| previous by date | index | next by date |
| topic list | next in topic |
Hello! I signed up originally to ask a question about a problem I was having, but I ended up solving it and thought I'd post the solution and my restoration/repair story instead. I hope this helps somebody down the line.
Backstory:
Recently I picked up a "slightly screwed up" AN1x for cheap off of craigslist. I'd always been a fan of this particular synth since I used to play one in a music store, and regretted never buying it when I had the chance. I'm pretty handy with screwdrivers, multimeters and soldering irons, so I thought this would be at worst an interesting project and at best I'd end up with something awesome.
Reviewing the synth before purchase- Mostly the only things wrong with it seemed to be physical. Some of the keys didn't respond, others had velocity issues. The touchpad had some strange behavior. It had obviously taken a fall. There were some pretty significant cracks on one side of the plastic and the metal underneath was bent at one of the corners. Miraculously it still had all the knobs. Its brain seemed to be fine though, so that was already very encouraging. I could control it with MIDI if nothing else.
First Pass:
Right off the bat I took it apart and dusted it out. I cleaned the key contacts with cotton swabs and rubbing alcohol. Lots of dark gunk was picked up from the contact board, which was a good sign for restoring the keyboard. Disconnecting/reconnecting the sockets for the touchstrip board seemed to fix it somehow. I guess sometimes you get lucky. I reassembled the keyboard components enough to test it, and pretty much all of the velocity problems were gone. Two of the keys that didn't work before still failed to make contact this time, so I took it back apart.
Inspecting the long rubber contact strip that's sandwiched between the keys and the circuit board underneath revealed that holes had been punctured in the rubber by the keys. There were little worn-out spots on the rubber strip where each key touched it, and the two keys that completely didn't work had holes all the way through. As an experiment, I rotated the rubber strip upside-down so that the contact side still faced the circuit board, but hoping that the keys would land on different unworn areas on the opposite side. It looked like they did, so I reassembled it, and the keyboard worked perfectly! Hooray!!! (or so I thought)
Here's a picture of the rubber contact. Its part number is VF834100 or VF83410R for future reference.
http://www.fullcompass.com/common/products/original/201468.jpg
With the synth seemingly fully functional, I proceeded to glue the cracks in the case back together and bend the metal bottom plate. It turns out that the sparkly dark blue color is just spray-painted onto a lighter blue plastic. It looks like maybe the same plastic from the CS1x. I discovered this when sanding/polishing the glue seams in the cracks. Oh well, I'm far more concerned with how this sounds and feels than how it looks.
The Problem:
For a couple of days, everything was going great. Then I discovered a subtle problem which is what I was originally going to ask about: I noticed while attempting to do a trill in legato mode from C3 to C#3 that it was actually causing a note-off on C3 and failing to note-on C#3. Likewise if I held down C#3 and pressed C3, it would note-off C#3 and not note-on C3. Either key worked just fine individually or in combination with any other keys. I decided to test for other problems like this, and discovered a pattern: C to C# and F# to G had this problem on every octave. My heart sank. A pattern like this could mean an electrical fault, maybe as simple as a bad connection to something more complicated like a burned out diode or IC (microchip). Factory reset didn't help. All the notes responded just fine to MIDI input, so it was definitely a problem with the keyboard hardware and not some kind of firmware/settings issue higher up in its brain.
Because the keyboard was misbehaving from the start, and because this was a fairly subtle problem, I had no idea whether it existed before I was messing around with its insides. Checking all the connections between the circuit boards revealed/fixed nothing. Pulling the keyboard contact assembly apart and put back together didn't help. I searched the internet to see if other people were having a similar issue with this or any Yamaha board, and there wasn't much out there other than the things I'd already checked. I opened up my CS2x only to find it uses a completely different type of keyboard assembly and circuitry, so I couldn't test for a bad part. Drat!
I'd read about diodes burning out on old keyboards causing problems in patterns before. Because of my fears of some kind of electrical fault, I downloaded the service manual to take a look at schematics for the keyboard contact board and keyboard scanner (KBS) board. ( http://www.deepsonic.ch/deep/manuals/yamaha_an1x_service_manual.pdf ) I busted out the multimeter and began to test every diode on the contact circuit. I was hoping it would just be a bad diode because they're simple to replace. Otherwise, it might be a problem with the KBS which would have to be swapped-out for a (potentially impossible to find) replacement.
The Solution:
None of the diodes were bad. I was filled with despair. My momentary dream of a perfectly-playing AN1x had been ruined. I resolved that I could live with this problem. I could MIDI-control it from another board if I had to. Before putting it back together, I took one last look at everything. Closely inspecting the contact circuit board itself, mapping out each key, I noticed a pattern: The contact circuits were broken up into patterns of six, starting from C# to F#, G to C, each octave. There were 10 of the groups of sixes, and then one extra for that last C on the end... Looking at the rubber contact strip, I noticed there were grooves cut in the contact-side that I hadn't payed attention to before. 10 long strips, and then one extra short strip at the end. OH!!! How foolish of me!! The contact circuitry was counting on the groups of sixes sharing a common terminal on one side. I'd effectively caused an offset-by-one error physically by flipping the touch strip around, bridging the wrong terminals!
I put it all back together, right-side-up, and the error was gone! The original problems with the worn out contact strip returned, but no more unwanted note-offs! So finally, to solve my problem with the worn out strip, I actually cut off one of the ends by about a 1/16th inch in order to offset the old contact points away from the keys so that they would touch fresh rubber. The offset was not enough to change the bridge pattern on the strip that the circuitry is relying on. I consider this a semi-temporary solution and will eventually order a new strip once this one starts wearing out. After testing, everything worked. Put it all back together, and aside from some cosmetics, it's perfect!
TL;DR:
Restored a cheap abused AN1x. In the process I caused a keyboard note-on / note-off problem in the pattern of C to C#, F# to G by putting the rubber contact strip in upside-down. Fixed it and ended up with something awesome!
Some post-project notes:
I've had some great success buying "screwed up" used instruments and gear for cheap and then restoring them to near-perfection with just a few hours of my spare time.
Often the only thing you need to do to fix a malfunctioning keyboard is to reset it (if possible) and open it up and clean the contacts on the keyboard itself, and on connectors between circuit boards. Once you put it back together you may have a perfectly functional board on your hands. If you can turn a screwdriver, you shouldn't be afraid to open your own gear up to give it a good dust-out and cleaning every once in a while. As long as you're gentle it is unlikely that you'll cause any real damage. Even then, mistakes can be fixed on your own with some additional time and research.
I regret that the AN1x is not physically designed and constructed as solidly as it deserves to be. The spray paint coating will wear off very easily and the plastic case will not survive even a modest accident undamaged.
Despite the cheap construction, the actual keyboard keys and action feels very nice and sturdy to me, especially when compared to the similarly-shaped CS2x and even some other mid to high-end boards. According to some websites, the same rubber contact strip is used in the PSR-7000, MO6, S30, SY35, QS300, and others. So maybe it would be possible to cannibalize those boards for parts. I don't know for sure, though.
In my opinion, the rubber contact strip is going to become brittle and wear out on you in the long run, even with gentle use. Especially if you jam hard or use aftertouch, which puts on a lot of pressure and eventually pokes holes where the keys touch.
Eventually I intend to replace the rubber contact strip with a new one which I've found costs around $30 to $40 on the web. Google the part number above to find a few suppliers. The firmware reports that it's version 1.02, I've seen version 1.04 floating around, so I might make the upgrade down the line.
I hope this first post wasn't too long-winded! I'm glad to see that there's a lively community for this synth, it's one of my favorites ever!