• Please review our updated Terms and Rules here

C64 IN PET S' CASE

Unfortunately, you have to make it.

You can start off with a bit of prototyping plugboard where you just plug the components in and then plug wires in to make the interconnections.

You can also use prototyping board and solder IC sockets and the components onto it and then insulated wires to make the interconnections (for something more permanent).

Dave
 
I found this picture on web.... Cbm 4064 video output:
 

Attachments

  • Immagine 2023-06-29 125709.png
    Immagine 2023-06-29 125709.png
    1.2 MB · Views: 26
unfortunately i can't find the bottom part or the gerber file :(
I am desperate!!
 
There are some better photographs here of the inside layout of the CBM 4064: https://www.vclab.de/projekte/restaurations/commodore4064-restauration/

You can see the cable (red, black and white) from the C64 to the monitor signal bodging PCB and the other cable to the PET 8032 monitor.

I suspect Ruud's monitor pinout is not 100% correct though - but as the wires are pushed onto the connector it shouldn't matter.

You just have to build Ruud's little circuit now...

Dave
 
I found this picture on web.... Cbm 4064 video output:
This looks really easy. In fact if you had a photo of the rear of this pcb you could reverse engineer it in a heartbeat. Even without that though, Ruud's schematic will match it, so you can simply place the components as shown on this pcb photo and make the tracks conform to the schematic.

On the notion of the LM1881 (post #20), these seem appealing as a composite signal sync pulse separator, they are not bad, but not perfect either.

Years ago I worked on a project to find the perfect sync separator for composite video. One that would extract syncs and completely ignore the picture content. It seems easy on the face of it, but it is not because the picture content alters the overall signal amplitude and this causes phase shifts in the extracted sync.The highest chance of instability occurs going from a frame of black to peak white and back again.

A superior IC to the LM1881 , with the same basic signal outputs is the EL4581, I used one in this project, see page 9:


All sorts of tricks were tried to make a perfect sync separator, with signal clamping, comparators with slice levels through the sync pulses etc. It was hard to find the perfect performer.

Then out of the blue a three transistor circuit turned up which out did them all. JVC in Japan designed it for the sync separator in the camera control unit of their KY2000 camera. It out-did all others, including the LM1881 IC.
 
Last edited:
>>> Is possibile use Pet power supply for c64 board?

Yes and no...

It looks like the same transformer is used (as in the PET) - but the C64 requires +5V DC and 9V AC.

The 9V AC supply can be supplied from the PET transformer - but the +5V DC has to be regulated and smoothed off-board and fed to the C64. If you look at the link I provided you with, you can see the little PCB that has been developed for the power supply.

>>> I am desperate!!

Back to desperate mode already?

If Ruud hasn't produced a PCB - then the only option is to either design one - or prototype the circuit up on (say) veroboard.

As this is a scarce PET model, then you will have to hunt around the internet for photographs. It seems there are component side photographs - but I have not come across a track side photograph so far.

The PCB looks to be single sided (hence the wire links).

Dave
 
I have put together a Bill of Materials (Bom) for this video card.

If anyone spots any errors let me know and I will update it. I have yet to find the part numbers of the diodes.

Dave
 

Attachments

  • CBM 4064 BOM.txt
    615 bytes · Views: 8
I read that these roms will probably be needed, is that correct?
 

Attachments

  • Screenshot 2023-06-29 alle 21.43.17.png
    Screenshot 2023-06-29 alle 21.43.17.png
    35.6 KB · Views: 9
Oh yes i found new pictures of the video board!
 

Attachments

  • DSCN4620.jpeg
    DSCN4620.jpeg
    2.9 MB · Views: 22
  • DSCN4621.jpeg
    DSCN4621.jpeg
    3.4 MB · Views: 21
  • DSCN4622.jpeg
    DSCN4622.jpeg
    3.5 MB · Views: 23
Major ⭐!

You see, no need for the desperation - you just have to put in some hard 'slog' searching the internet...

Photograph DSCN4622.jpeg looks to be the key...

All you have to do now is to persuade someone to design a PCB from the photograph.

There are online PCB design tools that can be used. The downside is that you generally have to purchase the PCB through them. Why don't you have a go at creating the PCB yourself? Learn a new skill...

We can check it for you before you purchase a PCB. I would be interested in one - as I know a few other people would be!

The three IC packages have a specific layout - but I would suggest purchasing the resistors, and capacitors before hand and making sure that the component spacing matches the components you have purchased. However, I would also try to purchase 'similar' components to what was used. The three brown ceramic disk capacitors are now much smaller. I would suggest replacing the two electrolytic capacitors with tantalum beads. We then need to work out what the best timing capacitors would be.

Dave
 
>>> Yes I would love to but I don't know how to proceed...

Pop over to: https://easyeda.com/. I wonder if there is an Italian language option for you?

You enter the schematic (follow Ruud's original).

Assign footprints to each component.

Move and orientate the footprints to match the solder-side photograph you have found.

Interconnect the terminals of the footprints (components) to match the copper tracks on the photograph.

OK, there is a learning curve to go up - but that is why we are here. Have a go - it is free - so it won't cost you anything financially.

The PCB is single-sided (from what I can see) so there is no hidden 'magic' to work out.

Dave
 
>>> Yes I would love to but I don't know how to proceed...

Pop over to: https://easyeda.com/. I wonder if there is an Italian language option for you?

You enter the schematic (follow Ruud's original).

Assign footprints to each component.

Move and orientate the footprints to match the solder-side photograph you have found.

Interconnect the terminals of the footprints (components) to match the copper tracks on the photograph.

OK, there is a learning curve to go up - but that is why we are here. Have a go - it is free - so it won't cost you anything financially.

The PCB is single-sided (from what I can see) so there is no hidden 'magic' to work out.

Dave
Great!
for example like this?
 

Attachments

  • Screenshot 2023-06-30 alle 14.31.00.png
    Screenshot 2023-06-30 alle 14.31.00.png
    187.6 KB · Views: 8
Yep, one component on the schematic down, more to go...

This is the schematic that you are now doing of course - not the PCB layout. That comes later. The software tool will compare your schematic with the PCB layout and ensure there are no inconsistent connections etc. These are called design rule checks. They also check that you have not 'goofed up' with your schematic (e.g. left inputs unconnected).

You have placed JP1 at the correct location. I think that the 'wiring connectors' are coming out of the left-hand side of the connector whereas Ruud's are out of the right-hand side. You need to mirror or flip the JP1 connector around so that the wiring is oriented the same.

I see it is in Italian as well :)!

Dave
 
So first of all I have to place the components and the integrated following only the ruud scheme? Just the top of the board basically?
 
Correct, just copy Ruud's schematic to start with. I think Ruud uses KiCAD rather than EasyEDA - but there should be a one-to-one relationship.

You are NOT doing the top of the board. You are entering the schematic. You will layout the PCB after you have entered the schematic.

Dave
 
Back
Top