I found an old beaten-up CD player (a Philips AK601) next to the trash container a while back.
I picked it up, found it not worthy of being taken back into use, so I decided to open it up and maybe gut it for parts. I found a couple of motors inside, but what really drew my interest was the human interface card, or the print which holds the buttons and LCD.
The card is a one-sided PCB, with on the backside (the green side which holds the traces) three ICs, and on the other side an LCD screen (with what seems to be two incandescent lightbulbs as backlight, 8 diodes, 13 buttons, one capacitor, one transistor, a 9-strand ribbon cable that connects to the motherboard and a hell of a lot (25ish) wire links.
The LCD is an interesting one: it has 6 7-segment digits and a few more items, like whole-word PAUSE, SHUFFLE etc, separate digits 1-6 (presumably for a cd changer) and two “cd” icons.
My curiosity was triggered by how simple everything looked and I started on a quest to understand this board.
Finding the connections
My first line of attack was following the traces on the board and see where they went between the 3 ICs, the LCD, the buttons and the connector cable. At first I tried this with a notepad and eyesight alone, but quickly lost track. I made a photograph of the board, put that photograph in Inkscape, and drew coloured lines over all the traces and links.
The ribbon cable connects at the bottom right, under the numbers 1-9. I traced all the lines from the ribbon cable to their ultimate destinations with rainbow colours.
I found that all of the 13 buttons (under the red dashed rectangles) were connected, via a diode, to 8 of the output pins of the smallest IC (a Philips PC74HCT164T) on on one side, and to two of the ribbon cable pins, such that one pin connected to 8 and the other to 5 buttons.
All of the LCD’s pins connected (with black lines) to one of the two larger and identical COP472WM-3 chips (C1 and C2). Interestingly, there was one wire that connected the two chips to each other (pin 16 of C1 to pin 15 of C2).
On the lower right of the board and under the LCD, there was some wiring that I didn’t fully understand connecting the lightbulbs via the transistor, capacitor and resistors to the ribbon cable.
And further to the right there was a section that was unpopulated with any parts, I marked this out in grey lines or areas.
There were also some lines that went under the IC’s and I couldn’t visually identify where they were going.
In the meantime, I had downloaded the datasheets for both the PC74HCT164T (pdf), which is a shift register, and the COP472WM-3 (pdf), which is a Liquid Crystal Display Controller, capable of driving “4.5” digit displays. The pinouts I found in the datasheets helped me to explain what all the wires were and also what the wires on the ribbon cable were doing.
The next step was to get out a multimeter and check where all the traces disappearing under an IC were connecting to.
And while I was at it, I drew diagrams of how the chip’s, LCD’s ribbon cable’s pins were laid out functionally. The ribbon cable has 9 pins, counting from left
- Data (connecting to all three chips, C1&C2 pin 7, C3 pin 1&2)
- Clock 🕒 (also connecting to two resistors)
- Buttons (2,4,6,8,9,10,11,12)
- Buttons (1,3,5,7,13)
- Chip Select for LCD driver chip (C1, pin 4)
- Vcc (=+5V)
- Chip Select for LCD driver chip (C2, pin 4)
- not connected
The most puzzling part for me was a small section with 4 resistors, a capacitor and transistor which connect one side of the lightbulbs. The lightbulbs connect to the collector of the PNP transistor, there is a 1.5kΩ resistor from the base to the Vcc, the emitter is connected directly to Vcc. There is also an 820Ω from the base to the negative side of an electrolytic capacitor, with that side also being connected via a 1.5kΩ resistor to the clock line, which is also connected to Vcc with 3.3kΩ resistor. And the positive side of the electrolytic capacitor is also connected to Vcc.
I have no information on the LCD whatsoever. It sits in a plastic housing — if there is a serial number at the back (which I doubt) then I would have to bend or desolder 22 pins to view it— something I don’t want to do. I did not fully understand how to drive a (generic) LCD, but what I understand from RC Tractor Guy’s video, you need to cycle the LCD backplane and signal lines with a 100Hz signal, so it’s not a simple case of connecting voltage to the right pins. A better (but longer) resource on how raw LCDs work is found in EEVBlog’s video series (part 2,3,4,5). All this knowledge is unnecessary luggage, as the LCD driver chips take all the hard work of driving the LCD away from me. But to understand the chip connections, you need to know that this is a multiplexed LCD which has 3 so-called backplanes and 19 segment pins to drive the 3*19=57 segments. Interestingly, the input to the chip is as if driving 4½ 8-segments (7 segments plus decimal dot) with the 4½ byte (4 8-bit segments plus a nibble for a “half” segment) which is 4.5*8 bits, but the output of the chip is 12 “column” or segment lines (in 4 groups of 3) and 3 “row” or backplane lines for 3*12=36 segments. The datasheet mentions this only sporadically, calling the last 4 bits “special segments” which can either be ½ segment (to make a 9-digit display with 2 chips chained together) or a SP (special character, like a second dot)
Anyway: the first (leftmost) 3 pins of the LCD are the backplane lines, with pin 1 (BP C) connecting to chip 2, and pins 2 and 3 (BP B, A) connecting to chip 1. The next 6 pins are the segment lines (groups 2 and 1, all the time in order A/B/C within that group) from chip 2, then 12 pins with all segment lines from chip 1 (groups 4,3,2 and 1), then one more segment line from chip 2, group 3 line C.
All the buttons connect to the PC74HCT164T shift register, there are 5 pairs and 3 single buttons connecting to a pin each, via a diode (oriented towards the shift register).
The other pin of the buttons are connected to two lines that go to the ribbon cable: 8 buttons share line 4 and 5 connect to line 5 of the ribbon cable.
The next step will be to see whether I can actually control this card, i.e. let the LCD do something and read something useable from the keys. From that I can go on further.
I already have an idea in mind for a serious use: I’m going to make a kitchen clock/timer out of this. That is, if I can actually make it work.