Motorola SC44125P - info sought

[Meterman]

I'm looking for information on this old 8-bit microcontroller, to try and determine what variant of the MC6800 it really is.

The pinout alone is what made me strongly suspect a 6800 variant, the highlighted pins being the only ones differing from a stock 6800.

I have a second one still in one of the boards it was used on (part of a 'power line carrier' control system), and for some reason I wrote "6800ATX" on the chip itself some years back.

What I am looking for:

Ways to help me narrow down which flavor of 6800 it is - opcodes in the accompanying 2716 EPROMs I dumped suggest it is NOT a stock 6800 (based on the opcodes used - 6801 is a possibility, 6809 is very doubtful)

It just about HAS to be one of the 6800 versions incorporating a ROM, so information on how to read out the ROM would also be appreciated.

[CPUShack]

If you had a spare we could open it up and see

I have a lot of early Motorola MC68xx and did not see a match

THis is a standard product, with mask set W2L but I could find nothing else in my collection with that mask

You may browse pics of 680x and see if you can find one
I suspect 6805 or 6809 (not 6809E though)

[frag_]

Pins 39, 39 are not connected?
Can you share EPROM dumps?

[Elar]

6802? Is pin 35 connected to VCC?

[yy2013]

Pin3 of MC6802 is Memory Ready, so it does not match exactly.
Also, pin 39 of MC6802 is EXTAL and it must be connected either to a crystal or to an external clock source.

By the way, MC6801 has a multiplexed addres/data bus, so the pinout is completely different.

[Meterman]

CPUShack: Out of the 8 (just thought of a possible 9th) I have in my possession, this is the only one that is loose and potentially available for decapping... if it wasn't also THE oldest - all the rest are 1983-1986 vintage (2 on plug-in circuit boards, 5 installed in complete units). Really wish I'd paid more attention in the late '90s when we were scrapping out some boards with that chip on it, for no other reason than automatic upgrade on returned units. Hindsight = 20/20 etc.

I do have access to an X-ray machine (designed for examining circuit boards etc.) at work... how well do you think the die would show?

Can you clarify the info on the mask set you speak of? I strongly suspect this chip has onboard RAM as well.

Elar, frag: here's a schematic I reverse-engineered of the board that has this IC on it. Pins 35 / 39 are left unconnected.

frag: How do you want the EPROM dump? I have both a raw binary (4096 bytes) and a text file with the hex dump. Since I don't have a tool that will let me readily view the data as 680x assembler, I've been manually (laboriously) decoding this into readable form - and slowly getting a better understanding of both the code, chip, and the device it goes into. It doesn't help that I'm far more used to 6502 code than 6800 code, though (ex- C=64 owner speaking).

[CPUShack]

The 3 letters before the date code denote the mask set, they are found on standard products only (custom ones dont get a mask set assigned except perhaps internally)

Not sure what an extra would show, usually can only see the metal layers/vias on the die but maybe be interesting

do any of your others have something besides W2L?

[frag_]

[Meterman]

There are four products I have at hand that used this chip: CCU (main transceiver), RPT (signal repeater), LMT (endpoint device), and FCT (configuration unit).

The chip pictured is from an early RPT logic board; I grabbed the one CCU logic board I had handy and the mask letters on that one are JP4, as is the one in the FCT unit. The oldest of the 4 LMTs I have has P8W.
I can verify the rest over the weekend - not like I have anywhere to go these days.

So I guess that means that when they were building these products they had multiple versions of the same part on the shelf?

As for pin 3, the pinout I posted is clipped from the only schematic I have that shows the chip in question (RPT Logic) - referring back to it, it's tied to Vcc and nothing else. Looking at the CCU board I have, it too has pin 3 tied to Vcc and nothing else.
The FCT unit however has a trace tying it to ~IRQB (pin 37) on the 6821 PIA.

EPROM dumps are attached - offset is 0xF000. The earliest LMT unit I have uses two 2716s, the other three LMTs were upgraded to a single 2732, and the FCT uses two 2732s (continuing, the CCU uses two 2716s, and the RPT used a single 2732).

The dumps in question come from the early LMT unit - I combined the two 2716 dumps into one file so I could verify it was the same as a dump from an LMT using a single 2732 (and they are, other than the unit S/N and two pairs of configuration flags).

[Elar]

Code clearly uses MUL command so it can't be 6800/6802/6808. 6809 has different object codes. What's left is 6801/6803.

[frag_]

Yes, code is for 6801/6803. So it must be it but with custom, different pinout.
Here is disassembly, reset vector points to address 0xF761.
Code looks sane assuming there is something at address 0xE000.

[Meterman]

frag: thank you so much for that disassembly, and I agree it has to be a custom 6801/6803.

I did notice the call to 0xE000 as well - could that be the on-chip ROM? Similarly, would the code be able to call on the on-chip RAM?

I did grab the other 'CCU' card and both are in agreement in having mask JP4. I still have to look in one of the other LMTs (getting at the processor board requires a full disassembly of the unit).

[CPUShack]

Very interesting custom 6801/03 then wasn't something they normally did unless in large volume, and they updated it through at least 3 masks, even more amazing

[Meterman]

So, here are the 8 parts I have on hand:

D/C 8132, mask W2L (pull from old RPT board)
D/C 8307, mask P8W (found in early LMT unit)
D/C 8323, mask W2L (found in second LMT unit)
D/C 8323, mask W2L (found in third LMT unit)
D/C 8530, mask JP4 (on CCU logic board #1)
D/C 8541, mask JP4 (on CCU logic board #2)
D/C 8541, mask JP4 (in newest LMT unit)
D/C 8541, mask JP4 (in FCT unit)

Doesn't seem to be much rhyme or reason after all?

[CPUShack]

Tells us a couple things, all are the same chip (so no difference in mask btw the various applications)

P8W probably had some issues so they went back to W2L, then JP4 came out later , probably was a shrink