3 key notes not working and can't write my savings to the banks..

Yes there's a way. Use Registration Memory (I shorten it to: RM) ..
RM have 8 banks and 4 bank numbers. For changing banks, press BANK several times.
Example :

• choose tone (eg:500)
• press BANK where you want to store. (eg:1)
• while holding down STORE button,press registration number where you want to store (eg:3) Release store button. Finish.

Next :

• choose tone (eg:038)
• press BANK where you want to store. (eg:1)
• while holding down STORE button,press registration number where you want to store (eg:4) Release store button. Finish.

Anytime you press RM bank 1 number 3 (one key), the sound is 500
Anytime you press RM bank 1 number 4 (one key), the sound is 038.

Notes:

• RM not only can save tone, but can store style,overall setting live performance,etc.
• RM functions are disabled while you are using the SMF player, song memory or demo.
• RM are also disabled while you are using the floppy disk drive.

I suggest a hard reboot / initialize. Istructions are in your manual or on the Korg website

those bubble contact strips will have to be completely cleaned, when dirty
the keys will be at max velocity(loud).

Read your manual regarding "fingered chords" and similar subjects. It sounds like you are in a single fingered chord mode...

The blow reed plate is backwards... either upside down or facing the wrong way. The open end of the reeds needs to be opposite to the draw reeds.

The graded action is only that there are different counterweights on the various keys. They are a sort of "J" shaped pieces of steel that vary in length. I don''t know how the graded part could stop working, as they just vary the key "feel" across the keyboard. Are you saying this key "feel" has been lost?

I have found my schematics. The F3# and the C3 and E3 all have common interconnects in the key scanning matrix.

The following diodes on the key contact board need to be checked:

D49, D50, D57, D58, D61, D62.

These diodes block the hazard paths of the scanning when more than one note is held down.

You MAY find some contamination is shorting one of the diodes rather than the diode being bad itself.

The scanning is of a conductive rubber key contacts so is high impedance as the contacts don't go near zero ohms when pressed. Leakage in the diodes can cause the problem. I would first visually look for shorts on the board affecting the conductors tied to these diodes.

Clean the board and contacts with 99% isoprophyl alcohol and Qtips.

You will have to remove the sets of keys that are held with several screws. I believe they are modular and the keys are in groups of white and two black groups per octave. Under the keys is a silicone strip of domes with conductive pills, two per note that are pressed onto the board. The two contacts close at different times so they can measure key velocity.

The key contact circuit board is in two parts, a high and a low section. To remove, you will have to remove all keys and the silicone strip over the section of board of interest.

We have a Yamaha keyboard YPT-330 note E and F sharp not working why?

The next step gets very complicated without test equipment.

It is important to realize how these keyboards work. ( I assumed you cleaned the black rubber dot that forms the contact.)

The keys are "scanned" by the electronics. Usually 12 bits are read at a time and the computer scans through the banks of keys that are usually 1 octave at a time and two per bank to enabling reading the velocity as the contacts for a key close one slightly behiund the other.

If there is any deformation of the contact bearing dome, that will cause a velocity sensing problem.

The keys are scanned as a matrix. To avoid current going backward through keys in other octaves, there are disconnect diodes that prevent coupling of one octave with another and between the two sets of contacts (for velocity sensing) within a single octave.

Do verrify that you cleaned the board and the black rubber dots with 99% isoprophyl alcohol and NOT rubbing alcohol as the latter contains an oil that worsens the problem.

Either the first or second closure matrix line for the "A" keys is bad.

This may be a cracked or shorted circuit board.

A bank of diodes is used to prevent sneak electrical path if more than one octave has notes pressed.

First thing is to clean ALL the key contact areas and the conductive rubber pills with 99% isoprophyl alcohol.

This involves dis-assembly of the key area. If you are not competent in electronics, best left to a pro shop.

The keys are scanned in a matrix, two contacts per key and all the individual notes are in common (all the "A;s", "B's"...) And each octave is strobed for the first set of contacts and then for the second set of contacts. an the whole octave is read in parallel.

The reason for two contacts is one closes first and then the other as a key is lowered. The time between these is measured and is the note VELOCITY which for a piano controls the loadness of the note. IF ONE of the two fail, the loudness will vary as yours does.