Intellivision Keyboard Component pics

[y-bot]

[y-bot]

I had evidently posted pics of this a long time ago but linked to my Photobucket so the links are broken now. Here they are again because someone asked me for them. I can't remember what is going on in these pics. Been a long time since I messed with it. A chip gets hot and it starts malfunctioning after a bit. Someday I will jump back in to this project but for now here are the pics. I get it from the family of Papa Intellvision and it appears to be different from the other pics I could find online at the time. Maybe modified or maybe pre-production? I have no idea.

[m-crew]

Thanks for sharing again...

 

Cheers

[decle]

I had evidently posted pics of this a long time ago but linked to my Photobucket so the links are broken now. Here they are again because someone asked me for them. I can't remember what is going on in these pics. Been a long time since I messed with it. A chip gets hot and it starts malfunctioning after a bit. Someday I will jump back in to this project but for now here are the pics. I get it from the family of Papa Intellvision and it appears to be different from the other pics I could find online at the time. Maybe modified or maybe pre-production? I have no idea.

 

Thanks for doing this the photos are a great resource.

 

I believe your first Computer II machine is a pre-production model, as demonstrated by the rather jumbled layout, patch wires, decidedly hacked daughterboard and possible use of EPROMS (the labels might be covering the erase windows)

 

The Computer III machine looks to be the same as RonTheCat's. The board layout has been tidied. The two white strips are small grounding fences designed to reduce electrical interference between the 3 sections of the board:

 

 

Here, the left section is the 6502 computer, complete with its CPU, TMS9927 video chip and heatsink, video RAM and the 6502 ROMs.

 

On the right is the interface to the Master Component, with the ROMs that are projected into its memory map at $7000. There are a couple of interesting things here. Firstly even though there is no CPU or video chip on the CP-1610 side, the non-standard GI bus design leads to a more glue logic and that means that it is larger than the more comprehensive 6502 side. Secondly, the 4K of CP-1610 ROMs are split into 3 chips, a single 2Kx10 RO-3-9502 (big 40pin DIP) and two 2kx8 9316Bs (two wider 28pin DIPs). I'm not sure of the reason for this split, but it would allow 2K of the EXEC to use 16bit instructions and no SDBDs, potentially making reducing the number of decles.

 

Between the two halves of the board is the 16Kx10 bit shared RAM and associated glue. This is implemented as 10 National Semiconductor MM5290J 16Kx1bit dynamic RAM chips. Again there are a couple of interesting design decisions. The 6502 divides its memory into 256 byte pages. The first 2 of these are special, page 0 acts a bit like an extended register set and can be accessed with shorter and therefore faster instructions than other addresses. Page 1 is typically used for the processor stack. The weird thing is that both these pages are implmented using the 16K shared memory. Presumably this was a money saving measure. The CP-1610 probably should not tinker with either as it could potentially cause problems for the 6502. So effectively there is 15.5K of shared memory, and all of the 6502 memory, apart from the video RAM, is shared with the CP-1610.

 

The other thing is that I don't believe that MM5290J is really a dual ported RAM, at least not in the way I understand it, as is claimed on Intellivision Lives (there is also only 16K of it, not 64K). According to the data sheet, while it does have two data lines, it looks as though they are exclusively in and out, and there is only one address bus (actually there is only half an address bus as it is multiplexed and requires two accesses to define a 14bit address). Therefore, it is not possible for the two CPUs to access the memory independently, instead the CP-1610 and 6502 have to arbitrate access to the RAM. Unless there is some clever clock phase fun, this might mean the introduction of wait states, blocking access by one or other CPU while its twin is reading or writing. Clearly, if this is the case, it would slow things down, and while this is probably OK for data that is genuinely shared, it might not be the best idea for page 0 and 1.

 

All very interesting.

 

 

[freewheel]

Those definitely look like EPROMs to me. Very cool. MS BASIC was on bloody everything back then - guess nothing changes.

[+Lathe26]

Has anyone dumped the BASIC cart? I strongly suspect 'yes' but still curious.

[mr_me]

Basic cartridge has been dumped. You can download an image file here. https://github.com/mist64/msbasic/tree/master/orig

 

That SN 766 KC could be what went out for the 1980 Fresno test market. That was followed by a redesign and delays. The computer iii revision is likely what went out at its 1981 release.

 

Interesting that the 16k ram is not dual ported. Then the two systems would have to be on the same data bus. Could they have interleaved the two systems so they don't interfere. Is the 6502 using one of the clock signals coming from the master component.

[Ron The Cat]

Good work Decle , it seems there were tinkering with the board during early production

 

For actual use of my keyboard component I guess I would need to get the image of basic actually on a cartridge to put in the back slot of the keyboard component.

 

An LTO flash cart will not work as it has a user interface to select , the load basic command on the keyboard from expect to load it directly.

 

Anyone know if the basic image works at all in emulation ?

Edited December 7, 2017 by Ron The Cat

[decle]

Good work Decle , it seems there were tinkering with the board during early production

 

For actual use of my keyboard component I guess I would need to get the image of basic actually on a cartridge to put in the back slot of the keyboard component.

 

An LTO flash cart will not work as it has a user interface to select , the load basic command on the keyboard from expect to load it directly.

 

Anyone know if the basic image works at all in emulation ?

 

You are correct, you will need a BASIC cartridge to play any of the tapes. Although this is not quite as unobtainable as it might first seem. The LTO Flash will not serve as a substitude for this as it can only supply programs to the CP-1610 CPU. BASIC is a 6502 program (it goes in the back of the unit in one of the slots behind the fold up doors) and even before it fails to work programmatically, or electrically, the 6502 cartridges are physically the opposite of the normal Intellivision ones. They have the 44pin connector in the cartridge, rather than in the hardware. So it is impossible to plug the wrong cartridge in the wrong slot.

 

As to emulation, well:

 

 

You need to use MAME, and the magic command for Windows is something like:

 

mame.exe intvkbd -cart2 C:\Path\To\The\Basic\Rom\Image.bin

[Ron The Cat]

That makes sense given its a different CPU.

 

Oh well another ten years to search on Ebay to bid for the Basic cart for $$$$$ .

 

Very nice that it can be emulated although nothing would beat Ron's paws on the KC with it.

Edited December 7, 2017 by Ron The Cat

[mr_me]

Might be cheaper to make your own Basic cartridge; doesn't look hard once someone makes a schematic.

 

The Basic cartridge is only needed for some tapes eg. Crosswords, Family Budgeting. The other tapes don't use Basic.

[+Lathe26]

Might be cheaper to make your own Basic cartridge; doesn't look hard once someone makes a schematic.

 

 

I had a same thought. Those 4KB EPROMs are standard parts (2 of them makes 8KB), as is the 74138 (a 3-to-8 decoder). I can't make out the last chip, but it is definitely a standard 74LSxx part.

 

Just for comparison, here's a proto version of BASIC with 12KB of ROM chips.

[decle]

Awww, you boys have spoiled the surprise!

 

An email went out to Frank Palazzolo requesting the 6502 cartridge port pinout last Sunday.

 

Veroboard and 44pin edge connectors were ordered and arrived yesterday.

 

 

With y-bot's images posted overnight, the mapping of the cartridge port was started this morning (the top side is proving difficult):

 

 

The ROM image was dispatched to a friend to be burned onto a single 2764 EPROM at lunch.

 

 

Merry (early) Christmas Ron!

 

 

 

 

Edited December 7, 2017 by decle

[+Lathe26]

Sweet!

 

Would it help if I provided physical measurements of the original box (i.e. the size of the spot where the cartidge was nestled)? I suspect it is unnessary since someone else likely can directly measure the exterior of the cartridge's plastic case.

 

It would be extra cool to be able to store a repro cart in there in the original box.

 

This project is both a surprise and exceptionally cool. Just knowing this is in the works will be my 2nd best Xmas present this year.

[+Lathe26]

Just curious: what's the mystery 74LSxx part? I can't quite see the number on it. My wild guess is that it is some kind of multi-input AND gate (or a cascade tree of AND/OR gates) to act as a simple address decoder that enables the 74138.

[Ron The Cat]

Wow Decle ...you don't waste time ! Did not know you could just make some of this stuff

[First Spear]

You guys are smoking my head. Just wow.

Thanks for sharing all of this! It would have been amazing for KR to see this in motion.

[decle]

All components have arrived



Now, if only I could complete the table to work out how they connect together...

Using y-bot's pictures I'm reasonably confident of the connections originating from the back side of the board, but the component side is really difficult to trace out.

Just curious: what's the mystery 74LSxx part? I can't quite see the number on it. My wild guess is that it is some kind of multi-input AND gate (or a cascade tree of AND/OR gates) to act as a simple address decoder that enables the 74138.

Like you, I'm not sure what the 14 pin 74LS chip is. The only 74LS chip I've been able to match up with the expected inputs and outputs (I'm assuming connections to the cartridge port are inputs to the chip, stuff going to the ROMs are outputs - see bottom of picture above) is a 74LS64 AND/NOR chain, which seems a bit odd to me, especially as, if correct, it would send its output to A11 of the ROM. I would have thought that something simpler like a 74LS00 quad NAND would be all that is required to generate /OE and possibly a /WE from the clock and R/W lines. Theoretically the BASIC cartridge should not need any additional support chips to map the 4K ROMs into a 64K address space, the 74LS138 can do it all if A15 is used as a chip select. Probably an error on my part.

 

It also interesting that the BASIC cartridge has 3 sockets on it. The KC memory map (pages 27-28) should not have room for a third 4K ROM. I guess it might have been mapped into the external I/O space between $8000 and $B800 or perhaps it was bankswitched, but that would have needed an external latch.

Would it help if I provided physical measurements of the original box (i.e. the size of the spot where the cartidge was nestled)? I suspect it is unnessary since someone else likely can directly measure the exterior of the cartridge's plastic case.

It would be extra cool to be able to store a repro cart in there in the original box.

 

It is a nice thought, but at the moment I'm just hoping to put together a functioning bare board.

You guys are smoking my head. Just wow.
Thanks for sharing all of this! It would have been amazing for KR to see this in motion.

 

Hmm, I don't know. Whilst I'm sure Keith was over the moon with people's continued interest in Intellivision, he was also a business man who was, quite rightly, protective of his toys. This can be seen in his admiration of the Nintendo lockout chip, and his less than positive attitude toward the work of Joe Jacobs and Dennis Clark. I suspect he might not have been totally encouraging of these efforts.

It will be very interesting to see how the Intellivision Productions stance toward the community evolves following Keith's sad and untimely passing. Speaking of which, beyond the recent flash sale, is there any news about the future eminating from Intellivision Towers?

Edited December 11, 2017 by decle

[mr_me]

Keith R might have reacted to this with something like... okay... don't you guys have something better to do with your time?

 

Since Keith R setup an office in France to program coleco vision cartridges, I think he understood the idea of third party development. I thought that Keith's position on Mattel's attempt to block third party cartridges and Nintendo's lockout chip were that they were both in violation of anticompetition laws. He understood about crossing the line when it comes to copyrights and trademarks. The homebrew community has a different attitude but Keith R tolerated it.

[decle]

 

I had a same thought. Those 4KB EPROMs are standard parts (2 of them makes 8KB), as is the 74138 (a 3-to-8 decoder). I can't make out the last chip, but it is definitely a standard 74LSxx part.

 

Just for comparison, here's a proto version of BASIC with 12KB of ROM chips.

KeyboardCartProtoBare.jpg

 

I'm still waiting on the cartridge pinout, but we do now know that the final chip is a 74ls86, thanks to this image, just posted in the ebay thread:

 

 

Another little mystery solved

Edited February 13, 2018 by decle

[intvnut]

Man... now I feel like I need to dig my KCs out of storage and open them up. :-) Of course, I have no idea where they are in storage, thanks to the movers. :-P

 

In my copious free time, naturally. :-P

[JohnPCAE]

It would be awesome if we had a schematic of the KC. I'd love to get a look at what they did to genlock the overlay video. I'm still working on my overlay video project (including color text, not just pushing it to white pixels), but it would be interesting to see their solution for getting a stable image. While achieving white text is relatively easy, creating color text requires exquisite timing precision since hue is based on signal phase.

[intvnut]

BTW, "Computer III" is mentioned on page 38 of this PDF. I wish that PDF had the full contents of the documents it's a preview of...

[intvnut]

It would be awesome if we had a schematic of the KC. I'd love to get a look at what they did to genlock the overlay video. I'm still working on my overlay video project (including color text, not just pushing it to white pixels), but it would be interesting to see their solution for getting a stable image. While achieving white text is relatively easy, creating color text requires exquisite timing precision since hue is based on signal phase.

 

I don't know what's in the circuit, but it takes a few seconds to find a lock. Watching it "cyclone up" to a stable title screen is something to watch.

 

I don't think they actually genlock in the PLL sense. Rather, I think the TMS9927 is being driven by the same clock as the STIC and so has a well defined phase relationship to the STIC. I have a feeling the CBLNK and SR1 signals combine with HSYN/VSYN outputs from the TMS9927 and some logic to gate the pixel clock to the TMS9927 until it's locked in place with the STIC.

 

But that's just a hind-side extracted theory, with little to go on.

[mr_me]

BTW, "Computer III" is mentioned on page 38 of this PDF. I wish that PDF had the full contents of the documents it's a preview of...

The "Computer III" engineering memo and the Keyboard Component service manual are there as separate pdfs. Which ones are you looking for?