Electronic keyboard repair (I hope)

Stuff with micros locking up, often is a heat problem to do with the power supply, especially if the item really is 20 years old, and more especially, if it makes use of a switchmode power supply. As a first move, I would feel inclined to pop a meter on the main 5v rail, and see where it's at, then leave it on there, and see if it's any different, when the unit locks up. A 'scope on there wouldn't hurt as well. The rail should ideally be no lower than about 4.85v and no higher than about 5.2v. The 'official' limits are a little wider than that, but it won't have been designed to run outside of the limits I've said. The rail should also be 'clean' with hash or ripple no higher than say 20mV.

If there are any issues with voltage level or noise, look to caps and resistors in the regulator circuit, for your trouble.

If the power supply turns out to be ok, the next place I would turn my attention, is to the system control processor's clock crystal. Old xtals sometimes get reluctant to oscillate, and will just stop after a short while, in just the way you describe. You can often 'prove the point' by taking the xtal out, and refitting it reveresed to its original orientation.

Finally, I have a dim recollection of a keyboard having a problem like this, some years ago. As I recall, it too had a couple of large logic boards, and ribbon style interconnects looping the boards and power supply together. The 'launch' level of the power supply was fine at 5v, but by the time it reached the second board, having passed through tracks from one end of the board to the other, and the in and out board connectors, it had dropped to perhaps 4.8v, due no doubt to age-deterioration of the connector plating, and the large current that all this logic draws on the older stuff. Anyway, it was enough to cause the thing to keep locking up. Eventually, I decided that there was no particular need for the supply to the furthest board to loop through the nearest one, so I gave it its own heavy gauge wires straight back to the power supply, which cleared up its problems completely.

Arfa

It`s just possible that a reset to factory settings might help[1]. There`s a list of Yamaha reset procedures here

I dont see one for the PSR-12 but it`s possible that one of the other PSR resets would work.

Ron(UK)

Here's some additional info:

The specs say "rated voltage - 9v dc".

I have only 19 yr-old $10 meter. I don't expect accuracy, but rather ballpark readings.

Directly from the 120v converter connector, it *looks* like 15v dc.

The switch/volume board has a 2 element wire that appears to feed the 2 main boards. There I measure 2v dc both with the unit functionaing -and- when it is frozen.

Did that make any sense?

Thanks, P

"Well, there's two trains runnin'. Ain't neither one goin' my way. One run at midnight, the other run just before day." - from "Still A Fool", Muddy Waters, maybe 1949

Anything's worth a shot.

But I'm having trouble interpreting such as:

PSR220 Press [+/YES], [-1/NO] + Power on

There's no +, -, YES, NO on the keyboard. There are 2 pairs of buttons with only up/down settings: pitch and beat. When the unit freezes, these buttons have no effect.

Thx, P

"Well, there's two trains runnin'. Ain't neither one goin' my way. One run at midnight, the other run just before day." - from "Still A Fool", Muddy Waters, maybe 1949

Yes...and no. The 2 volt reading is meaningless unless you can determine that it was measured at the point where supply voltage enters that particular board. There might be some sort of marking at the terminal...perhaps "+5v" or "B+" or "Vcc". That the voltage does not change when frozen is irrelevant.

That said, if it *is* the main supply terminal, it is probably low, and the unit is just barely limping along with reduced voltage even when functioning.

The 15vdc reading from the 'unloaded' (IOW not connected to anything) supply is also irrelevant. The relevant reading is taken when the supply is actually supplying current to the keyboard.

You need to find the main power supply board, and trace forward from where there. That board (or portion of the main circuit board where the DC adapter connects) is where the excess voltage is regulated down to a level that the actual circuits use.

jak

Power supply board: Takes input from batt. or converter Connections: Black wire - screws to alum. kb frame "to HP" - 3 element, goes to speaker "to DM" - 4 element, goes to 1684 PN1 (lg. board) elements are labeled DG +5D AG +5A +5D and +5A measure 5v DC to black wire "to SW" - 8 element, goes to switch/volume board elements are labeled E 5v B T M S E E M measures 15v DC to black wire

The two large boards (2 x 12+ ") are labeled 1684 PN1 and 1684 PN2 with markings for the voicings, various effects.

Any help?

Thx, P

"Well, there's two trains runnin'. Ain't neither one goin' my way. One run at midnight, the other run just before day." - from "Still A Fool", Muddy Waters, maybe 1949

"+5D" will be the 5v digital supply - ie the 5v supply for the digital electronics, and "+5A" will be the 5v analogue supply. "DG" and "AG" are the corresponding digital and analogue ground returns for those two supplies, so will be the references to measure the supplies against. You really need to be able to measure the +5D supply with a reasonable amount of accuracy - to at least 0.05v.

Arfa

That's a big help. I had no idea how they code such stuff.

To what extent will such an accurate measuring device empty my po' wallet?

Goes without saying that if I get 5v from +5D when it plays and 0v when it freezes, I've likely found my problem?

Thanks, P

"Well, there's two trains runnin'. Ain't neither one goin' my way. One run at midnight, the other run just before day." - from "Still A Fool", Muddy Waters, maybe 1949

Weeeelll, yes ... But more likely, it would be something like 4.9v when it's working ie just limping along on the lowest threshold that the micro can work at, and 4.7v when it's wrong. With digital electronics, particularly older garden variety such as this, a few decimals of a volt are enough to make the difference, which is why you need a meter that can read to half a decimal. Such a meter is not expensive, and if the one you have, even if it is a $10 job, is digital, then it should be capable of this level of resolution. The trick is how *accurate* it actually is, and that is something that you can't really know without comparing it to a 'standard' or someone else's much more expensive meter. I am shortly to be away for a few days, so will only get back on here another time or two. There are lots of others on here who should be able to help you through this. Good luck with it, and hope you manage to get a result.

Arfa

Ouch.

It's analog.

I spotted a Sears Craftsman dig. meter on sale $30 but sez "DC accuracy to 0.7 percent for accurate measurements." Not sufficient for the job?

Your help is -very- much appreciated.

I continue to try to measure differential voltage between the ekb when working and when "frozen". Just measured both putting the meter under the microscope of my eyeball. If there was any difference at all, there was a snatch-hair more voltage when "frozen". Reading was like 4.9 v.

Dunno where to go from here. Any help most welcome.

Cheers, Puddin'

"Well, there's two trains runnin'. Ain't neither one goin' my way. One run at midnight, the other run just before day." - from "Still A Fool", Muddy Waters, maybe 1949

the

the

so

to

For some odd reason it is often difficult to find a 0 ground point on music keyboards. Usually the battery negative is usable for that.

Unpowered of course

I would undo and remake any ribbon connections and closely inspect all surface mount chips with a powerfull magnifying glass and bright light , followed by looking at all solder joints.

then powered

Use an empty plastic ball point pen barrel to push all chips, tap other lumps ,waggle wires etc while powered up to see if anything changes , with a note engaged by sticky tape.

-- Diverse Devices, Southampton, England electronic hints and repair briefs , schematics/manuals list on