There were books sold with software printed in bar code form - Paperbyte books.

You can probably print pages, tape them together and scan them into your machine like a player piano.

So many cool ideas!

For some reason the LP style disc reader just seems to stand out.
I bet something like a 10" diameter white arborite disc could be pulse burned with a 1 watt blue laser and probably manage 64K easily.
A binary stream of dots about 32 DPI. Reading and writing becomes easy, only require one laser for write and a decent optical sensor for read.
A single lead screw working like the mechanics in a typical floppy would do the job.

I have a 250mw red laser here, maybe I will do some testing on various surfaces over the winter and see what happens.
It's won't be as fast as a 1 watt blue, but I can certainly burn surfaces with it, and I modded it for digital modulation.

Anyone know of a decent spiral calculator that will to the math based on thickness of a line and space between a line?
This one is the opposite...

https://rechneronline.de/pi/spiral.php

Assuming 64K using .03 pitch dots, I would need a spiral length of (8*65536*.03) = 15,728 inches.
I imagine this is doable (~1300 feet) on a disc of under 12" in diameter.

The reason I like the laser burnt disc or paper idea is because I want to see the data. It would be much more retro cool to show someone the media, and say.... "see those dots? They are ones and zeros" right before loading it up to see the magic happen on screen.

Brad

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Sounds like you're at about the density of the LP record, which is approx 200 lines per inch (variable) - so first question is whether a lead screw approach can track such a dense spiral (and how it can be aligned in phase onto the track.)

The bad thing about floppy controllers is, that they would refuse to write data on a non_formatted floppy disc.
You can write the paper only once with that laser, so writing the formatting before writing the data won't work.

Hmm... there are ROM listings for the 1541.
Technically, the 1541 is "just another 6502 single board computer".

Would be interesting to know, if a modified 1541 PCB could be re_purposed to handle your "proprietary" storage medium... somehow.

Depends on the floppy controller. On an Amiga computer the computer always just read and wrote complete tracks. Mfm en and decoding took place in software.

Thinking about it, I reckon it will be mechanically much easier to use concentric tracks, like a floppy, and not a continuous spiral like an LP or CD. No need to follow the track: it's open loop. But with the disadvantage that the datastream is not continuous, and when arriving on a new track it might take half a rotation on average to find the start of the data. Some signalling needed in addition to raw data bits.

The one advantage I will have in my mechanical design is that the same hardware will read and write.
On a commercial product like a floppy disc, the manufacturer will have to deal with all kinds of tolerance variations.
If I use the mechanics from this old project of mine, I should be able to get into the required accuracy...

https://www.youtube.com/watch?v=-CVN1L6jhDE

I still have these parts sitting in my shop.
Yes, that video is LONG, the action starts at 2:50.

If I can find an inexpensive and readily available thin material to make a robust 12" platter, then perhaps using a laser to burn small holes might work well.
For reading, the laser could shoot a low power read beam back and a photo sensor could clock in bits.
Arborite would be perfect, but not sure if would be easy to pop a .05" hole through with a laser.
Perhaps a small drill bit instead. At 12" diameter, it should still be possible to store 64K with that size of punch hole.

This will be a fun project to work out over the winter along with the final circuitry for the main V-74 system.


Brad

Nice project. A tenth of a millimeter mechanical accuracy sounds about the right degree. Promising!

On second thought... in case you only want to read and not to write:

Some of the PCB manufacturers can make a PCB of almost any shape you want,
and I think this includes "round with a hole in the middle".

Now, if one would write a little C program that reads a binary for generating
a script which draws the tracks of dots on the top copper layer of the PCB
in Eagle\KiCad\etc., he could send the resulting *.brd file to a PCB manufacturer.

While "read only" isn't exactly what's needed,
I think a "PCB disc" might be a nice tool for testing/adjusting the hardware\software of the drive.

Edit:
Of course, writing a little C program that generates a *.bmp from a binary
for printing the dots of tracks on paper would be less expensive for a start...

...and some of the PCB layout software out there might be able to import *.bmp,
probably mainly for only importing logos and such, but it might be worth a try.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I'm not familiar with the RS-274X extended Gerber format.
But there are Gerber viewers out there, and in some of them it might be possible to edit the Gerber file.
Just make sure that your aperture list matches the aperture list of the PCB manufacturer.

The hole in the middle of the PCB would require an Excellon drill file.

Solder mask isn't supposed to be much of a problem.
For instance, in Eagle 6.4 everything drawn on layer 29 (tStop) later would be free of the top solder mask on the PCB.
Just drawing circles (at a reasonable trace width) for the tracks of dots on layer tStop would give you plain copper for the dots.
//I don't know how to do this in RS-274X Gerber.

The PCB Disc is a great idea, I would like to see this done someday!
A laser hitting the disc and a photodiode catching the beam would probably work at the CNC tolerances we are discussing.

In my implementation though, I have to maintain these rules...

- Must be able to easily and inexpensively make new program recordings at home.
- Must be built from components I have in this room.

While I do have LOT of stuff, I don't have a PCB router, well only A small one!

I also modelled a 12" disc with a track in SolidWorks, and it only held 11K, so the disc is simply not dense enough.

So far the paper roll is still the top solution.
I am swaying towards the burning laser + roll instead of laser printer + end to end taped sheets though.

On a more current note, the one design issue I will be up against this year is the monitor bezels.
I don't intend to use the 1702 monitor bezels because I want more of the PET look, and I want to make them myself.

This is exactly the look I intend to copy....



This is most likely vacuum formed plastic.
My first thought is to carve the bezels from wood by hand and then cover them with epoxy and attempt to get the rough look later.
I don't care how much work it is, but I do want to get close.

I would consider farming this out as well if I could find a plastic prototyping service that didn't cost a fortune for 2 of these.
Not sure what's out there these days, but a few years ago I looked at 3D printing something, and it was crazy expensive.

Any suggestions here would be great. I want these completed before spring.

Cheers!
Radical Brad

For a textured bezel finish, it reminds me of the 50's... ARRL manuals. Homebuilt radio gear was finished in either Black Krinkle finish or Gray Hammertone finish. You can still get the black textured paints... they generally refer to them as wrinkle or textured... but I still prefer krinkle

Project is looking great!

_________________
Regards, KM
https://github.com/floobydust

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?