Beep Beep I'm a JP

A while ago I took on the ownership of a rather ill Roland JP-8000 keyboard. I fixed it up with a bit of PCB surgery and got it running, but at the time I didn't replace the tact switches or clean the sliders. I was just concentrating on seeing if I could make it work and I didn't want to spend money on extra parts if it was going to be a lost cause.

Well, as we know, it was not a lost cause and the synth works great, but it still had some dodgy front panel switches and some scratchy sliders and potentiometers.

A few months ago I went at the sliders from the front, squirting in some contact cleaner and working them around. This cleared up the scratchiness and intermittent connectivity on all but the volume pot (which we'll get to later in this post), but the contact cleaner left some pale coloured stains on the plastic of the case that really frustrated me, so I put it aside and didn't bother replacing the switches right away.

Well, fast forward to earlier this week and I decided to take a closer look at those stains. They hadn't just magically disappeared as I hoped they might after the solvent evaporated, but mysteriously they did wipe away when I scrubbed at them with a microfiber towel soaked with Simple Green. I honestly don't know what the chemistry behind it is, but I'm glad it worked!

That success got me excited for this synth again, and gave me the motivation to replace the tact switches and get it over the finish line into being fully functional, so onto the bench it goes.

And while I'm in there, I'm definitely going to be dealing with the dust inside the display.

So off comes the back, since this synth is a bottom-loader.

To get to the front panel PCBs, the keybed needs to come out. You might think this would be a lot of fuss and bother but actually it just comes out with 5 screws.

And then a few more screws gets the main PCB and audio jack board out of the way.

A few more screws and yup, there's all that dust that was hiding inside the display.

Anyway, out comes the right-side front panel PCB, since it's a bit easier to remove than the other two.

Now I should say that my sense of frustration was only half of why I hadn't done this job already. The other half is that I kept buying just enough tact switches to repair this synth and then using them to repair other synths that use the same tact switches. So I finally did the smart thing and bought Enough Tact Switches.

A little desoldering and resoldering later and the first panel is done.

The middle and left panel PCBs are a bit trickier and it tends to work easiest if you remove both of them at once, so out they come.

And one is rebuttoned.

This is also the panel board with the main volume control pot. The rest of the control potentiometers have a little pinhole on the top surface that you can squirt contact cleaner into, but for whatever reason this stereo pot only has holes on the sides.

So I gave it some squirts now that I have access to it, and measuring the resistance as I rotated it showed that all the jumpiness had disappeared.

So that just leaves the left-side panel.

It went exactly like the others, and we can scoop up all the old switches to get a good look at the pile of e-waste.

I'm sure there's a whole host of sea creatures that are lining up to chew on these tasty morsels of soon to be ocean microplastics.

Speaking of things that are plastic becoming smaller things, a few spots on the control panel PCBs had some foam tape that was used to keep the component leads from, over time, sawing their way through some overlaid cables. As luck would have it, every single piece of tape managed to cover the leads of at least one tact switch, so I had to pick off the sticky foam, which had somehow turned into something resembling astronaut ice cream at some point over the past 30 years.

But that left the PCB exposed, with its razor sharp cable-shredding leads just itching to shred some leads. Luckily I happened to have some fabric wire harness tape handy, so I used that to armor things up.

So everything goes back inside, and we can flip it over to see how much cleaner the display... oh for the love of-

Ok, take two, flip it over and...

Yes, that's much better.

So with everything properly back together, the synth still works great, all the buttons now work, and the volume pot is as smooth as butter.

And now the only thing left to do is play it.

Plastic Squirts

So I picked up a "3D Printing Pen" this week.

It's basically a handheld version of a 3D printer extruder, and in this case powered via USB C. You shove 1.75mm filament up its poop chute and it vomits out a semiliquid version of it from the other end when you press the button. This then quickly hardens and, by manipulating the pen, you can shape this plastic into a variety of different forms.

I opted for the fancy version which incorporates a display on the side. A hilariously small display.

I wanted to make sure I had some decent control of the temperature, and experimenting with that without a display didn't seem like the most productive use of my time.

Speaking of unproductive uses of my time, I decided to give it a test run by "printing" a plastic YouTube play button, mostly because they included a play button shape printed on a pack-in card with the unit and it was easy to trace over.

This did mean that the top surface of the paper came up with the plastic, since apparently ASA sticks to coated paper fairly well I guess.

But the exercise did validate that the pen works, and I'm at least vaguely capable of manipulating it.

While most people use these for creative hobby projects, my goal for this is much more utilitarian: I'm planning on using it for rebuilding portions of broken plastic in my various synth repair adventures. Since this broken plastic is most commonly ABS, getting a pen that could melt ABS (or ASA in this case) reliably was an important factor, as many cheaper models only melt PLA. This pen seems to do that job fairly well, so I'm looking forward to putting it to work.

Prune Me Like Your French Lavender

But not like the Spanish Lavender, that would be different.

Following up on the rosemary pruning this week, I decided to give the french lavender a haircut since the flower stalks were basically done.

And you'd be able to see that in this photo if my camera hadn't decided to focus on the pea gravel in the background instead of the flower stalks in the foreground.

Well anyway, if you don't prune lavender it can get super leggy and weird looking, so let's give it a trim.

I actually pulled a few branches out of this one earlier that were dying back, so it's looking a bit peculiar. The other one is much more photogenic.

And much smaller than it was previously.

Go Into the Light

I picked up a small EL panel for a project that is coming up soon-ish, and it occurred to me that before I go ahead and install it, I should make sure that it works, and also figure out what side the light comes out of.

This sounds like something that should be an easy task, but it's actually a bit tricky because EL panels generally take a pretty high voltage and high frequency to actually turn on. Luckily I have a well equipped workshop.

So first things first: yes, the panel does work, and it emits light out of the white side, not the other side that looks a lot more like what light should come out of. Go figure.

To get things to glow (somewhat, at least), I used a multi-tapped 24v HVAC transformer, and fed a 400Hz sine wave into the low voltage side. In theory if I put 24v into it, I should get 240v out of the 240v tap.

Well, the first problem there is that my function generator can only put out 10v.

And the 10v is peak, not RMS.

And then on top of all that, most of the time these cheap transformers are over-wound so that the low voltage side actually puts out a fair bit more voltage than it's actually rated for, since the voltage tends to sag under load.

All this combined to make the output voltage only around 55v RMS, but it was apparently enough to still dimly illuminate the EL panel, so I guess in the end it did its job.

Roses are Mary

The rosemary along the front of my house tends to be somewhat vigorous in its growth. Previously I had been going in with my pruning shears and cutting things back, carefully making sure to prune the long branches back deep inside the shrub so that I wouldn't create a thin, flat wall of foliage on the outer edge that shades out the interior.

Then I got a hedge trimmer and did the lazy thing once or twice, and boy howdy did the rosemary grow a thin, flat wall of foliage on the outer edge that shades out the interior.

It doesn't really show in this picture but a lot of the foliage is not as vibrant as it should be, and when I took this picture I had already pulled a pretty large dead branch from the bottom of the first bush.

Well, I decided to keep going and do a full thinning to allow more light and air to get into the middle of the rosemary so it wouldn't just turn into a green skin overtop of a bunch of dead brown underneath.

This looks a lot more like what rosemary should look like. The closest one is a bit naked on the bottom, but it had a lot of dead branches down low, so it'll take a bit to fill back in there.

Curiously, despite doing a pretty hard pruning, it somehow manages to almost look bigger than when I started. I can guarantee that I removed quite a bit of material, though.

Will I go back to pruning with the shears from now on? No, obviously not. In fact I still used the hedge clippers to shape things back once I had thinned out the growth with the shears.

Speaking of things that are growing, the sunpatiens seem to be reasonably happy with their home.

They didn't really do a whole lot at first, but over the past month or so they've been really puffing out with new growth.

Feeling Tacky

Today on the bench is a familiar fellow, the Yamaha FB-01. When we last saw this happy little black box it was suffering from a shorted capacitato, which I fixed on video.

At the time I noticed that the tact switches were a bit iffy, but the ones used in the FB-01 are a slightly peculiar 2-pin rectangular switch that I didn't have replacements for. I ordered a pack of them, but they took a week or two to show up, so I set the box aside and didn't really think about it.

Until now, that is.

So let's dig in.

I think the front panel might be removable without pulling off the top case, but it's only 6 extra screws so not really a big deal.

Disassembling the front panel, we find my favourite thing: a broken screw standoff.

Oh goodie. Well let's glue that together with some superglue so it can fail to bond even slightly.

I'm sure it'll be fine.

[Narrator: it was not fine]

Moving on, the button panel PCB comes out quite easily.

And on the flip side we can see that they didn't exactly spend a lot of time laying out these traces.

If it works, it works I guess.

Desoldering went smoothly.

As did soldering in the new switches.

And for the grand reveal, did the super glue stick?

Of course not.

But no worries, I'll just apply some more and then screw everything back together before I notice that it didn't stick again.

I'm sure it's probably fine.

Blue Gravel

So recently a local hobo decided that they were bored of stealing copper and would try branching out into stealing tools instead.

Well jokes on them, I don't keep any tools in my van. As such, nothing was stolen, though a few other vans in the neighbourhood were also broken into and probably suffered some losses.

Anyway, the good news is that a new window is under $100 on Amazon, because of course it is, so let's get to fixing this mess.

For the sake of the algorithm, this beast is a 2007 Dodge Sprinter 2500 based on the Mercedes NCV3 platform.

With that out of the way, we need to get the inside trim panels off, and that starts with removing the lower panel. It's held in by the two quarter-turn knobs on either side, and then just hinges down and lifts out. Don't forget to unplug the light at the bottom.

I think this is meant to be kind of a storage cubby thing? I dunno.

Next up the trim panel on the door pull needs to come off to reveal a pair of bolts underneath.

It's got some pretty stout clips at the bottom that need to be pried open, then the rest just pops off.

Or maybe it's supposed to come off from the top first? I dunno, I'm not a mechanic.

Removing the trim around the door lock is technically optional but helps free up the door a little easier.

There's two screw covers at the top corners of the door panel that need to be popped out to reveal the final pair of screws, then after unscrewing all six the door panel just pops off with an assortment of the usual at-least-three-break-every-time-you-remove-it plastic snap fittings.

The bowden cable for the door latch snaps out of the handle by prying the outer jacket fitting out first, then unsnapping the clip around the L-shaped end, which I apparently failed to take a picture of.

It's easiest to unplug the window controls from the wiring harness at this point, but I'm going to need them a bit later so removing the white plastic clip allows the controls to be removed from the door panel.

That way it can be plugged back in and left to dangle.

Now this is the part where things diverge a little from the usual method one might employ on most cars. Since the van's windows are significantly taller than they are wide, it's not possible to slide them in from the top by angling the leading edge down. Instead, Mercedes designed a large cutaway on the inner door panel that the window goes through from the inside.

But first the window regulator needs to move out of the way. It's held in by two bolts and two nuts, one of which is loose enough to explain why my door kept rattling so much every time I opened and closed it.

Good thing the nut didn't back itself all the way out, I guess.

Anyway, with the 4 bolts removed and one wiring harness clip popped out, the window regulator can be removed and set aside.

And now there's clear access to shove the glass inside the door.

Speaking of the glass, let's get that unwrapped.

I am genuinely impressed that it made it here in one piece, all the way from China.

Miracles do happen, I guess.

Anyway, the glass goes in bottom-first... I think. I know I tried both top first and bottom first and I'm pretty sure top first didn't work, so if going in bottom-first doesn't work for you then try the other way.

Then to get the window regulator back in you need to slide it at least partly out of the way. The simplest thing to do is to just shove it right up into the track and tape it in place at the top.

My window track was sticky enough that it probably could have held itself up by friction, but also I didn't want it to suddenly come crashing down inside the door while I was working and shatter into a fresh new pile of blue gravel.

On the subject of working, I need to work the remaining bits of old window out of the regulator's clamps.

And then once it's bolted back into place, tightly this time, I can lower the window down into the clamps.

I snugged it up just finger tight at first and ran the window up and down a few times to make sure it was properly aligned. Remember that part earlier where I hooked the window controls back up and left them dangling? This is why.

Once the window was seated in the right spot in the clamps, they can be snugged up. Don't overdo it here, it just needs to grip the glass firmly.

Anyway, now I just need to peel off that sticker from the glass and... oh, I probably should have done that before installing the window.

Lesson learned.

Moving on, the inner door trim needs to go back on, so I'll start with refilling the plastic snap fittings because of course some of them broke.

And now there's three different colours of them in this one panel.

The reinstallation is just removal in reverse, and with everything back together things are looking good.

And with any luck I won't have to do this again soon.