JANOME MC6000 Sewing machine board CHIP IDENT. CHALLANGE

[PerpetualWalnut]

Recently my grandmother has retired her old 80's Janome MC6000 for a modern Brother somethingsomething modern equivalent. The MC6000 is mechanically starting to wear out and adjustments are getting little finicky to say the least, but I thought I would try my hand at giving it a good cleaning and lub once again along with maybe even backing up the ROM chips if it had any for good measure. It still works. I'm just looking at doing a partial restore as I'm a fan of 80's computer tech; aren't we all?

Anyways, while the yellowing plastics are soaking in a retrobrite solution I thought I would go ahead and start looking at the electronics. I've managed to figure out that the four larger NEC branded chips are two programmable microcontrollers, one PROM, and an IO expander. I've found the datasheet for the 8049, however, I cannot seem to find any more information on the other three chips other than what they are. No pinnouts, no nothing.

Any thoughts?

[technix]

Grr those custom parts...

I'd say that since you already have the board out, you may want to socket those specialized chips (or just every chip) so there can be help with replacing them with modern reimplementations in the future.

[Ian.M]

The D8243C is not custom - its an IO expander without any ROM:  https://datasheetspdf.com/datasheet/D8243C.html

The 23128-209 is almost certainly a mask rom.  The -209 would specify the mask set used to 'program' it. 
The 8049-348 is almost certainly a clone of an Intel MCS-48 series 8049 MCU, with  2K mask ROM.  Again the -348 would tie into the mask set and thus the contents.
The 8041-122 is almost certainly a clone of an Intel MCS-48 series 8041 UPI (slave peripheral MCU)  with 1K mask ROM.  Again the -122 ties into the mask set.

Pinouts *should* match the original Intel MCS-48 series (mask ROM variants), but timing and other electrical specs are likely to be subtly different.

Knowing the above for the two MCS-48 family MCUs isn't going to be much help as I doubt they support program readout that was available fro the PROM and EPROM versions of the family, and if they don't, the programming voltage applied during readout will destroy them.  You should however be able to dump the 23128-209 mask ROM, and replace it with an EPROM programmed with its contents.

[technix]

Given the fact that this piece of equipment is 80s vintage, we likely have three pieces of abandonware here. We maybe should archive this if possible.

The 23128-209 is almost certainly a mask rom.  The -209 would specify the mask set used to 'program' it. 

Dump it and archive it. 27C128 likely can go into this socket.

The 8049-348 is almost certainly a clone of an Intel MCS-48 series 8049 MCU, with  2K mask ROM.  Again the -348 would tie into the mask set and thus the contents.
The 8041-122 is almost certainly a clone of an Intel MCS-48 series 8041 UPI (slave peripheral MCU)  with 1K mask ROM.  Again the -122 ties into the mask set.

It would likely require decapping to dump those chips. If you can get a dump, archive them too. If you want modern replacements, it would likely be better to use some modern hundred-pin MCU like ATmega2560 for this instead, and instead of the two separate MCU's use a single piece of software that handles both.

[amyk]

Mask ROM is readable on an 8049 (and possibly 8041) if it's a 100% clone: https://www.sbprojects.net/projects/8049spy/index.php

Contrary to popular assumption, mask ROM microcontrollers should have a readout mode. It may be undocumented but it is there for the purpose of production testing.

[gamalot]

They are μPD8049, μPD8041, μPD23128, and μPD8243

https://datasheetspdf.com/datasheet/UPD8049.html

https://datasheetspdf.com/pdf/1098227/NEC/UPD8041/1

https://datasheetspdf.com/datasheet/UPD23128.html

https://datasheetspdf.com/datasheet/UPD8243.html

[PerpetualWalnut]

Thanks for the info! I never heard of datasheetspdf before. I wonder why it never showed up in any of my searches...

In the 23128 datasheet it says "Pin Compatible with 2764"

As for the other two chips, I'll investigate the undocumented read mode if it exists.

[technix]

If we have the readouts, and if the 23128 is not programming code but pure data, we may be able to create a replacement board that uses 27C128/27C256 in place of that 23128, and ATmega2560 or a 5V Cortex-M chip on a big adapter board in place of the two 8042 MCU's. In fact if we used a fast 5V Cortex-M chip we may even be able to use the fast 32-bit core to run emulators so the original code can still be used and developed on.

[PerpetualWalnut]

If we have the readouts, and if the 23128 is not programming code but pure data, we may be able to create a replacement board that uses 27C128/27C256 in place of that 23128, and ATmega2560 or a 5V Cortex-M chip on a big adapter board in place of the two 8042 MCU's. In fact if we used a fast 5V Cortex-M chip we may even be able to use the fast 32-bit core to run emulators so the original code can still be used and developed on.

That would be interesting. I was thinking of figuring out how this machine's patterns are formatted so that I or someone else could make their own ROMs for them for custom monograms. Also curious to see if the original programmers left any eastereggs in the code. 

[PerpetualWalnut]

I have a MiniPro programmer, however it does not support intel microcontrollers or these NEC variants. Does anyone know of a way to add custom chip pinnouts to the MiniPro? I know that there is a few open source alternatives to the software, but I can't seem to find where the chip definitions are.

[Ian.M]

Caution: You need to find some original NEC documentation for the readout mode as the μPD804x datasheets specify abs. max. voltage on any pin with respect to ground of +7.0V, which is obviously incompatible with applying +12V to EA for readout mode as the 8049spy project did for Intel MCS-48 chips.  I wouldn't experiment on rare chips from (mostly) working equipment unless I had spares.

[PerpetualWalnut]

Caution: You need to find some original NEC documentation for the readout mode as the μPD804x datasheets specify abs. max. voltage on any pin with respect to ground of +7.0V, which is obviously incompatible with applying +12V to EA for readout mode as the 8049spy project did for Intel MCS-48 chips.  I wouldn't experiment on rare chips from (mostly) working equipment unless I had spares.

Oh absolutely! I'm not even going to de-solder any of these chips until I am satisfied with information on them and maybe even test on a few experimental chips!

[amyk]

Here is someone who claims to have successfully dumped an NEC 8049: http://www.precise-data.fr/index.php/blog/3-tuto/14-dump8049

The NEC datasheet also has a "Program/Verify Timing (ROM/EPROM)" diagram, as well as claims of "Fully compatible with industry standard 8039/8049/8749" and a nominal 18V "EA program/verify voltage high level", so I wouldn't worry about it getting damaged at 12V. It claims you only need 1mA max on that pin anyway. I would definitely test the stability of the voltage and readout ability on some extra parts, however.