Repairing a used Yamaha KX88?

[dgm]

Hello.

I recently bought a used Yamaha KX88 from eBay, wanting a good piano-action 88-key controller, on a budget. The person who sold it to me told me that the only imperfections were some light scratches, failing to tell me that the pitch wheel was sticky (lot of friction when rotating, as opposed to the mod wheel which rotates smoothly), and that the velocity response on the lower end of the 'board was rather erratic.

Are either of these two problems fixable without having to forfeit a large sum of money to a repair center? I read on another website that someone with a velocity response problem on a KX88 was able to solve the problem by lubricating the "micro-switches" for each key. Problem is, I have no idea what a micro-switch is.

Any help is much appreciated. =]

[PaPi61]

Originally Posted by dgm

Hello.

I recently bought a used Yamaha KX88 from eBay, wanting a good piano-action 88-key controller, on a budget. The person who sold it to me told me that the only imperfections were some light scratches, failing to tell me that the pitch wheel was sticky (lot of friction when rotating, as opposed to the mod wheel which rotates smoothly), and that the velocity response on the lower end of the 'board was rather erratic.

Are either of these two problems fixable without having to forfeit a large sum of money to a repair center? I read on another website that someone with a velocity response problem on a KX88 was able to solve the problem by lubricating the "micro-switches" for each key. Problem is, I have no idea what a micro-switch is.

Any help is much appreciated. =]

A major problem of the KX88 is that it's unable to transmit MIDI velocities greater than 100.

[rokuez]

Originally Posted by PaPi61

A major problem of the KX88 is that it's unable to transmit MIDI velocities greater than 100.

the yamaha kx88 goes up to 127


KX5 and many of the DX series boards like the DX11 - DX21, and DX7 go 100 velocity

not kx88 tho; it definitely does 127 but u may have to strike the key(s) with some force

[dgm]

Getting a velocity response of 127 isn't my problem -- it's just that, you can hit a key with a certain force and get a response of 110 and then hit it with the same force a second time and get 63.

[danielb]

Originally Posted by rokuez

KX5 and many of the DX series boards like the DX11 - DX21, and DX7 go 100 velocity

Not true - the DX21 does not have a velocity sensitive keyboard.

[kpatz]

Cleaning the key contacts should solve your velocity problem.

As for the pitch wheel, most likely it's just out of alignment. Taking it out and putting it back in correctly (making sure the wheel is installed on the pot correctly) should take care of that as well.

Not too difficult if you aren't afraid of opening it up. Otherwise, find a tech, it shouldn't take too long and won't be expensive to repair.

[dgm]

Thanks for the advice, guys.


I opened up my KX88 today, and found out that the pitch wheel mechanism in my unit was actually missing a piece that was essential to smoothing out the movement! So, I took out a couple extra pieces, which allowed it to function almost as well as it should. Which I'm totally happy with. =]

Tomorrow I guess I'm going to unsolder the keyboard from the chassis. Which is pretty scary, since I've basically never soldered anything in my life. I'm confident in my ingenuity though.


Another quick question -- there were these pieces of cardboard that overlaid the circuit boards behind the buttons and sliders. Apparently, these sheets of what looked like thick aluminum foil were supposed to lay flat against the cardboard itself, but the foil lost its "stick" over the years and was just laying in the chassis. What on earth was the foil for in the first place, and can I just throw it out?

[kpatz]

Originally Posted by dgm

Tomorrow I guess I'm going to unsolder the keyboard from the chassis. Which is pretty scary, since I've basically never soldered anything in my life. I'm confident in my ingenuity though.

You don't need to unsolder anything to remove the keyboard. The connectors should unplug on at least one end (probably the end going into the circuit boards) and a few screws hold the keybed assembly to the chassis. Depending on how the screws are installed, you may have to unscrew them from the bottom of the chassis or from inside the chassis after removing the top cover and anything else that's in the way.

Another quick question -- there were these pieces of cardboard that overlaid the circuit boards behind the buttons and sliders. Apparently, these sheets of what looked like thick aluminum foil were supposed to lay flat against the cardboard itself, but the foil lost its "stick" over the years and was just laying in the chassis. What on earth was the foil for in the first place, and can I just throw it out?

It could have been for RFI/EMI protection, but if the "foil" wasn't connected to anything (e.g. grounded to the chassis) it likely isn't doing much of anything. I'd remove them, just so they don't touch something and cause a short circuit.

Post some pics of the inside while you're doing it.

[dgm]

^ You're a life saver. =] I didn't think of taking pictures, but it's probably the least I can do to help out the community.

Last night when I was doing pitch-wheel surgery, I noticed that there were these two little wires directly soldered onto the keyboard at the left end, and there was seemingly no way of disconnecting them at the other end either.

I can't really leave those wires connected to the keyboard while I'm working, can I? I don't think I'll even be able to rotate the keyboard to turn it upside down if those wires remain connected.

[kpatz]

Originally Posted by dgm

Last night when I was doing pitch-wheel surgery, I noticed that there were these two little wires directly soldered onto the keyboard at the left end, and there was seemingly no way of disconnecting them at the other end either.

I can't really leave those wires connected to the keyboard while I'm working, can I? I don't think I'll even be able to rotate the keyboard to turn it upside down if those wires remain connected.

Those sound like the wires going to the aftertouch sense strip. Where does the other end go? To a circuit board? If the wire is soldered on both ends, if it's a small board you might be able to remove that entirely and take it out along with the keybed assembly. Either that or it's *part* of the keybed and it comes out with it. Post some pics and I can advise better.

[dgm]

The solder joints are circled in red:



As you can see, the soldered wires are encased in a grey insulation, which you can follow to the end of the picture. They're bundled up with a lot of other wires. Past the end of the picture, those wires are bundled with *another* set of wires, and I really can't figure out where any of it goes from there. Seems like it'd just be a huge pain to try and disconnect the other end, wherever that may be, and un-bundle all of the wire, etc.

[kpatz]

You might find that wire is bundled with other wires going between the keybed and the main board, and you can disconnect them all on the other end.

Otherwise, unsoldering them and resoldering them won't be hard. That is the aftertouch strip, like I suspected.

[dgm]

Finished up the contact-cleaning job, and took some photos while I was at it. For the record, it actually really did make a difference in the playability of the keyboard. Responds much more like a pro MIDI controller should. For those who have any velocity-response issues with a similarly constructed keyboard, here's how it's done.

First, unscrew the chassis, desolder the keybed, unscrew the keybed (easy). Next, you're going to want to tilt the back end of the keybed upward, revealing the sockets where two cables connect the keybed circuit boards to the rest of the instrument. On the left and right edges of the socket, you've got these little black arms that swing outward to unlock the cables.



You can clearly see the two black sockets and locking arms in the photo of the bottom of the keybed below.



Within the blue rectangle marking on the photo, you can see an example of one of the 88 triple leaf spring contact sets. It's composed of three parallel strips of metal, with an upper, middle, and lower spring.



When a key is resting, the middle spring is in contact with the upper spring (all of the photos are essentially upside-down):





In this photo, the key associated with the shown contact set is halfway depressed. You can see the white arm of the key pushing "down" (again, upside-down photos) on the middle leaf spring. At a certain point, the upper spring was stopped at its resting point, and now, the middle spring is sandwiched between -- but not touching -- the upper and lower springs:





In this photo, the key is fully depressed. The middle spring eventually was pushed far enough to come into contact with the bottom spring, and push it past its resting point:



The key strike velocity is measured, of course, by the amount of time it takes for the middle spring to traverse the gap between upper and lower contacts.



Refer to the photo of the keybed bottom. Here, we see two red circles. These are two of the numerous screws that hold the circuit boards and leaf springs to the keybed. You'll need to remove these.

After removing the boards, you'll want to take a cotton swab and dampen it with some cleaning fluid. I used rubbing alcohol, and it worked well, but kpatz recommended contact cleaner. You then carefully lift each spring from the plastic stopper just far enough to give your cotton swab room to lightly scrub the metal-on-metal contact points, but NOT far enough to irreversibly bend them. I found that trimming some of the cotton off of my swabs made them slimmer and more suited for the job.

After you're done, just reverse the disassembly to put it back together. Your keyboard should have much better velocity control, especially if it was oooold and dusty like mine was!

[dgm]

Update:

I just went ahead and took out each and every key, cleaned off the lubricating grease from each key fulcrum and black key guide, re-greased the keys, oiled the guides, and put 'er back together.

It worked like MAGIC. The keys slide down frictionlessly and evenly now, whereas they had previously been sticky and inconsistent.

The lubricants I used were:

Fulcrums: Roland Pink Key Grease -- part no. 17041858 (ordered by phone)
Black guides: DuPont Teflon Silicone Lubricant (bought at Lowe's)

Both of these are lubricants are designed for plastic-on-plastic application. Vaseline and WD-40 would be convenient -- until they started to corrode and crack your keys, that is.

The silicone lubricant isn't as effective at the fulcrums, so Roland's special goop was necessary. However, if applied to the key guides at the front of the keys, the friction of the key sliding over the pink goop actually applies too much counter-torque to the key rotation as the key is being pushed back upward by its spring, and it'll prevent you from doing fast repetitions, e.g. trills. So the thinner Silicone spray lubricant was ideal.


This KX88 is really close to being in mint condition. I highly advise anyone with an old keyboard to go through the aforementioned contact-cleaning and re-lubricating procedures. Hopefully this thread helps out someone else with a less-than-phenomenally performing old keyboard!

[kalpajazz]

Hi dgm, I have a KX-88 bought a year ago, and has the next problem with the key velocity: In the velocity range below 80 works in a normal way... but with values higher velocity jump abruptly 80 to 91, then to 110, then 115 and then to 127!!!! (these values are measured through MIDI-OX). My question is: this is a problem in my unit or occurs in all the KX??? Thank you for your help!!!

[ScoobyDoo555]

Glad you got it sorted - I LOVE my KX88. Also got an A80, and whilst great for live work (nice light action) and lots of master-controller options, it just doesn't have the feel of quality that the KX does.

I've been lucky with mine - it doesn't haev any issues with the keybed.

To the poster above, it just sounds like issues with 15/20 year old keyboards - just one of those things, and with the correct maintanence (sp?) will serve you well for a great more many years

Dan

ps - great write-up!

[kalpajazz]

Hi, ScoobyDoo555!
Then... you can confirm that your KX-88 sends all values of key velocity??? (100, 101, 102... 112,113,114)...

[vtester]

I have 2 KX88 controllers. One had a well known problem like controller can't be switched with MS1...MS4 buttons. It always goes in DUAL node and became be stuck for control, but still allowing to play.
I had found several recommendations like standard RESET, replacing of the battery (my was good), making a loop from MIDI OUT to MIDI IN (I don't understand this trick). Nothing could help.
I remember, that I found one sticky key earlier. This was a reason. The same response controller gives with disconnected keyboard flat cable.
Yamaha should add this information to his F.A.Q. for KX88 or similar.


You can see a dry dirt from some liquid shit.dfegad


Next 2 photos shows important ZERO initial positions (2-nd photo) of upper and bottom contacts for measuring speed/force of hitting.

[ScoobyDoo555]

Originally Posted by kalpajazz

Hi, ScoobyDoo555!
Then... you can confirm that your KX-88 sends all values of key velocity??? (100, 101, 102... 112,113,114)...

BUgger - missed this
Yes, I can confirm that - checked as many as I could, and it seemed to transmit them all.....

Sorry about the LATE reply!

[kalpajazz]

Originally Posted by ScoobyDoo555

BUgger - missed this
Yes, I can confirm that - checked as many as I could, and it seemed to transmit them all.....

Sorry about the LATE reply!

Thanks for your response scoobydoo! After disassembling my kx-88 could see that the mechanism that reads the velocity value is different from the one shown in the pictures ... is only the metal part and has no white plastic triangle. The contacts are clean, as if the mechanism is new.
Anyway, clean some, but the problem continues.
It is strange that the problem is in the 88 keys (the sensitivity values jump from 80 to 110, for example) in all keys). This makes me suspect that the problem is not the keyboard but the
electronics.
Anyway, thank you very much for all your help ... the forum is really good!
Best to you all

[donkeyote]

It would appear that the MS1 through MS5 buttons on a KX 88 need to have all the keys up to work. When up, they rest on Contact 1 (when played they travel to contact 2 for a velocity/time calculation), so if you have any non working keys and the reason is they're not making initial contact, then there's no way to get out of dual mode (default setting) since MS1 toggles single/double. In addition, the rest of the buttons also become non-operational once a MS button has been pushed. You can sort of test this by playing keys (in a functional unit - which mine is now!) and pressing MS1 and then other buttons. Confused - don't be, just make sure all your keys are working well and you're not playing upside down. I fixed my key by opening up the unit, tilting the keyboard back, removing the key (don't get me started on key removal--major pain to put back in, even to pullout) and bending the contact switches to work again. No amount of cleaning would work. You should see some resistance vs. open between the resting Contact 1 and the key has been played Contact 2 i.e. a healthy key up.

[Rael]

After a few years involving a move and recently finishing my basement I dug out my old gear, bought a couple of new pieces, upgraded my DAW software and built a new computer. So the home studio is now set up again and getting ready for some fun.

My KX88 (I'm the original owner) had a problem with internal goo on the contacts that killed C2 almost completely. I just tore it down and cleaned all of the contacts. I also removed the keys, cleaned them up and relubricated the fulcrums. The action is fine and I've got C2 back, still loving this keyboard.

One remaining problem. The keyboard is sending out a continuous stream of portamento time messages as well as occasional SysEx messages. This is happening with portamento turned off. I reset the keyboard by turning on while holding down both the Bank A and Bank B keys but still have this stream.

Could my CS4 slider be having issues (sending continuous portamento time changes)? The CS3 slider (LFO speed) sends SysEx messages so could be a culprit as well. I plan to pull out that board anyway to replace the battery so can use contact cleaner on the CS1-4 sliders. I would welcome any ideas on how to track this down.