Been looking around on google for this.

Officially the 6502 became available in September 1975. I'm wondering what the earliest date code for one produced by MOS would be? I've seen a 6501 as early as the 34th week of 1975 but most of the white ceramic 6502s I see are 76 or 77. The one on my OSI 300 is the 38th week of 75.

I think you have a very early one there! It would be good if you'd post a photo.

Here's a couple:

https://drive.google.com/file/d/1YgCAIS ... sp=sharing

https://drive.google.com/file/d/1Jun55x ... sp=sharing

Thanks! I do like the wibbly-wobbly hand-routed PCBs from those days!

From cpu-world.com forums:

I have a date code 3675. Bought new and never used. Anyone have any idea what it is worth? I have an older date code in my Apple 1 and I'm considering swapping them out.

They're back!

This was made with TopoR.

_________________
Curt J. Sampson - github.com/0cjs

Interesting! I'd imagine even that date code has the ROR bug, and it sounds like your Apple 1 would have it too. I was wondering if any Apple 1s had the ROR bug; was yours homebuilt or purchased from the Byte Shop or Apple?

_________________
Curt J. Sampson - github.com/0cjs

Lovely!

Mine was bought from the Byte Shop. I don't know if it has the ROR bug to be honest. It's #91 on the registry.

If your Apple 1 works, then you can enter a test program to check whether ROR works. This is perfectly safe; the CPU will produce the wrong result (basically an ASL that doesn't update the C flag) but otherwise will keep running fine. You're probably more familiar with WozMon than I am, for entering the following: This should print "ROR BUG" or "NO ROR BUG", based on whether ROR($01) sets the C flag - it should, but the buggy CPUs won't.

If you do have the ROR bug, I'd be interested to confirm whether my 6502 classification routine correctly identifies even the early 6502 as NMOS - which it should, since it doesn't rely on any ROR behaviour. I'll reproduce it here in hex format, and with an Apple 1 compatible output routine: This version includes the extra check for a naive 6502 emulator (which implements undocumented opcodes as NOPs). I've slightly modified it to produce an E for that case instead of a lower-case n, because the Apple 1 only supports upper-case.

A real NMOS 6502 should print N; a 65SC02 would print S; a 65C02 with the Rockwell instructions would print C, and a 65816 or 65802 would print 8. If you get anything *other* than N on a ROR-bug 6502, let me know and we'll puzzle it out.

I was thinking that if I owned a computer that cost as much as a house, I'd be nervous about even turning it on. Then I reveled in the delicious irony that the Apple 1 was one of the computers that was supposed to put an end to the era where a computer cost as much as a house. :-)

_________________
Curt J. Sampson - github.com/0cjs

That is indeed a (golden?) delicious irony.

I tried the two above programs in the Pom1 emulator, which evidently includes a naive 6502 simulation - NO ROR BUG and E were produced, respectively. I doubt anyone has taken the time to produce a 100% accurate emulator in this respect.

FWIW, my go-to emulator for quick-and-dirty tests is the on-line Apple 1js page; it also produces NO ROR BUG and E. I'll run it on my replica SBC when I finish up the current soldering mess on my workbench, but I have little doubt it will produce `NO ROR BUG` and `n`; it's a Rockwell NMOS 6502 with a very dodgy emulation of the video output to a serial port run on an Arduino Nano.

_________________
Curt J. Sampson - github.com/0cjs

Not to be a curmudgeon, but why are there two topics going on elderly 6502s?

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!