Inside my Yamaha piano

Krw is immune from accusations of senility. His output has always been wort hless, so there can't be any degradation as he gets older. And since krw is a no-trick pony he's scarcely in a position to criticise someone who has m astered even one one trick, not that Jim's a Bob Widlar or a Barry Gilbert.

In fact, there may be no actual contacts at all.

... Phil

On Sat, 1 Nov 2014 16:07:38 -0700 (PDT), Bill Sloman Gave us:

You are senile for stubbornly using the news client you use.

Not an accusation. A historical, proven fact, right here in the archives. You stupid, uncivil, thumb-to-the-nose bastard.

Is the piano meant for performances? If so, it has to survive rain, strong EMI, being next to megawatt speaker decks, and traveling tens of thousands of miles in trucks and ships. A ribbon cable would rattle and wear a hole in itself unless it was laboriously wrapped and clipped down every few inches. A bundle of wires works well with differential signals and it's much less prone to damage.

Some brands routinely get it right and are quite reliable while others ( infamously Mackie) get it routinely wrong and there is nothing but trouble.

Matching the ribbon wire to the particular IDC plug is critical, otherwise it is intermittent city soon as the PVC coating begins to harden.

... Phil

Yes, I had to take it apart because it was misbehaving - one key had lost its velocity sense, and was sounding loud every time it was pressed.

It works properly now, after I've reassembled it, but I've no idea why - I couldn't see anything wrong with it, and given the way it appears to work, there's very little that's specific to one key. Perhaps there were some contaminant particles which got released when I took it apart.

For each note there are two contact areas. Each contact area consists of two interlaced circuit board tracks that are not connected together. When a key is pressed they get bridged by what I take to be a resistive rubber. This certainly suggests some velocity sensing arrangement, but I'm far from convinced that this is how it works, since experiments I did while it was disassembled, but connected, indicated that one of the contacts is arranged to be pressure sensitive (i.e. how hard the rubber is pressed down). Indeed, it wasn't clear that the rubber for the second contact area (which is displaced upwards relative to the first) would always make contact at all.

This seems rather crude. No hall sensors. No optical switches. I'm surprised that it works, but it clearly does, and has done for 10 years or so.

The entire keyboard is a diode matrix, and there is an on-board IC that is presumably cycling through the keys, and sending information through the eight wires.

Bear in mind that I'm talking strictly about the keyboard - the distinct part of the piano that contains the keys. All the rest of the piano is separate from it, including power switch, audio jacks, display panel, etc., so the eight wires are for dealing with key movements, and nothing else.

Sylvia.

I was not griping. I don't see any basis for thinking that I was griping. It's all in your imagination.

Sylvia.

Actually, I'll take some of that back.

A closer examination shows that the rubber pad has two tips, one of which is almost touching its contact area when the key is up, and one which is considerably further away.

So as soon as the key is pressed a bit, the first contact area is bridged, but the second is only bridged a while later, when the key has been depressed almost to its fullest extent. Thus indeed providing a velocity measurement.

Quite what the failure mechanism is that leads to loud notes still eludes me. It implies that the two contact areas get bridged close together in time, yet there appears no plausible mechanical mechanism to cause that spuriously. If the distant contact area is contaminated so that it appears bridged all the time, the note should be unplayable, since it would appear to the IC to be pressed continuously.

If it happens again, I suppose I'll have to do more experiments. Maybe I'll get lucky, and it won't.

Sylvia.

Nothing senile about it. Google offers better service than Thunderbird. Your addiction to the interface you learned to use more than a decade ago does suggest a certain ossification of your mental processes. The demand that everybody else share your disability is typical of those who have aged to the point where they can't master anything new.

You're wrong. He is a masterbaiter.

From your description that follows, I suspect the surface of one of the rubber pads "lost some of its conductive properties". E.g., a film on it or the contact area to which it mates (?)

In your followup message, you've sussed this out.

Large keyboards have a big incentive to "cut corners". I've seen some that simply drove a resistive ladder to a "common" to sense which (ONE!) key was pressed.

Which still leaves the protocol up in the air. Obviously serialized but no idea as to whether it is a differential signal, multiple "bits" at a time, etc.

OK. That wasn't obvious from your description. E.g., a "keyboard assembly" could conceivably have a power switch nearby (which connects to the balance of the piano via the same cable bundle).

And I see you haven't changed at all, for years now.

Jamie

I suspect the "first down" area is reluctant to make a contact. Either because of a foreign mass (which could be dislodged) that it has to flex around *or* a film that has developed on the contact area (of the PCB or rubber pad).

So, more force is required to make that connection (which is absorbed in the natural flex of that contact area).

Meanwhile, the "second down" area has traversed its gap and similarly made contact. Thus, the relative times of these two contact event appears shortened.

I had a PC keyboard that perpetually had this problem. I would disassemble it every few months and wipe the contact areas with alcohol. Then, reassemble it and be good for several more months.

While this was a trivial operation (remove keycaps, split case, clean, reassemble) and somewhat "therapeutic", I finally "woke up" and realized I should just replace the entire keyboard as it was far from a "one time" event!

Mine, and most these days, have carbon-impregnated rubber contacts just like those in remote controls. You can see pictures of them here:

Bullshit.

If I was designing one of those, I would seriously consider having the IC on the keyboard assembly spit out MIDI. That might simplify the construction of the main board and allow reuse of components they already make.

And the UnoLinux nym is teaching his little brother "Big Jamie" how to become a master-baiter himself. Good change for you, Mildred Foolbrick! You both need an occupation of some kind. It seems that you already have a hobby.

What's wrong with you? There was no gripe that I could detect. Are you deficient in reading comprehension? They have classes for that if you are not too brain-dead to learn.

That is just crazy. Google's "news" "service" is legendarily substandard.

Nonsense. It works better on a purely technical basis.

He rollin; you hatin....