Denial

I need a bit of help here.

Since this thread started, I was worried only the 6561 would be explored. Then Kakemoms came along and has posted electron microscope images of the 6560 as well as optical microscope images.

Now, if you look at the die shots catalog on Visual6502.org there are a few versions of the VIC-II (6569 and 8565) are available there but they too are PAL only. See: http://visual6502.org/wiki/index.php?ti ... collection

It so happens I have a ceramic version of the 6567R5 I can sacrifice but before I tried opening it and potentially made a giant mess of this small thing, I thought I'd try opening up an old Pentium 75 with my Dremel. All I was able to do is destroy 2 of my carbide disks.

What the heck? Are these things made up of Star Trek's fictitious 'duranium' compound?

Sidenote: What I tried to cut into is the metal cover on the underside of the chip. The shiny line around the black cover...



Can somebody tell me what is required to make a clean cut through which won't damage the die ? And preferably not destroy my cutting tool...

Somebody obviously had more success than me...
http://mixeurpc.free.fr/SITE_x86-guide/ ... %20P54.jpg

Hot air gun.
The cover is soldered in place.

Brad

eslapion wrote: Can somebody tell me what is required to make a clean cut through which won't damage the die ? And preferably not destroy my cutting tool...

Radical Brad wrote:Hot air gun.
The cover is soldered in place.

Brad

eslapion wrote: Can somebody tell me what is required to make a clean cut through which won't damage the die ? And preferably not destroy my cutting tool...

I do have a very good hot air gun which will liquify tin-lead solder.

WIll this trick also work with this:


Also, this begs the question: Since the silverish line is indeed solder, why did it destroy my disks ? Isn't that material supposed to be soft ?

A diamond impregnated Dremel cutting blade would work on the ceramic, but as said hot air is the way to go. In the past, I heated the cap on a chip with a propane torch briefly (don't burn it) and it came right off.

eslapion wrote:

Radical Brad wrote:Hot air gun.
The cover is soldered in place.

Brad

eslapion wrote: Can somebody tell me what is required to make a clean cut through which won't damage the die ? And preferably not destroy my cutting tool...

I do have a very good hot air gun which will liquify tin-lead solder.

Its not tin-lead, that would not work on ceramic packages. Most often its an Indium alloy (In-Sn or In-Ag). Indium tend to wet most things (unlike Tin), so you probably end up covering your cutting disk with it.

Hot air should work.

Kakemoms wrote:

eslapion wrote:

Radical Brad wrote:Hot air gun.
The cover is soldered in place.

Brad

I do have a very good hot air gun which will liquify tin-lead solder.

Its not tin-lead, that would not work on ceramic packages. Most often its an Indium alloy (In-Sn or In-Ag). Indium tend to wet most things (unlike Tin), so you probably end up covering your cutting disk with it.

Hot air should work.

Well, it didn't 'wet' my cutting disks, it reduced them to splinters in the blink of an eye. Good thing I was wearing eye protection!

When looking at the seam material on the edge of the metal cover on my 6567R5, it doesn't look like solder as it did with the Pentium 75. It looks like a mixture of silver and gold so I am reluctant to put a torch on it. Also, I am worried about a potential thermal shock; having your face sprayed with bits of super hot ceramic and metal isn't a very appealing prospect.

eslapion wrote: Well, it didn't 'wet' my cutting disks, it reduced them to splinters in the blink of an eye. Good thing I was wearing eye protection!

When looking at the seam material on the edge of the metal cover on my 6567R5, it doesn't look like solder as it did with the Pentium 75. It looks like a mixture of silver and gold so I am reluctant to put a torch on it. Also, I am worried about a potential thermal shock; having your face sprayed with bits of super hot ceramic and metal isn't a very appealing prospect.

Well, sounds more like you should steady your Dremel before you start.

A better option is to use a CNC mill: https://openbuilds.com/builds/openbuilds-minimill.5087/

Kakemoms wrote:Well, sounds more like you should steady your Dremel before you start.

A better option is to use a CNC mill: https://openbuilds.com/builds/openbuilds-minimill.5087/

I did steady my Dremel. I cut lots of other stuff with it but I guess if you hit either the ceramic or the black cover just a little bit then your disc is a goner. Using the Dremel would require hitting the seam material perfectly. Too much for me at this time.

A CNC is something I have considered over the last few months for a variety of use, including getting PCBs faster.

Abra is nearby and they offer this little beauty: https://abra-electronics.com/3d-printer ... e-kit.html

... I'm thinking about it...

Perhaps that nice old 6567 would rather not die, and instead become part of my large collection of ancient ceramic chips?
... just checkin!

Brad

Radical Brad wrote:Perhaps that nice old 6567 would rather not die, and instead become part of my large collection of ancient ceramic chips?
... just checkin!

Brad

I would really like to reserve this specific unit to become a documented IC. I want to make sure I 'open it up' correctly.

lance.ewing wrote:Logisim is nice for working things out for small sections, but to build up the whole chip in one schematic, I think we'll need something else. Suggestions welcome for suitable tools.

Yesterday I discovered that the CS3410 course at Cornell University has an "evolved" version of Logisim called Logisim Evolution. Their download of this includes a few custom components built for the course, one of which is a 128x128 LCD component. You give it an X/Y position and an RGB colour and it plots that pixel. That got me thinking, and by the end of the day I'd created a modification of this component (Logisim and the library components are written in Java, so right up my alley) that is instead a 284 by 312 LCD screen. My idea is that the X/Y position would be tracked internally by this component and I'd just feed it a steady stream of RGB pixels on each dot clock cycle. With this component now available, I might have a go at connecting all the various logisim simulations I created together and see if I can get them working in unison to produce something visual in this LCD component.

Just had a read up on logism. Looks like a nice tool.

That sounds like an exciting project - how were you planning to drive the inputs to the (WIP) simulated 6560?

beamrider wrote:how were you planning to drive the inputs to the (WIP) simulated 6560?

If you mean how am I going to provide the screen RAM, character ROM, and colour RAM, then what I'm planning to do is make use of the ROM and RAM component simulations that come with logisim. Logisim should allow a file to be loaded into a ROM component, so I should be able to load the character ROM into one. It may be that I use the ROM component for character ROM, screen memory and colour memory since nothing is actually going to be changing the memory, except for me manually doing it. I'd be preloading them with data, e.g. colour memory might be all 0 (which is the default for that component anyway) and screen memory might have 0-255 repeated twice. I'd then have the address and data pins connected up to those components in the way that you'd normally expect.

Something I hadn't really played around with before is that you can create sub-circuits that by default appear as a box with a name and its inputs and outputs (you can customise the appearances). So I'm planning to have a Horizontal Counter, Vertical Counter, Clock Generator, etc. as sub-circuits that are then connected together in a higher level circuit. In fact what I'll probably have is the top level circuit with a 6561 component (as a box with inputs and outputs, i.e. the pins) and the ROM (and maybe RAM) components mentioned above would also exist at that highest level. Then drilling down into the 6561 would show what is basically a block diagram of the 6561 functionally, where each box represents a sub-circuit, and then you can drill down into each of those.

Now that I think about it, at the highest level I might set up an ultra minimum VIC system, something I was playing around with last year on a breadboard with the real chip. It didn't have a CPU but instead used DIP switches to set the control register values. I had it working so that you set the address on one set of DIP switches, the data on another set, then push a button and it gets written into the 6561. I also had the current value stored in that control register being read and shown on some LEDs to confirm it had been written. Worked really well. The breadboard circuit made use of 74LS245 chips to keep those reads and writes in the correct clock phase, i.e. phase 2, so that it didn't clash with the VIC chip reads. I should be able to set up the same type of circuit in Logisim. Logisim Evolution has a DIP switch component, a Controlled Buffer component, and obviously LEDs, so everything I'd need to set the same thing up. Because I'd have complete control over the clock, I could pause it at a certain point, change a control register, then manually clock it for a bit and see what happens.

The biggest thing it is missing is analog components, since it is a solely logic and digital circuit tool. What that will mean is that I won't be able to fully simulate the parts that deal with analog, such as the generation of the chrom and lum outputs. I'll need to feed the colour code of the pixel into an RGB lookup table and then have that digital value go into my custom LCD screen component.

Interesting. Have you thought about making a Verilog or vhdl file, and simulate it there? Lattice Diamond is free and includes a simulator (Active vhdl).

RAM/ROM is also supported there, plus you can get the output on a monitor if you buy a $22 card (or I can do it for you).

Kakemoms wrote:Interesting. Have you thought about making a Verilog or vhdl file, and simulate it there? Lattice Diamond is free and includes a simulator (Active vhdl).

I'll certainly take a look at Lattice Diamond. Currently I don't know Verilog or VHDL but they're on my list to learn. One of the nice things about Logisim Evolution is that it can generate both Verilog and VDHL (as long as you stick to the components that it supports that for). My plan was to get the simulation working first in Logisim Evolution and then generate the Verilog and/or VHDL as a starting point for moving it into another tool, perhaps something like Lattice Diamond. I'd hopefully learn a bit of VHDL and Verilog by looking at what is generated by Logisim Evolution.