My Tyros5-76 Changed to a 49-note keyboard

[RoyB]

For 2 days last week, my 76-key Tyros5 decided to re-invent itself as a 49-key arranger.

On boot up, it only sounded the keys from C2 to C6 (the 49 keys on a standard 49-note keyboard) - all keys below and above this remained silent. Numerous attempts at factory/system resets and reinstalling the latest firmware failed to resolve the issue.

When I ran the 'Keyboard type check' in the on-board test diagnostics program, instead of reporting something like 'OK (FSX76)', it reported 'NG (FS49)'. Presumably Yamaha uses a common keyboard eBus for a variety of keyboards and it is supposed to automatically detect the correct type of keyboard in use.

So, it is thinking it is a 49-note keyboard. I found this also affected my installed expansion memory packs - it failed to locate them. When I checked using YEM on my PC, the '49-key' Tyros now appeared as a new keyboard in addition to my normal 76-key Tyros. Expansion packs intended specifically for a 76-key Tyros could not be assigned to it, neither could commercial expansion packs linked to my original instrument definition file. In fact, my '49-key' Tyros now generated a different instrument definition file than it did when it was a '76-key' Tyros.

So, it had truly re-defined itself as a 'new' 49-note keyboard (all other functions on the keyboard performed normally).

2 days later, it decided it had had enough of being a 49-note keyboard, and booted up normally in 76-key mode. It has behaved properly ever since.

Has anyone else experienced anything like this on any other Yamaha keyboard, or can anyone offer some sort of technical explanation for this 'temporary' malfunction?

Regards

Roy

[overover]

Hi Roy,

if a Test result is "NG" ("Not Given"), this means that the Test routine in question could not be performed successfully.

I think no 49-note keyboard was "detected", but this Test simply failed, and "FS49" is probably just the standard keyboard driver that is loaded if no particular keyboard type could be detected.

I suspect that "only" an internal connector which has to do with the internal keyboard connection is loose or has poor contact. However, since there are various internal connectors for this (e.g. E-BUS / Key-Scanner / Key-Matrix), I recommend having the T5 checked by an experienced Yamaha service technician.


Best regards,
Chris

[RoyB]

Thanks Chris.

If no keyboard was detected, why were certain keys still able to be playable and respond successfully to the key velocity test? In the Tyros service manuals, it implies that if a 'NG' result is returned, it would still report the actual (proper) keyboard type.
Also, would this change the instrument definition file?

I would also add that just before this happened (in proper 76-key mode after playing correctly for a session), all the G notes on the keyboard suddenly failed to play properly - which to me seems odd, because it is a pattern (all the G notes) rather than a 'random' key failure. It was on a keyboard reboot that it then went into 49-key mode. 

Regards

Roy

[Toril S]

Roy, don't you understand? It is G for GHOST! I think you need an exorcist, not a keyboard service person!

[overover]

Thanks Chris.

If no keyboard was detected, why were certain keys still able to be playable and respond successfully to the key velocity test? In the Tyros service manuals, it implies that if a 'NG' result is returned, it would still report the actual (proper) keyboard type.
Also, would this change the instrument definition file? ...

Hi Roy,

I don't know what exactly is going on internally if the Test item "049: Keyboard Type Check" returns "NG". The fact is that the T5 keyboard was only properly recognized if either "OK (FSX61)" or "OK (FSX76)" is returned.

As already mentioned, I think that in your case, despite the Test result "NG", a standard 49-note keyboard driver is available (for whatever reason).

I consider it extremely unlikely that this error will change the exported Instrument Info file. In principle, the content of this file is based on the Serial Number of the device, the Model ID and the data relating to the installed Flash ROM Expansion Memory.

... I would also add that just before this happened (in proper 76-key mode after playing correctly for a session), all the G notes on the keyboard suddenly failed to play properly - which to me seems odd, because it is a pattern (all the G notes) rather than a 'random' key failure. It was on a keyboard reboot that it then went into 49-key mode. ...

This could also come from a loose internal cable / connector (as mentioned), for example in the area of the Key Matrix circuit boards (MK76L, MKC, MKH) or the Key Scanner circuit board (EMKS). The Key Scanner IC (E-VKS) may also be defective .

The most likely thing for me is an internal loose contact ...


Best regards,
Chris

[RoyB]

Thanks Chris.

I'm tending to think beyond a loose internal cable connection at this stage (but not ruling it out entirely). It has behaved perfectly for 6 years, and for at least the past 12 months it has been on a solid worktop and has not been moved a mm.

My initial thoughts were either a mis-behaving component or some reaction (perhaps a component or soldered joint) to fluctuating room temperatures - it is a room that can get quite cold at night. Having said that, it's been in the same place for 6 years and not bothered about the temperature before when it has been colder.

Something definitely changed the instrument definition  file - the hex data was different, YEM recognised it as a different instrument (but with the same serial id and amount of flash memory), and expansion packs linked to the original .n27 file could not be assigned to it.

Is anyone able to decode .n27 files? I have copies of the original .n27 file and the changed .n27 file.

All other tests returned OK. And it has been back to normal for several days now (but I haven't tempted fate and run the test diagnostics while it is working OK).

Toril: yes, G is for ghost, and I have lived in houses before where unexplicable events occurred. In my previous house, for a few years every January/February, the circuit breaker for the downstairs lighting would mysteriously trip in the middle of the night over several nights, even though circuit tests showed the circuit to be OK.  It never happened at any other time of the year but for 4 or 5 years happened regularly in January/February. My wife (who is quite sensitive to these things) also felt a 'presence' in the bedroom at those times.

Regards

Roy

[RoyB]

I now have some new information which is interesting.

I have now run test diagnostics whilst it is working correctly. The keyboard type check is now properly returning 'OK (FXS76)'.

However, what is interesting is the first test - 'Version'.

'Version' is now returning 'Model Name : Tyros5-76 (E), which is the model I have.

However, when the keyboard was malfunctioning as a 49-key unit, the 'Version' returned 'Model Name : Tyros5-61 (e)'.

So, when the keyboard error occurred, did my Tyros boot up thinking it was a Tyros5-61, tried to find a 61-note keyboard, couldn't find one (because mine is 76-notes) so defaulted to a 49-keyboard state? 

In which case, this might mean that that the real cause of the problem was that there was an error when it read the model version information (which could also explain why the generated .n27 file is different).


Regards

Roy

[Toril S]

Very interesting indeed!