Let's Set Things Straight

And by "things" I mean the knobs on my JX-305.

They're a little bit out of line, particularly the knob for sustain, which is quite noticeably bent towards the keys.

These knobs had clearly taken a bit of a hit at some point.

And they're also kind of filthy, so let's dig inside.

First, the back comes off.

Then the main PCB is unscrewed and moved out of the way.

Then after unscrewing a few brackets we merely lift the control panel free from the... Son of a-

Ok, power switch comes out next, I guess.

Then we can finally pull the front panel board out.

As we can see, the pots are definitely filthy, and they've definitely been bent out of line. Those little legs are supposed to sit flush against the PCB.

I wasn't initially sure if these had been soldered askew at the factory, but looking at the back side it's clear that the solder joint had, at some later point, been deformed by the pins being levered up.

So that explains things.

Also just looking at this picture it seems I failed to notice that the pins themselves need resoldering. Excuse me a moment.

So let's see, where were we... Ah yes, so, flipping the board over, it becomes pretty clear that the pots are out of their proper alignment.

And some of them have even been hit hard enough to start to fold back the tabs on the bottom.

But once those tabs are pushed back into place, and once all the pins are resoldered (ahem, all the pins), things are looking much more straight.

Thanks to these refreshed solder joints.

Which are so shiny that they're basically impossible to photograph. Oh well.

Shoving things back into the case, we can see that the pots are aligned with the front panel much better than before.

And with the knobs on, everything is looking pretty much perfect.

Or at least as close to perfect as I'm gonna get it.

A Polarizing Subject

So I recently picked up a Boss SE-70. This is not to be confused with the Boss SE-50 that I already purchased a little while ago, although put a bookmark in that, we'll get back to it in a moment.

The unit was sold "for parts or repair" due to a nonfunctional display. Turning it on, we can see that indeed it is nonfunctional.

I had assumed that the likely cause was that the battery had gone dead and the contrast settings had got messed up. Then, without the display to be able to navigate back to the contrast setting, well, you can see the problem. Or not see it, as the case may be.

As I suspected, I was able to find an angle where the text on the screen was just barely visible.

However, navigating to the contrast settings and changing the value, while it did affect the contrast somewhat, did not fix the display. It just went from almost blank to entirely blank. I did some diagnosis and found that all the signaling, including the contrast voltage, were correct, so it was in fact a dead display.

It does have a nice bright backlight, though, so it's got that going for it. What doesn't have a nice bright backlight is my SE-50.

Switching the SE-50 on reveals a screen with a very sharp, high contrast display, but a pitifully dim backlight that's barely brighter than when it's completely turned off.

This of course means that I have one display with a good backlight, and one display with a good LCD panel. I think we both know where this is going.

While the newer display in the SE-70 has a clone chip, the one from the SE-50 has an original Hitachi HD44780 LCD controller.

Which isn't really relevant here, I just thought it was neat.

Anyway, to swap the backlight we first need to desolder the power leads.

It would be nice if the backlight would slide right out, but it's still stuck under the metal bezel. We'll need to twist the tabs on the backside and lift it off. Note that it also holds the LCD panel itself to the zebra strips, though after a while they tend to become fully adhered to both the LCD and the PCB, so there's not too much danger of knocking them loose. If you do, though, just put them back in place and they'll work just fine.

We'll need to do the same disassembly on the SE-70's display to get the good backlight out.

It was at this point I noticed that the bezel on the SE-70 display was also in better shape than the SE-50's bezel, particularly regarding the part where it wasn't covered with the decayed gunk of what was once some black foam. Instead it was covered with actual black foam. So I took both the backlight and the bezel from this unit to combine into the final working unit.

And speaking of combining things, the backlight looks pretty good behind the SE-50's display. And with the bezel on and connected to the SE-70, things are looking mighty fine indeed!

And it still works on the SE-50 too.

The SE-70's display is definitely toast. It's vaguely possible that only the polarizer is destroyed, but a whole new display is only about $3 which is less than a sheet of polarizer film, so I'm just going to go that route.

Anyway, since I haven't played around with the SE-70 yet, I decided to put the frankendisplay into it until the new displays show up.

And I must say it looks mighty fine.

As for the old display, I was thinking of putting it in the SE-50 just to fill the hole in the front of the unit.

But the shorter cable from the SE-70 doesn't reach between the connector near the back of the SE-50 and the front panel where the display mounts.

So it's just gonna look a bit empty for a little while.

A small price to pay, just a temporary inconvenience.

Rotato

Today it's the MC-303 back on the bench.

After I had initially repaired it, I noticed that the rotary encoder was a little skippy. It would sometimes miss steps, go backwards, etc, which indicated that it likely had some contact issues inside it. So, apart it comes.

We'll need to remove this happy little camper from the board.

And luckily this unit is old enough to have delicious leaded solder, which slurps up so nicely.

To properly clean the contacts, we'll need to open it up. They're not exactly designed to be opened, but it's relatively safe to do once or twice.

With the limiting factor being how many times you can bend these tabs back and forth before they fatigue and break off.

But given it's lasted this long before failing, it'll probably be a good while longer before it fails again. and needs another disassembly.

Inside we have three contacts and a spoked metal disc. One contact rides around the solid part of the disc making constant contact, while the two others ride around the inner and outer spokes respectively, making contact 90 degrees out of phase with each other.

At least that's what they're supposed to do when nothing's getting in their way.

Hmm, what's this? Zoom, enhance!

Well that's no good, there's a little flake of plastic sitting on the back of one of the contacts. If that got between the contacts and the disc it wouldn't function very well. I wonder where it came from?

Yes, I wonder indeed. It looks like a thin ring of plastic was overmolded onto the outer circumference of the spoked disc, and it has over the years cracked and broken away. Well, that explains things, better clean out the shrapnel.

That's looking much better. Now it's just a matter of slathering up the shaft with some friction grease and getting things back together.

And together it goes.

As does the rest of the machine.

And I'm happy to report that the encoder now works perfectly.

I do have one other task on the plate for this unit...

But it will have to wait for another day.

If It's Worth Doing Once

So the time has come to crack open the MPC1000 again, hopefully for the last time.

The replacement pad sensor sheet came in to replace the one with the unfortunate manufacturing flaw. So, let's dig in.

The top case comes off in the usual manner, and the pads just pop right off, revealing the sensor sheet support bracket underneath. With 4 more screws it can be flipped over, and we'll release the flat flex and then unscrew the adapter PCB to free the bracket (and its attached sheet) from the unit.

A little preview of the before and after, the new pad sheet doesn't have the unfortunate manufacturing defect cutting through the trace for pad 4.

But before I put the new one on I have to peel the old one off, and it's properly stuck down. Slow peeling and cutting the adhesive with a hobby knife is the way to go here.

And after about 20 minutes of that it's off.

Or mostly off, at least, there's still plenty of adhesive residue left that we really need to get rid of so that the new sensor sheet lays flat against the bracket.

To deal with this I opted to soak the bracket in d-limonene, and the cheapest source of that I could find was Zep citrus degreaser concentrate.

I managed to only spill a little bit on my toes, so I now have very clean toes and an unusually clean spot on the floor of my car hole. Could be worse, I guess.

I let it soak for about 40 minutes, then it was time to go at it with a scotch brite pad.

The fact that the adhesive was now orange seemed like a good sign.

And it was a good sign indeed. Just a little bit of scrubbing with the pad and some water peeled the adhesive right off.

So that was a lot less work than the first time I tried.

After a rinse and dry the bracket is clean and very grease-free, so it's time to reassemble.

Peeling back half of the backing paper gives you something non-sticky to hold onto while you carefully position the sheet.

And once it's lined up, you just pull out the rest of the paper and smush it down.

Then plug it in.

Which is a bit tricky since at least this particular revision of the board doesn't have the type of connector with a release mechanism, so you need to use some force to jam it in.

Anyway, reassembly happens.

And we use the pad calibration screen in JJOS to check our work.

And it looks like luck is on my side this time, all of the pads work!

I guess now I just have to figure out how to use this thing.