Panasonic FS-PC1 internals/teardown

Par George.rm

Supporter (15)

Portrait de George.rm

31-08-2018, 08:05

Ever wondered what makes Panasonic's 48 dot color thermal printer tick?
This doubt struck me past 2AM today. So I opened it up and I didn't need to wonder anymore

Overall view of all components

Overall view of the board:

Close-up of the CPU and print controller area. The CPU, a x86 NEC V20, is an enhanced Intel 8088-compatible (the CPU in the original IBM PC) and as such, a 16-bit core (albeit with an external 8-bit bus). Printer has more brains than host system? Print controller is by Seiko Epson, and you can also see the Font-ROM, Kanji-ROM and FW ROM, for a total of 10Mbit of installed ROM.

The area with the drivers for the printhead

DRAMs, 2x 64K*4b, for a total of 64KB

PSU area. Transformer is a weird slim unit sitting flat under the printhead rail.

Printing unit. Nothing much to see here.

The also relatively unremarkable rear of the above

That's it for today, hope this is of some interest to someone. I was too lazy to pull the board to photograph the other face.

Ask away!

!login ou Inscrivez-vous pour poster

Par hamlet

Scribe (3838)

Portrait de hamlet

31-08-2018, 08:29

Thanks for sharing!
Printers are often still unnoticed technical wonders. With their mechanical parts even more than mere circuit boards.

Par sd_snatcher

Prophet (3498)

Portrait de sd_snatcher

31-08-2018, 17:22

Nice job! Maybe it's a good idea to dump its ROMs, so others can fix their printers when maskROM bit rot finally arrives? Smile

Par George.rm

Supporter (15)

Portrait de George.rm

01-09-2018, 16:04

Theoretically MaskROMs don't rot (in the usual sense of the word, applied to nonvolatile memory devices), as the bits are not set by the charge of a floating gate, but rather by the metallization layer (made using a specially-prepared litography mask, hence the name). They're akin to a punched card. They do fail in other ways, though.

In other equipment, OTP ROMs, being just an UV EPROM without a window, do rot, since the gates eventually lose the charge, leading to said rot.

In any case, I'll dump them when I get the time to desolder the ICs. My reader will be waiting, I want to see the x86 code that runs it, and perhaps see if I can change the fonts on it Evil

Par sd_snatcher

Prophet (3498)

Portrait de sd_snatcher

02-09-2018, 04:13

Say that to Leonard. He already had to change tons of bit rotten maskROMs. And Belavenutto's A1ST had that problem too.

Par Manuel

Ascended (18392)

Portrait de Manuel

02-09-2018, 09:18

Amazing to see an "XT PC" inside an MSX printer Smile

Par RetroTechie

Paragon (1563)

Portrait de RetroTechie

07-09-2018, 11:08

Googled that SLA7490F (the quad flat pack) - it's a gate array from the SLA7000 series, with some 4900 'raw gates' in it (hence the "490" in type #). Manufactured by S-MOS Systems, which is a little surprising since that was a US company. All other IC's are made by Japanese companies (Sharp, Nec etc) which is common for Japan-built equipment of that era. Logo on that gate array is from Seiko Epson Corp.

Those D6345's on the bottom are some kind of drivers. Like ULN200x but with serial input.

Manuel wrote:

Amazing to see an "XT PC" inside an MSX printer Smile

XT without all the PC peripherals.. Big smile Low power x86 chips (especially 80186 derivates, and later '386s) were once commonly seen in embedded applications like this. I once worked in a big warehouse where pick & place hand scanners used a 80C86 chip. Ran some version of DR-DOS iirc! These days it's mostly ARM, some MIPS, and a long list of CPU families optimized for very specific jobs like engine control or whatever.