How "vintage" is the PAL-1?

54 views
Skip to first unread message

Magnus Olsson

unread,
May 9, 2022, 4:31:44 AMMay 9
to PAL 6502 computer
The following question is more for fun than anything else, but I've been thinking about how realistically vintage the PAL-1 is, hardware-wise. From a software and UI point of view it's an exact clone of the KIM-1, but some of the components weren't around in 1975.

I first got interested in electronics a few year later, around 1977 or 78, so I'm basing the following on what I remember from that time.

The 6502 and 6532, as well as the TTL support logic were all around in 1975, of course. I think the 556  time circuit is vintage as well.

I think that the RAM and ROM chips are too large for 1975 and are probably from the early 1980s - is that correct?

The large-digit LED displays are late 1970s, I think. The original had smaller digits relative to the footprint of the component.

The PCB is higher density than the original KIM board, perhaps because it has more layers. Could they make PCBs like that in 1975 or is that newer technology?

And I'm not sure that the ceramic capacitors in 1975 were as small as the new ones for that range of capacitance,  but that may be wrong.

As I wrote, these questions are just for fun and for completeness. The PAL-1 is certainly vintage enough for me, though it would of course be nice to have an original KIM-1.

Magnus Olsson

unread,
May 9, 2022, 5:41:51 AMMay 9
to PAL 6502 computer
I forgot the 7805 voltage regulator. That may also have been around in 1975. 

Oh, and let me stress that this is not intended as any kind of criticism of the design of the PAL-1. It's sufficiently accurate to its period for me.

GN L

unread,
May 9, 2022, 9:38:21 PMMay 9
to Magnus Olsson, PAL 6502 computer
This is a very new point about the system!

The PAL-1’s strategy is to find the middle point between the vintage and the price, the more vintage the higher price, vice versa.
> --
> You received this message because you are subscribed to the Google Groups "PAL 6502 computer" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pal6502+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/32b70da0-9758-4d2a-94e4-750f6683ffbbn%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

Best,
Liu

Hans Otten

unread,
May 10, 2022, 3:39:28 AMMay 10
to PAL 6502 computer
The PAL-1 is not 'vintage', it is a replica with old 65XX
 and modern parts.
A real KIM-1 is not possible, the parts are not available anymore.

Magnus Olsson

unread,
May 10, 2022, 4:46:09 AMMay 10
to PAL 6502 computer
Yes, I'm aware of that. That's why I put "vintage" in quote marks. Perhaps I should have used the term "period" instead (meaning technology and design true to the period). 
The question, anyway, was not if you could build a "real" KIM today but which of the components of the PAL-1 are not "period", i.e. too modern to have been available in 1975.

But the other side of it is also interesting. The 6530s were custom programmed for the KIM-1 so I suppose they are impossible to get today (unless you can find a scrapped KIM-1, but then it would be better to repair it than to appropriate the circuits for a new machine).
Which of the other original parts are difficult or impossible to obtain today,  even as pre-used components salvaged from other systems? 
 

Magnus Olsson

unread,
May 10, 2022, 5:03:46 AMMay 10
to PAL 6502 computer
It's a point which interests me a great deal, actually. There's a whole spectrum between preserving a real museum artifact (which I suppose the real KIM-1 is) in working condition and just running an emulation on modern hardware. 

If I was just interested in programming 6502 assembler and trying out some old software I could have run it on an emulator.  But the PAL-1 has two features which I wanted in addition to that: it has a real physical keypad/LED display front panel, and (unlike, say , the KIM Uno) it runs on real 8-bit hardware so I actually got a chance to work with that technology again.

It's not like owning a museum piece, or even an exact replica, but it's close enough to recreate the experience without that nagging feeling you have when using emulated hardware that it's not "real".



On Tuesday, May 10, 2022 at 3:38:21 AM UTC+2 liuga...@gmail.com wrote:

Hendrik-Jan Megens

unread,
May 10, 2022, 7:43:54 AMMay 10
to PAL 6502 computer
Unfortunately I don't own a KIM-1, but I do own a pretty good reproduction:


This is as good as it gets in the 2020s - but even then the 6530 ofc had to be replaced by 6532 + ROM. 

Dave Williams told me that sourcing time-appropriate parts was difficult and costly. Since this reproduction cost me three times the price of a PAL-1 this is evident. 

Even then it didn't have an early generation 6502, so even with the appropriate architecture you might still have to consider the different generations of silicon. 

The PAL-1 really brings together the vintage architecture with smart decissions on cost-effective implementation. The replacement of the 8x 2102 with the ubiquitous 6264 is one such decision. 

Being cost-effective, the PAL-1 preserves the vintage experience in great detail, including the goal to serve as a learning and tinkering device. Even an original KIM-1 couldn't do that. These machines are now so valuable, even if I owned one I'm not sure I would dare use it.

I was looking for the year the 6264 was introducted, but couldn't find it. I'd be interested to learn.

Hendrik-Jan

Magnus Olsson

unread,
May 10, 2022, 7:48:18 AMMay 10
to PAL 6502 computer
I had forgotten that there were different "generations" of the 6502. How large are the differences between the early ones used in the original KIM-1 and the one used in your replica?

GN L

unread,
May 10, 2022, 7:56:36 AMMay 10
to Hendrik-Jan Megens, PAL 6502 computer
I guess the 6264 was introduced in late 80’s
> To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/24a2e16d-f709-4f16-b56c-b7f806494147n%40googlegroups.com.

GN L

unread,
May 10, 2022, 8:10:57 AMMay 10
to Magnus Olsson, PAL 6502 computer
In my knowledge:

The NMOS 6502 basically has four types, 6502, 6502A, 6502B and 6502C, faster and faster from left to right. (https://en.wikipedia.org/wiki/MOS_Technology_6502)

The most advanced 6502 is the 65C02, which is currently on production by TSMC, the 65C02 also has some variations over the years, I don't know much about it. Talk about the KIM-1 platform, the biggest issue of a 65C02 is the interrupt circuit needs to be modified.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/d923aff8-471d-4947-b9c5-0821dc58c387n%40googlegroups.com.

Magnus Olsson

unread,
May 10, 2022, 8:23:11 AMMay 10
to PAL 6502 computer
My personal view is that the 65C02 is a different animal - it uses a different technology (CMOS rather than NMOS) and is not 100% compatible. It would be non-period in a much more serious way than your use of more modern memory chips in the PAL-1, since it would actually change the way the system behaved. 

Michael Doornbos

unread,
May 10, 2022, 9:17:20 AMMay 10
to PAL 6502 computer, Magnus Olsson
The W65C02S is backwards software compatible, but not pin compatible. The Corsham KIM clone board is actually jumper configurable to use a legacy 6502 or a new 65c02S and ships with a brand new one. Mine is date stamped July of 2021. 

To me, a DIP package 6502 that's brand new is just fine. I have a lot of confidence this one will last beyond the end of my practical lifespan (I'm 47). I actually have about a half dozen of them in various projects, including a VIC-20 with an adapter board. 

It does have a distinct feature which is very usable on Breadboard builds: it will maintain state without the clock running. Very useful for Ben Eater's breadboard computer for example. 

Anyway, very cool that all of this stuff is available these days, both old and new! 

Happy Tuesday!

Michael 


--
Michael Doornbos
Imapenguin.com
Evadot.com
michaeldoornbos.com
myzenith750.com <—- yes I’m building an airplane in my garage (hangar)
828-450-7418
HAM: N4LNX 

Hendrik-Jan Megens

unread,
May 10, 2022, 9:20:08 AMMay 10
to PAL 6502 computer
I've been trying to find differences between the different manufacturers and 'generations' of the 6502. I'm assuming that at the very least the fabrication scale might be different. But I really can't find much on that. 

But referring to your question: my PAL-1 has an UMC 6502 from '88, while my KIM-1 replica has a Rockwell chip from '97 (R6502P). So that latter one is about 22 years separated from the original 6502 and 21 years from the launch of the KIM-1. 

Hendrik-Jan

Jim McClanahan

unread,
May 10, 2022, 4:20:50 PMMay 10
to Hendrik-Jan Megens, PAL 6502 computer
So far as I know there wasn't a significant redesign of the NMOS 6502 after the first one where they fixed the ROR issue and reduced the size slightly (to meet the original design goal--the silicon footprint of the first run was a bit larger than their target). Most of the effort went into integrated chips (like Rockwell was known for) or some of the cost reduced versions (like the 6504 that was in a smaller 28 pin package an could only address 8K). I'm not an expert on that by any means, but it seems like the CMOS versions were where a lot of the focus shifted to.

UMC is interesting. I can find where companies like California Micro Devices (GTE Microcircuits), Rockwell, and Synertek were approved second sources for the 6502. I can't find much about UMC. The UMC chip (so far as I can tell) demonstrates all the same behavior with the unimplemented op-codes as the original NMOS device did. That makes me think it is likely they were using the same design. But they don't seem to be a major player and I can't find how they fit into any merger or acquisition with one of the approved second sources.

There are online photos the the NMOS Rev D chip:


Thanks,
Jim W4JBM

GN L

unread,
May 10, 2022, 10:37:40 PMMay 10
to Jim McClanahan, Hendrik-Jan Megens, PAL 6502 computer
Now I know the NMOS 6502 has D revision~ Thanks Jim!

UMC is one of the largest fabrication plants in Taiwan, it’s also the first semiconductor company in Taiwan (1980).

It’s known for copycat the Nintendo FC (NES) chips, the UA6527 and UA6528, which most of the FC “compatible” game consoles are based on these two chips from UMC. Then UMC made the MOS65xx clone, a lots of Apple II clones, game consoles and electronic dictionaries are equipped with UM6502 processor. I have a FC clone using UMC chips when I was about 10.

I don’t know if these chips from UMC are clones or cleanroom implementations by UMC. In fact they had lawsuits with Nintendo/Ricoh and the United States (green x86) in 90’s.

And now, UMC is a global IC manufacturer with no lawsuits ;)
> --
> You received this message because you are subscribed to the Google Groups "PAL 6502 computer" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pal6502+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/CAHb_ag8bUv16zcVJdOF3yiTycF0C_K%2B%2B%3D40zBua4fwK_%2B9Uqsw%40mail.gmail.com.

Bob Leedom

unread,
May 11, 2022, 9:19:13 PMMay 11
to PAL 6502 computer
re: jim3's comment, "...there wasn't a significant redesign of the NMOS 6502 after the first one where they fixed the ROR issue..."
I had thought that my KIM-1 was pretty close to the original. I didn't find out that there ever was a change to the ROR until recently, when people playing my KIM-Venture game on even older KIM-1s reported that it was impossible to get past the dragon. Eventually we discovered that the failure of the ROR instruction (the only one in the whole program!) to operate properly makes the "Is Dragon Dead?" test always result in "No!". 
Bob Leedom

Neil Andretti

unread,
May 12, 2022, 4:42:16 AMMay 12
to PAL 6502 computer
Here's a description of the very early 6502 bug: https://www.pagetable.com/?p=406

ror.png

Magnus Olsson

unread,
May 12, 2022, 9:51:50 AMMay 12
to PAL 6502 computer
Speaking of the compatibility between 6502 and 65C02, doesn't that only apply to the legal, documented opcodes of the 6502? The 6502 had  a number of illegal opcodes that actually did more or less useful things and some programmers relied on these (see e.g.  How MOS 6502 Illegal Opcodes really work – pagetable.com). What I've heard is that on the 65C02, some of these opcodes are used as legal, documented instructions, and that they do different things than the illegal opcodes on the 6502. Is that correct?

Jim McClanahan

unread,
May 12, 2022, 9:26:25 PMMay 12
to Magnus Olsson, PAL 6502 computer
There is a program called proc_kim on my GitHub site that uses the different behavior of a NOP instruction and BCD math behavior as tests to see if a processor is running the original "undocumented" (or "unimplemented") behavior or the later 65C02 enhancements.

I have tried it with the PAL's NMOS processor and with a Rockwell 65C02.


Thanks, 
Jim W4JBM

Reply all
Reply to author
Forward
0 new messages