2023-03-08: I had an opportunity to troubleshoot a Mazak SQT18M Mark II with the Mazatrol T PLUS Control / Mitsubishi Numerical Control System M535LHY a while back... Just stumbled on the pile of boards and thought I should probably take some better pictures and do a quick post on it. This particular section is about the Mitsubishi QX141 Processor Card and its related Circuit boards. Kinda funny that I spent almost 9 years running Mazak CNCs and never once thought of examining the controller circuit boards unless there was a problem... There are a lot of things going on here!
QX423A
The above card is the Mitsubishi QX423A Memory Card (C1N634A442G51 Revision B). The main ICs of interest are the HM628128BLFP-7 SRAM (70ns, 128kword x 8-bit) and the HM6207HLJP-25 High Speed SRAM (25ns, 256kword x 1-Bit). This would be the 512KB version (there are also QX424 - 1MB and QX429 - 1.5MB Versions of this board). It might be fun to solder a couple extra memory chips to this board and figure out which parameters need changed in the CNC...but that is not within the scope of anything needing to be done right now...and I'm pretty sure the customer wouldn't be happy with me coming back to play with their machine again. Now if only there were beautiful hi-res pictures of the other board versions to compare this board to...
There is also a Lattice GAL20V8B-15LP Logic Array as well. (Sticker says GL1056 B 6G). Here is a link to a dump (ZIP-JED, 4KB) from the IC
QX722A
There is an interesting and somewhat techy article about the i960 processor here, curtesy of Ken Shirriff. It's always worth reading his blog even if it goes over your head a little!
This is the Mitsubishi QX722A P-Bus Communications Circuit Board (C1N634A531G52A Revision E). This particular board was the one that was giving a communications fault error. I was unable to figure out if it was this particular board or the base QX141 board causing the issue - it was easier replacing both boards and reloading everything a single time. There are an amazing amount of CPUs on all these CNC boards. This is the first one (in this writeup): an Intel i960 RISC-based processor. Assisting it are 4 LH52258AK-20 32KBx8 Static RAM ICs. Then there are two MELDAS MAC202 BHSC HG51B218FS microcontrollers, as well as a MELDAS MAC211 MCPA 4L01F1474 microcontroller (at least I assume they are microcontrollers - I can't seem to find any info on them). There are also two Fujitsu MB8441-55 8k x 8-Bit SRAM Chips. This board appears to be all microcontrollers and SRAM in a very compact form factor!
There was one questionable spot on the board... Pin 9 (DR2 Input) on the MC34051 (Dual EIA-422/423 Transceiver) appears to be just barely [dis-]connected. At the time, my bare eye-balls weren't capable of following the trace far enough to beep test it. If only there were a high resolution PCB diagram with copper traces marked... There was also the weird question of the red paint marks on the chips... Was someone working on the board previously?
QX141C
Mazak / Mitsubishi QX141C C1N634A617G51B Revision D Circuit board Top.
M-BUS is upper left white Connector
C-Bus is lower left black connector.
S-Bus is upper right right-angle Burndy Connector. P-Bus is the straight white connector found in the upper-middle right.
Bottom of circuit board
Outward Facing / Front - Where all the cables get connected!
I should probably include this here; Don't play with the DIP switches on this board without knowing what you are doing. Sigh... just don't ask...They should be off for normal operation.
Edge Connector that interfaces with backplane
Motorola MC68040RC25 CPU and Lattice GAL16V8B-10LP (GL1179 *7F) Logic Array. Configuration dump of GL1179 (ZIP-JED, 4KB)
Mitsubishi 301W000 QX141(-1) 13G Firmware (ZIP-BIN, 128KB) on a Texas Instruments TMS27PC010A-15FML LX76Z4YK8. This is a 131,072 x 8-Bit OTP (one time programmable) PROM
HN58C256P-20 Firmware (ZIP-BIN, 32KB) 32,768-word x 8-bit Electrically Erasable and Programmable CMOS ROM in DIP-28 package
Two more MELDAS Chips: MAC208 PLCA HG52E24AFD and MAC205 MEMC M60063-0508FP. Can we guess at functions: Programmable Logic Controller A and Memory Controller?
...and two more: MELDAS MAC201A DBSC HG62F43F11FL and MELDAS MAC203A SYSC HG62F22R98FH. System Controller and ?
...and one more: a MELDAS MAC204 IRQC 4L01F1480. Interrupt Controller? I wish I could find out more information on these...if only for curiosity's sake.
QX818B
Version Label: M515LWY 279W006-B0 Main
Edge Connector
When reinserting, please pay attention to the marking 1L: it means install it on the left side!
...and right here is the point where I realized I would not be able to dump the memory from this card. There are a total of twelve (12) Intel FLASH E28F020-150 2048K (256K x 8) CMOS Flash Memory Chips..
Back side of board... A bodge wire? Board is also marked BN634B548H01A Revision A. Made in USA by Mitsubishi. This board is part of the C-BUS and plugs directly into the QX141C Main Board
QX818B
QX812 M515LWY PLC / Ladder Program PLC-CMN P342A2
Pay attention to the 1R - it means this card gets installed on the right...
Edge Connector
CIN634B 596G51A Revision A Made in USA by Mitsubishi. Two Mitsubishi M5M28F102AVP-10 High Speed CMOS Flash Memory Chips 1048576-Bit (65536-Word x 16-Bit). These would almost be readable, but since the ladder is available in published form and is easily accessible from the control, it seems a little pointless.
...and nothing but traces on the back side of the board.
Did I fix anything? Well maybe... I soldered a bodge wire on the QX722A; but did it fix anything? No idea... Can't test it until there is a problem with the currently working set... Or I could send it back to Mitsubishi for evaluation and a bill... I think it'll end up back in the bottom of the electrical cabinet waiting for the machine to be retired.
As always, clicking on an image will give you a higher-res version! If you repost or further dissect these boards with info from this page, please link back to the original and let us know where we can read about your work!