SL-990XP weird behavior - triggering unintended notes when playing certain key combinations

[joeyluck]

I just started having a weird issue with my SL-990XP and wondering if anybody else has ever encountered something along these lines.

When I play a couple notes such as D#5 + F5 and then I play the note, C2, while holding that chord, it triggers E5 as well. Playing any other note with that chord seems fine. If I move the notes up or down one step, such as to play E5 + F#5 and then play the note, C#2, while holding that chord, it triggers F5 as well.

This behavior continues this way for a good amount of the keyboard. And it just started out of the blue.

If I hold D#5 + F5 and then press and C2 + B1, the MIDI goes crazy and sends 100's of messages for E5 note on/off in a matter of a couple of seconds. If I hold D#5 + F5 and play C2 + A1 is doesn't trigger any unintended notes or behave funny

Any idea what is going on here? Aside from Reason, I've tested the SL-990XP with several different pieces of hardware, running directly to them and trying different MIDI cables, and it's the same weird behavior each time. I tried to find if there was some sort of reset for the SL-990XP, but didn't see anything. It's not exactly a feature rich keyboard either. All the keys by themselves (and when avoiding these particular key combinations) otherwise work just fine.

Any ideas what could cause this weird behavior?

Thanks!

[Delora Software]

It sounds like your SL-990XP is suffering from "ghost notes". This is something that happens when keyboard scanning electronics based on a row-column scheme have a component failure. Virtually all "modern" (post 1970s) music keyboards use matrix scanning, as do QWERTY computer keyboards. In fact, if you see an array of more than a handful of buttons there is probably a matrix scanner behind it.

Computer keyboards usually avoid the ghosting problem by implementing a so-called "N-rollover" method using some combination of hardware and firmware. This works because people do not tend to press a lot of keys simultaneously while typing. However if they did then there is a chance for ghosting once the N+1 key is held. Music keyboards need to play chords so a more robust method is needed that uses additional electronic components, typically one diode per key. This approach is immune to "ghosting".

However, if a diode fails and becomes a short-circuit (aka "shorting out") then that key can cause other keys to ghosts whenever it is held. (If it fails by becoming an "open" then only that key is affected; it no longer triggers its note). Sadly this almost always requires a trip to a repair technician, or some time on your own electronics bench if you are DIY inclined.

This Wikipedia article is enough to get the gist: https://en.wikipedia.org/wiki/Keyboard_matrix_circuit.

[joeyluck]

Hey Doug, thanks for the explanation and link. I figured it was something along those lines.

What I found interesting is just how specific it is. I have to hold a series of those exact notes and a specific key on the opposite end of the keyboard... Move or remove a single note in the chord on the right side or move the note on the left side and all is fine. Only if I shift those affected notes together the same amount of steps is it reproducible on other keys. All key combinations using the affected notes play fine. I can play any combination of single notes affected, any combination of two notes in those chords and single notes affected, and just about any combination of any number of notes across the keyboard, just not those few in that exact arrangement.

So it's actually somewhat avoidable (for now). With how specific it is, I'm actually not sure how long it's been this way... I'll probably look at opening it and cleaning it today. Even if that won't help, I have nothing else to do lol. It is very clean though as it's always covered when not in use and doesn't move anywhere. I sent an email to Studiologic support and will see what they say. If something needs to be repaired, they might be able to narrow down to exactly what parts, what nodes, based on the very specific behavior.

[QVprod]

I remember researching the old Studiologic keyboards some years back and seeing that a lot of them had midi issues over usb but worked fine with a midi cable.

[joeyluck]

I remember researching the old Studiologic keyboards some years back and seeing that a lot of them had midi issues over usb but worked fine with a midi cable.

Thanks. This one is 5-pin only. No USB. Old school.

[hurricane]

I use an SL-990 Pro, and if it ever acts funny (which is rare) I take the power chord out, wait a few secs, and re-start my midi interface. That worked for me, although my issues didn't appear to be hardware related. There was this other time where the B-Flat key would not make a sound no matter what. Turns out I accidentally hit some buttons in Logic that did this.

[joeyluck]

I use an SL-990 Pro, and if it ever acts funny (which is rare) I take the power chord out, wait a few secs, and re-start my midi interface. That worked for me, although my issues didn't appear to be hardware related. There was this other time where the B-Flat key would not make a sound no matter what. Turns out I accidentally hit some buttons in Logic that did this.

Thanks. Yeah I tried everything I could think of. Today I opened it up and took at look at what I felt comfortable messing with. I only opened up the case, used some canned air here and there, and disconnected and reconnected the ribbon cables.

It's very clean in there since I always keep it covered, but just a little dust in the right place can certainly be problematic. Or as Doug mentions, maybe it's a short...

It's definitely not a behavior I've ever encountered on a keyboard. Other times with other controllers I feel like it's been a straight forward culprit key or octave entirely that would cause a similar behavior of unintended notes, no matter what combination of keys. But this has to be a very specific key combination up and down the keyboard. It does seem to only involve the last two octaves in conjunction with a single key in the second and third octaves.

[joeyluck]

To better illustrate what is happening, here are some images.

Green keys are those that are played.
Red keys are ghost notes.

1. This triggers a ghost note:


2. If I add the A here—no ghost note:


3. If I move the low C to ANY other key (such as to D in this example)—no ghost note:


4. If I move the notes together the same amount of steps, in exactly the same arrangement from example 1, I get a ghost note:


5. But again, like in example 3, if I move the low note—no ghost note:


6. Just another example, the same as examples 1 and 4, with notes shifted, getting a ghost note:


7. And again, like in examples 3 and 5, move the low note and no ghost note:


And to add to this, any variation of the spacing of notes on the right hand side will not result in a ghost note. And no single key causes a ghost note.

It's very specific. And with every combination, even those that cause no ghost notes, involve some keys that are culprits in other examples.

And it would seem that these key combinations are across two different PCBs. The PCBs are about $75 each. I'm still waiting to hear back from Fatar, but I'm curious if replacing the PCB above C would be the best bet in a first attempt to fix...or if it could be both PCBs...or maybe the main board? I wonder about the latter, because that's what ties them together, otherwise they aren't connected right?