There are certain key combinations that when struck are triggering notes that are not being struck. Specifically the open fifths G to D anywhere on the keyboard, and Db to Ab anywhere on keyboard. No other combinations (noticed so far that is). When I simultaneously strike (for example) G-D, it plays G,D but also Db.  If I hold (sustain) the G and D then the Db plays repeatedly but in odd rhythms. Same thing happens with all G-Ds on the MOX8, as well as any Db-Ab (where a G sounds although not being struck).  The extra note that plays is sending out midi data for that note as well (if that is useful information).  Thanks for the help!

Although it’s possible to have problems with key matrix decoding, contamination and/or corrosion of the key contact p.c. boards and their connectors are more typically the cause of problems such as you’ve described. I checked your member profile, and seeing that you’re in Hawaii reminded me of this:
http://www.motifator.com/index.php/forum/viewthread/475543/

Strange things can also occur via MIDI, but if the problem is still apparent with the MOX disconnected from other MIDI gear, then the above should be investigated.

Thanks. Sounds like an AC environmentally controlled room in Hawaii may not be enough.

So if corrosion
etc what am I looking at? A new logic board? Is this something I can explore on my own with a meter and a magnifying glass? I ask because I’m in the last month of scoring a film for Tribeca and there are no repair people on the island I live on so I’d need to pack and ship.

There’s trouble in paradise.

Thanks.

PS
And if corroded contact in logic board would that affect notes in such a pattern, and over the entire length of keyboard?

Only occurs with open fifths starting on any G (G-D) or any Db. But interestingly if I play the g or d singly, no problem. And if I play a third note, eg GBD, no problem. Only if I play G-D as an open fifth does this occur, and throughout all octaves. Would corrosion cause this kind of pattern potentially?

PPS Some inaccurate info I posted.  Whenever I hold a G down the D sounds a Db as well, even if a third or fourth note also is being played. I.E., if Iplay GBD the keyboard plays GBDbD, If I play GD the keyboard lays G-Db-D.

Just trying to be clear that corrosion of logic board can cause this, and if I can open up and locate problem and repair without sending off island. Thanks. Sorry for the repeat posts.

Please note that I said “contamination and/or corrosion”. The key matrix is decoded by a microcontroller that’s found on the LCD board. Proper decoding requires that the data be a solid “high” or “low”. While corrosion can cause poor conductivity between parts that should be connected, it’s possible for contamination to cause connections that shouldn’t exist. Such conditions can create key data that’s not solid, and the microcontroller might misinterpret which keys are being depressed.

In the MOX8, the keys are addressed in 14 groups of six, plus a single group covering the lowest four keys (A-1 through C0), for a total of 88. That requires 15 lines to the microcontroller. Each of the 88 keys has a pair of contacts, for a total of 176. The key matrix allows the controller to need only 12 more lines to read those 176 contacts and determine which key(s) is/are depressed, and at what velocity.

The Graded Hammer 88 keyboard uses three contact boards to cover all the keys, designated GHL88L (A-1~C2), GHL88M (C#2~C5), and GHL88H (C#5~C7) (low, middle, and high range, respectively). The low and high boards are each connected to the middle board via a 17-pin connector. The middle board is connected to the LCD board via a 27-pin flat cable.

If you really want to try to repair the MOX8 yourself, I suggest that you download this MOXF Service Manual:
https://elektrotanya.com/yamaha_moxf6_moxf8.pdf/download.html
In many respects the MOX and MOXF are the same or sufficiently similar. The keyboard areas are.

Due to the sharing of data lines, problems on any of the three contact boards can cause problems in any octave. You could try disconnecting the 17-pin connections going to the middle board from the low and high boards, and see if the middle keys (C#2~C5) then function correctly. If so, the possibility of a problem on either the low or high board is more likely. You might also check the connection of the 27-pin cable.

Page 115 of the Service Manual PDF shows the 17-pin interconnections (called out as “230d") between the contact boards. Page 172 shows the GHL88 keyboard circuit.

Of course, there could be a problem elsewhere, such as with the microcontroller (page 169 shows the circuit which decodes the keys), but in my experience it’s not common.

Good luck.

Warning: For your and your equipment’s health…
Only those who have sufficient experience and knowledge of electronics and the precautions necessary to safely service a MOX8 (or any other electronic gear) should attempt to do so. If you don’t have the qualifications, please leave repair work to those who do.

Thank you for taking the time to reply with such a thorough answer.  I actually put a new logic board in just a few years back, but as you say maybe it’s the logic board--again. Wondering if MOX’s are prone to problems. I have 6 other keyboards (same Hawaii studio) and none have ever had issues, not in decades. Mox design flaw? Just my bad luck? Anyway, this is clearly over my head to figure out but there is a certified Yamaha tech repair guy in Honolulu that I will forward your response to along with the keyboard. Thank you for all the useful input. Really appreciated.

You’re welcome.

For anyone who’s interested in understanding how a key matrix works, I found a few relevant articles:

http://www.instructables.com/id/Figuring-out-a-Key-Matrix-Scan-Matrix/

http://www.openmusiclabs.com/learning/digital/input-matrix-scanning/keyboard/index.html

http://pcbheaven.com/wikipages/How_Key_Matrices_Works/

The last one explains the “ghosting” problem - how a key appears to be pressed even though it actually wasn’t. The diodes should prevent the problem, but conductive contamination on a contact board, or one or more leaky diodes (conducting somewhat in the reverse direction), can allow it to occur.