I am looking for a 1970's MOS 6502 for a project. Any ideas where I can look?

Ebay? Find a Vic 20 or Apple II, they should be cheap enough on ebay, and you can pull the chip out then.

I looked in my own inventory and didn't find any. But I'm curious-- Why specifically 1970's, MOS Technologies, and NMOS?

Thanks for the help guys. I'm looking specifically for a 1970s MOS brcuase I am building an Apple 1 replica, and am trying to find parts that were used on the original.

If you have any other thoughts or leads, I'd love to hear.

If the goal is to have a unit that functions exactly like the original, including illegal opcodes, then any NMOS 6502 will work, and you may have some luck tracking one down on eBay or similar. If illegal opcodes don't matter to you then you should consider a 65C02. It's pin-compatible with the NMOS part and is readily available.

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

The very first 6502's lacked a very important instruction; I don't remember exactly, but maybe the LSR or ROR or both. Replacing the processor with one that has the instruction can't hurt anything though.

Especially if you have an Apple I board, either replica or original, I'd use one of the 1 or 2MHz NMOS 6502's. I don't know if the 14MHz+ current-production ones with the fast rise times would work on that kind of big board.

ROR.

The CPUs manufactured by MOS after acquisition by Commodore fixed this bug independently of WDC.

If you can ensure fast rise and fall times for the CPU clock, you should be able to. In my Kestrel-1 experience, every pin on the 65816 was tolerant of absolute garbage without malfunction except for the CPU clock input.

In my nascent SBC design ('816 powered) I'm using a can oscillator to drive a flop, whose Q output becomes the Ø2 clock. Also, the /Q output is used to drive the /OE input of the '245 data bus transceiver. The flop's output has a faster rise and fall time than that of the oscillator and can produce more drive.

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

The 65C02 is usually more tolerant of a poor quality clock than the 65C816. The 65C02 includes a schmitt trigger on the clock input and the clock usually only drives the CPU, which then generates the phase-2 clock for the rest of the system. I had no problems putting a modern 14MHz part in a BBC micro board. If you do put a new WDC part in an old board remember to bend up pin 1 on the 65C02 - this pin is an output on a WDC 65C02 but ground on everybody elses parts!

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

Isn't that only a few inches across? If the Apple 1 board is as big as I think it is (nearly a square foot?), and has no ground plane, it would be much more likely to produce problems for a processor with fast rise times.