NSFW! - You must be 18 or older to view this thread! - NSFW!
The other day I decided to take our Yamaha AN1x
apart to gaze upon it "in the buff" and scope out the naughty bits. This keyboard is a synthesizer, of the "analog modeling" class. That is, it is a DSP implementation of an analog synthesizer, so there are things like VCOs, VCFs, VCAs, function generators, etc. in there but all are done numerically rather than with analog circuitry ala MOOG. It can do a pretty nice pipe organ, and the EP patches aren't too bad, but as you might expect it isn't nearly sophisticated enough to make realistic piano sounds. Polyphony is a low 10 I believe - this is more of a lead instrument than anything else. I don't think Yamaha made too many of these, ours was a Sam Ash floor demo so we got it pretty cheap a long time ago.
My main interest in disassembling it was to examine the X-Z controller on the left hand side of the instrument. Being X-Z, it senses where your finger is on it in the X (side to side) direction and how hard your finger is pressing in the Z (downward) direction. If you've ever played with an AN1x for any length of time you know how lively and responsive this controller is - I've never encountered anything quite as musical on any other keyboard. Patches generally use the X axis of it to set the cutoff frequency of the low pass filter. When the order of the filter is high, even slight changes to the cutoff frequency can make a dramatic change in the overall sound. It's loads of fun to play with. OK, on to the disassembly.
Here is the AN1x disassembled. You have to remove every freaking screw on the bottom to access the insides. The bottom then lifts off and everything is securely mounted to the top plastic section. I don't know about you, but in the past I've been burned by removing too many screws, so it always gives me a queasy feeling when I end up having to remove them all. Anyway, you can see the massive PWBs for the keyboard assembly and for all the I/O, along with a central main board, and a tiny board for the X-Z controller located above the pitch & mod wheels (shown here on the right since the unit is flipped over).
A view of the main processor board. That's a lithium backup battery at the upper left, with 2Mb SRAM below it and a general purpose H8 processor to the right of it. A daughter card is plugged into the center of the main board.
A view of the daughter board unplugged and sitting on top of the main processor board. Same exact two chip set on both, must be some kind of DSP, perhaps with associated ROM. That's some DRAM and I believe the D/A converter in between.
A view to the right side of the guts. Here you can see the pitch and mod wheels, which BTW are only supported by their respective pot shafts & pot bearings (not the most rugged arrangement). Above them is the X-Z touch controller. The touch pad mounts under the tan PWB, though I had removed it by the time I took this picture, so the PWB screws are shown removed here. The pad plugs into that white connector near the top of the PWB. That's a dual op-amp (4558) on the PWB, not much else, though the trimmers are something of a surprise - EEs always try to design those out if they can - they are expensive, and literally no one wants the added manufacturing step of adjusting them on the assembly line.
The X-Z touch pad removed and sitting on my bench. To remove it you have to remove the 4 screws holding in the tan PWB, then slide the plastic sleeve on the electrical connector forward so that it releases the printed mylar cable.
And here is where the crying starts. I really thought the X-Z touch pad was capacitive in nature, like those laptop mouse pad thingies. When I first opened up the AN1x I was somewhat shocked to find almost no circuitry supporting the touch pad (see previous photo) but when I peeled back the top protective plastic on the pad, the awful sickening truth was laid bare. My beloved touch pad, with which I'd had so much fun, so many fond memories, turned out to be a cheap contrivance made of carbon deposited conductive plastic! I had been taken in by a simple resistive sandwich! There are some things in life one is better off just not knowing, and for me this was one of them. I went and took a shower but couldn't wash away the tawdry, used feeling I had.
Afterward, listless, I played with the touch pad for a while on the bench. Of the five connectors, two are for bulk pressure (~20 Ohms pressed hard, infinite Ohms untouched), and the other three behave like a ~9k Ohm potentiometer - but only when there is a finger pressing on the touch pad. Plugged into the AN1x, I saw a 15ms period inverse 5V sawtooth when the pad wasn't being touched, and almost full 0-5V control range for both axes when touched / pressed (minus a diode drop perhaps).
Here are some of the parts on the main board:
HD6413002FP16 - H8/3002 Microprocessor, Hitachi
HM6281288LFP-8 - 1 Mb (128K x 8 bits) SRAM, Hitachi
LC321664AJ-80 - 1 Mb (65536 words x 16 bits) DRAM, Sanyo
HD62098 - Stereo D/A? Hitachi
Yamaha XS467 - house numbered, uP ROM?
Yamaha YSS236-F - house numbered, DSP?
Yamaha XT113B0 - house numbered, DSP ROM?
I buttoned it back up and checked the keys, wheels, and touch pad - all worked as before, but much of the magic was gone.