Saturday, January 7, 2017

SWF A-T1501 Control Panel


Sometime in the late 90's my wife purchased an fifteen needle automated embroidery machine. Specifically a SWF A/1501.  For those not familiar this type machine here is a link to a similar machine in action SWF B/1501. This is later model of the same machine, it has similar control panel. About 8 or 9 years ago the machine required that the operating system be reload each time the machine was started and any saved patterns were lost. Eventually the operating system would not load and the machine would just sit with "SYSTEM LOADING" displaying on the LCD screen.

Some research on the internet allow us to determine that the embroidery machine utilized a embedded computer control system based on a processor similar to the Intel 80386. The computer is a SBC (Single Board Computer). The system as shipped included a 3 1/2" HD floppy drive, two serial ports and one parallel ports for input. For storage rather then a hard drive the SBC included a DishOnChip 2000 by M-System.
SBC from SWF Control Panel

Replacing the DishOnChip cured the problem, at the same time we purchased a floppy drive to USB drive upgrade. See PLR Electrionics for more information. While it was great this resolved the problem we still in the dark about how to repair the control panel.

Once again the machine stopped loading the operating system. Replacing the DishOnChip did not resolve the issue. Attempting to troubleshoot the problem was futile. We purchased a ISA backplane board see Advantech passive backplane board for more information. Tied to standard AT power supply and an ISA VGA card we were able to boot the computer but it would not display any information on the screen. It was determined that to accommodate the LED and LCD screens on the control panel the BIOS had been modified.

The simple solution was to purchase another SBC, a review of the parts book that accompanied the machine indicated that the part is number 02-100A-AT01, CPU CARD (ROCKY-386) ASS'Y. A review on the internet lead us to Penn Sew. There had to be another way. The description ROCKY-386 is miss leading there was a SBC labeled ROCKY-318 but it has a completely different form factor. Same processor, the PC104 socket is in a different location nor does it have the inline keyboard socket that the original card has.

All attempts to load DOS via 3 1/2" floppy drive were successful but the output viewed on the LED screen was limited and no information was available mapping of the keys on the control panel so it was not possible to enter any commands. We found and purchased another SBC with a similar form factor. The existing SBC had both ISA and PC104 bus systems, the PC104 is utilized in the control panel. The standard BIOS on the new SBC could communicate with the VGA card but NOT the the LED and LCD screens on the control panel. What we needed was another SBC of the same type with a standard BIOS.  We finally found one we could afford in China.

Dis-assembly and repair of the SBC

SWF Control Panel
The back of the SWF control panel has four Phillips head screws and three thumb head screws. To begin the dis-assembly loosen and remove the three thumb screws. Note the two sections are hinged together.
Back of SWF Control Panel
Moving back to the front of the machine pivot the control panel on it's hinge and look for the two data cable connections and power cable connection coming from inside the machine. Carefully disconnect the two data cable, each cable connection has two push points to release the connector. The front of the power cable connector has a release lever.
SWF Control Panel hinged open
Once the data cables and the power cable is disconnected, the four Phillips head screws in back can be removed. This releases the control panel from the machine.
Back of  SWF Control Panel on the bench
The 3 1/2 floppy drive is on the left. The board on the right is the I/O board, all communication with the outside world goes though that board. The two black data cable sockets at the top communicates with the X-Y axis drives and the needle head(s). The several white and red wires on top right communicates with the LCD screen. The ribbon cable on the top left communicates with the VFD control panel. Below the VFD cable on the left is the power cable for the floppy disk drive (FDD). The next cable down on the left communicates with LED screen. Finally the three ribbon cables on the lower left of the I/O board the parallel cable and two serial cable connections. Note on the bottom of the I/O board are the two serial ports and the parallel port for outside connection to Control Panel. Disconnect all of the cables connected to I/O board except the LCD cables. Remove the four Phillips head screws and then I/O board can then be removed.
I/O Board ready to be removed
Grasp the I/O on the side opposite the parallel and serial ports, slowly and carefully leverage the I/O board up. The I/O board is connected to the SBC below it via the PC104 connector built into the I/O board.
Back of I/O board note the PC104 connector
The danger here is to damage PC104 pins.
SBC after the I/O board is removed
The cable connected on the upper left side of the picture is the keyboard connection. The ribbon cable in the center is data cable for the floppy disk drive. The ribbon cable on the right of the picture is the parallel cable. There are four standoff screws that need to be removed and the SBC can be removed.
SBC on the bench
Once the standard BIOS was installed in the SBC it could then be tested on the backplane board. It was quickly revealed that the real time clock was defective. This type of real time clock contains a 3V lithium battery, a crystal for timing and memory for current time of day and CMOS. The modified BIOS was apparently getting to the point were it tested the real time clock and simply halted.


1 comment: