Extreme refurbishing: Episode 2 - Atari 800 XL - Part two

... or biznaz continues :O




This is the second part of "Extreme refurbishing - episode 2" feat. Atari 800 XL
The first part can be found here

A curious case of plastic case repairs :>

Ok, back to work. A whole case was missing big chunks of plastic and was broken in several spots. In order to repair it, I needed a decent and cunning plan ... OK, maybe not cunning ;P

First, I needed to clean all parts from mud, grease and other nasty stuff that stuck to it.

Btw., I've lolled at this heavily. Atari still had an original foil on a badge :D
It kinda worked, it protected the badge ... only ... mwahahaha :D

Pics after initial cleaning.

Plastic welding method was applied to fix some minor cracks.

I also had to straighten one of mounting poles.
I've figured that heating plastic with a hot air gun will let me straighten it with a steel rod inserted in it.
Here is a small vid of how it went

and an outcome

Next, I needed to level both broken sides of the top part.
That is necessary to add stiffness and also serves as a base for repairs that followed.
In order to do do it, I had to prepare pieces of straight steel wires that were heated up with a gas torch and then left on a plastic and sank into it.

Unfortunately, a video is not showing what I wanted it to show but at least it shows a general idea (hopefuly).
Yeah, I know, I really gotta 3D print some sort of cam holder :)

With a case straightened, I was finally able to work on a mesh for the polyester resin. The plan was to make it out of copper wire and solder it all together. It looks rather ugly but it works.
Pics are worth 1000s words so have a look how it was done :)

The molds

At this point, I've figured (lol, no comment) that I didn't have another Atari 800 XL to take measurements so I was kinda stuck.
I've asked on Atari forums if someone has a spare 800XL case. I needed it to get all original measurements and to create a mold out of it. Fortunately, few of guys stepped in and helped me to deal with this problem.
Big thanks go to Perinoid (cheers buddy:)) who sent a complete case, along with an original keyboard which was also needed, but more on that later. With the original case, I could continue my work on 800 X(hel)L.

I needed to create a container for a mold. Obviously, I've used something that I already had on a shelf which was a piece of aluminum pictured below.

... but to make things more complicated there is a sub-story here ...

A year ago, I've bought a pack of molding silicone which I intended to use for this project and this is where Fail number one comes into play ... just wait for it

With all bits and pieces prepared, I've started assembling a molding setup.

It turned out that old silicone of that type will never cure. I don't really know why. I've even tried keeping it in 50-60 C for few days but that didn't help either.I've waited for two weeks and it didn't change its consistency at all.My best guess is that it simply was too old.
Sadly, I've ended up cleaning that whole mess and had to start from scratch.
I've ordered fresh molding silicone and just in case a molding polyurethane combo(which turned out to be a great choice).

Ok, back to work. TAKE 2 :D
I've started by doing another mistake ARRRGH. I didn't cover inside holes and obviously, (duh!) polyurethane came through all of 'em. I had to cut it a bit in order to pull out the original case, but in general, a top part came out pretty well. Polyurethane was a good choice because it forms a relatively strong bond and I was able to pull it out without any major damage to the mold itself.

A bit later I was ready to make a use of a freshly made mold YAY \o/ !

I've decided that I will use a polyester resin <-- that was a major mistake because it cures too quickly. More on that later.

When curing process ended I've removed "a ready" part from the mold.
At this point, I should have already noticed my mistake but somehow I've missed it :/

The problem was that a cast wasn't straight at all. This is because polyester resin tends to heat up and sorta shrinks during a curing process.
I should have used a resin with 24h+ curing time instead.

Anyway, back to the bottom part. It still needed some work.
I needed a slightly different approach here and I decided that I'll create a special box for this mold. Obviously, I did the very same mistake with resin choice as before.

This time, I chose that I will use molding silicone for a mold itself. It turned out later that it was a mistake and a necessity at the same time.

Necessity, because silicone is a bit less "stiff" than polyurethane so dismantling a mold (without doing a damage to it) is way easier, especially when there are a lot of holes where silicone goes through. However, the major problem was with air bubbles trapped in it. Here is where a vacuum chamber would help a lot.
You will notice what I mean on pics below.

Degassing a silicone alone would also yield way better results than above.
Actually, If I knew how much of work it will cost me to sort it all out (with a poor outcome btw.), I would have definitely stopped and built a vacuum chamber of some sort first.
Anyway, I moved on and started to build another mesh for a resin in a similar fashion as before - soldering a copper wire frame

Ammonium Chloride was used during the soldering process, simply to remove copper oxides and make the soldering process easier.

Some tests

No worries , horizontal wires were added later ;)
Here it is after demolding.

Putty and further < cough >fails< /cough > errr ... work

At this point, just prior to applying putty, I've finally noticed and pinpointed a problem with a polyester resin and its shrinkage that I've mentioned before.
Below, you can see how far from perfect that cast was.
But HEY, we learn as we go right ?!

Regardless poor effects, I had to move on and fix all mistakes. For that, I had to introduce some game changers - putties FTW! :)

Initial layers always look nasty ..

..but after a bit of work, it started to look a bit better.

At this stage I've also addressed a missing part near a joystick ports.

The top part was still missing a large chunk of plastic that had to be filled with a resin.
However, instead of a typical mold, I needed a polyurethane "footprint" of original cases texture, if I can call it that way. I've simply spilled a polyurethane mix onto an original part that had an original texture and then used it as a mold for further work <-- btw. that was another fail LOL

Here is why. The fill came out way too concave (a.k.a shrinkage problem) so the whole idea with applying texture failed for now.
Anyway, I had to fix all that mistakes again. I've started off with drilling holes as in the original case.

Initial file work and its outcome :)

Truth is that I really wanted to sort out rounded edges but I didn't have a proper file to do that job. I've ordered a few from China but to this day I didn't receive a package :(
I did my best with tools that I had and moved on to next stages.
A layer of putty was applied on, a mentioned before, concave part.

Also, I thought that I will be able to fix all "mold bubbles" problem by redoing a resin surface using an original mold. It kinda worked on old bubbles but without vacuum, it introduced new bubbles during the process :(
At least this sorted out uneven surface that was there due to resin problem.

At first, I was really happy with results because of nicely restored port descriptions but a while later I realized that it will be gone once I'll apply putty to sort out bubbles :(

Anyway, work had to be finished so I've corrected all ugly points as much as I could by injecting drops of resin into holes and then sanded it off. Finally, I've applied the first layer of undercoating.

Next, I've applied a first layer of (again) custom made spray paint that exactly matched an original color.
Bottom part.

I still wanted to get a nice original texture on the top part.
My cunning plan was to apply a texture by reusing a resin+polyurethane method.

As you can see, an outcome is quite decent but again, without a vacuum, it is all ruined by random bubbles. In the end, the idea was not so bad but it failed because I simply didn't have all necessary tools.

To wrap it all up, obviously, it could be done better. And YES I need new tools <-- work in progress ;)


Yup, that's how it looked :D
An early prototype of banana type ergonomic keyboard lol

As mentioned in the previous post, thanks to Perinoid I was able to supply missing keys and complete a keyboard. Here is how it looked after cleaning.
I've also trimmed and sanded off all keys that required it.

Yes, I know that I should have whitened all yellowed keys but I was simply too impatient lol. Hopefully, I'll sort it out later.
Soooooo with a clean frame and a keyboard, I started ...


Having all parts sorted out, I was finally able to assemble 800 XL.

Grand finale

Just to sum it all up. Yes, it could be done way better, I am aware of that. I should have used different resin. I also know that some sort of vacuum chamber is a very handy tool and would sort out quite a few problems. I've skipped describing several minor fixes because this post is already a long one and I wouldn't like anyone to TL;DR it ;P
Well, I've learned quite a few lessons this time so hopefully, it will result in a better outcome for upcoming projects. Oh, Did I mention that there are quite a few projects waiting in the queue? :D

So this is it, this is how an Atari looks now.


LAMERS - poor but miserable :D

Jesionen - for donating a CPU to this project
Perinoid - for sending me a A800XL case and keyboard
String - for wanting to send a A800XL case :>
Kaz - for bringing morale up :)
SOS - for giving this Atari to TDC
TDC - for successful trolling :>

... and obviously a mandatory video dedicated to TDC :D


Chaos is your redemption .... better run .... better hide

I come from Internetz :>