6502.org

As of this moment, they don't have any in stock. Other than the PLCC-44 version which I can't work with at the moment.

Where else can I buy one in the US (other than eBay)?


**EDIT**

I guess I could get a PLCC-44 to DIP converter and go that route. But I already have 65c02's on hand.

The only place I've bought them is from WDC itself. I called with credit card in hand, was off in a few minutes, and the parts showed up at my door a couple of days later.
You don't even need to do that. Besides the common thru-hole soldertail PLCC sockets, there are wire-wrap ones available too (although the latter are quite expensive). See http://uk.rs-online.com/web/c/connector ... LCC+socket .




Note that these don't take any more board space than the PLCC sockets found on commercially made boards. Others here on the forum have come up with other ways to do the same thing with a little more work to save a lot of money. Jeff is one that comes to mind. Hopefully he will give a link to where he has posted pictures.

Oh yeah. Forgot that I can buy from WDC directly. So used to dealing with Mouser.

I'd love to get into wire wrapping. But I don't have any wire wrapping tools.

Plus, I read more and more on how people aren't doing wire wrapping any more. Seems to me it should still be a popular method for getting more permanent circuits built. I guess the PCB houses are getting cheaper and cheaper these days.

Part of the 6502 primer: "Answering Wire-Wrap (WW) Questions and Doubts" http://wilsonminesco.com/6502primer/WireWrap.html

HA!

I've been reading the primer as we speak.

Here is a post detailing the PLCC-to-wirewrap solution that Garth mentioned. Actually two different but similar solutions -- "Plan A" and "Plan B"

"Plan C" is for SOIC's, but could be adapted to PLCC.

J.

According to their web site, they've stopped doing direct business, and are now referring people to the regular distributors instead. See: http://www.westerndesigncenter.com/wdc/HowToOrder.cfm

If none of the distributors have something in stock though, and if they won't even sell any of the PQFP packages to distributors, I'd call them up and point out the nonsense in what they are saying.

On the Propeller you can sample the pins at 1 sample per clock by running 4 cogs that are each 1 cycle out of phase of each other.

It's probably not too hard to program a Propeller to function as an address decoder for bank switching, however you can't make it as fast as that: your program has to read the address bus of the 65C02 and generate an address for the memory chip or module, and if you use the trick mentioned above, the cogs will interfere with each other when they put the address on the output pins.

The fastest way to implement an address decoder with a Propeller would probably read the value from INA and modify a JMP or JMPRET instruction accordingly. The JMP instruction then jumps to a few instructions that store a value onto OUTA and return to the main loop to wait for the next clock pulse. Based on my experiences with Propeddle and L-Star I reckon you can probably get this to work at some clock rate between 1 and 2 MHz depending on whether you want to store the table in the hub or not.

Another problem is that if your circuit only has one Propeller, you're going to run out of pins really quickly. My L-Star project uses all pins of the Propeller and uses monochrome video using the 1-pin video driver. It's possible to free up one additional pin (for bank switching a 128K chip perhaps) by using the 1-pin keyboard driver instead of the 2-pin standard PS/2 driver, but then the user has to press the space bar after a reset, and the keyboard lights won't work.

The bottom line is that bank switching (in a world where you are using a Propeller to bitbang a 6502) will probably require an extra chip, and if that chip only has to do bank switching, it probably makes more sense to find a different solution than a Propeller for that.

===Jac

More than likely, I will be using 2-3 Propellers for my design. But not explicitly for address decoding or glue logic.

I still think a 65C02 running at 2MHz is a nice machine so I would be more than happy to hit that target. In fact, some other designs I want to build using the *cough* *cough* Z80, would run at a maximum speed of 4 MHz. So, I'm not trying to build anything massive.

After I build a few 8 bit machines, I want to tackle either a 16 bit or maybe 32 bit machine. But by that time, I will be using FPGA (which I'm learning) with maybe a Propeller as the main CPU (or ATMEGA) or maybe even a 68K running as a softcore. Lots to learn and do.

But my first computer is going to be pretty simple. I will more than likely use some discrete logic for memory decoding and 2-3 Propellers for video, audio and I/O.

Also, I've been trying to crunch some numbers. Nothing concrete. But I plan on running the Propellers with a 6.25 MHz crystal which gives you 100 MHz. Four clocks per instruction (minimum) yields 25 MIPS. Or, 40ns per instruction.

65C02 running at 2 MHz is 500ns per cycle....250ns per half cycle. So, in theory, that gives me a maximum of 6 Propeller instructions per half-cycle. With real life being probably 2-4 instructions. So, not much can be done other than clocking a shift register, etc. So, I get it now.

I mean, you would think 100MHz microcontroller could handle a measly 2MHz 6502....

But at least I understand why it can't for many things. Not without devoting several COGS to get parallelism. But like you said, if I were to do that, might as well use discrete logic.

Lessons learned.

FYI,

I didn't notice that Mouser tells you an estimated delivery date for out of stock parts.

Looks like the W65C816S6PG-14 is expected to be in stock on the 21st of this month! (Monday).

Hope that's true. I might go ahead and buy a couple just to have.

You can place the order and they will ship when they have inventory. If you wait until they have inventory before ordering you may discover that everyone else beat you to them.

Already way ahead of you. Got two coming soon.