label recess, which needs to be on the bottom, though there are none in on the original.

These are the compents, so far. The ungrouped stuff in the middle is the case seam, which isn’t done yet.

One thing I have learned: reference objects/surface as far back/up as you can. In other words always reference anything before adding anything else to it. Like if I started the case bottom after the vents were added, then I would have a lot more surfaces to look thought.

[Brenda]

I necromancied the computer case project. At this point, I added another feature slot area on the front of the case. With draft angles, this was complicated. If a mould is made, it would need plugs that come in from the front.

Because one of the case goals was to be reconfigurable, I came out a rear panel fill detail, which allows interchangeable sections. I’ve tried to use similar design language.

The Raspberry Pi 4, not drawn by me, is shown for scale.

Someone wanted fans, so… With the cover removed, you may notice the fan cut-outs for 3 of 80mm fans that take in in from the bottom. There are cover plates for use by people who want to passively cool the case. These can be vented, too, but in the original case, the slots in the front take in convection air to cool the whole PC board.

I have to open up my computer to measure the original keyboard, and draw that out. Then I need to make some attachment and supports for it.

[Ironically, since I bought my CoCo 3, the ebay price has gone up to an inexplicable going price of $350. My 1987 CoCo 3 has almost no wear on it, and even smells like new plastic. The one I was forced to leave behind in 1989 had more wear on it.]

At this point, I haven’t merged all the parts, but it’s close. I need revisit the screw holes, as well as place holes for the fan.

I’ve tried to make the case accommodate various sizes of boards and adapters.

The case lid will also allow 3 of 60mm fans. The SD/Feature areas are also meant to be interchangeable. I must have dropped the rear boss supports on the wrong layer, but you get the idea.

The most complicated thing about this whole project: drawing the perimeter seam, which has a 1mm inset (like a woodworking reveal but an innie). In wanted the seam to be closed which requires a bit of tension, from the .25mm gap between the screw bosses. Yet, the X and Y needed a bit of clearance, so the parts fit. The draft angles on all of this give it just the bit of extra complexity.

With draft angles, the rear fill plates became a thing, too.

It’s a bit smaller than one might think. I may make a extended wide version, too.

Generally, I like the industrial design of this old case. The Tandy 1000 (below) was a great design, too. It’s just cheap plastic, but they managed to humiliate IBM with it.

Here’s a PS2 for comparison:

[hjbedoya (Hector Bedoya)]

hi, do you still have the files?

[Brenda]

Yes, I do. Presently, I am working on a vehicle project.
I know that it’s annoying when I switch projects–but hey–I have to put up with it too.
Ironically, it’s almost done.

[Brenda]

I took a look at the case project, today.

I know where I got tripped up: to continue, I need to measure and draw the keyboard, which means opening my (near) mint-condition 1987 CoCo 3. Ironically, the computer still smells new, and looks newer now–than the one I had to leave behind in 1998.

I added the last few non-keyboard related ribs in the case.