Casio DBX-100 repair attempt

This bad boy from either 1986 or 1988 (sources differ) is one of a handful of Casios with a full dot matrix display, and has a metal case with a mineral crystal. Such a shame Casio doesn’t make them like this anymore… An EL backlight on one of these would be the bomb. Unfortunately this doesn’t even have an incandescent let alone a LED… Actually, I wonder why digital watches with incandescent illumination were ever a thing, since the first digital watches used tiny LEDs…

Anyway, I’ve had one of these sitting forgotten in a drawer for years, until recently… Unfortunately, Casio didn’t make the buttons out of anything durable, so they’ve all either disintegrated by now, or the watches with remaining buttons are unused to keep them that way. Here’s my example:

I bought a dead example of a later (and otherwise inferior) DBC-62 for its buttons, and managed to remove them and trim the surround to fit (a bit disappointed with that effort, but I can probably hide it), but the damn contacts are at a different pitch, such that the outer two would completely miss. It also seems I’ve damaged the traces during removal anyway, since they seemed to have lost continuity from a visual inspection… I had some old conductive paint, which is the only way I can think of to repair and reposition the traces, but it’s fiddly business…

The contact strip is wider and comes off the button pad at a higher position too, so I’ve had to enlarge the slot in the case. Looks like it can still fold back to the proper spot inside for the zebra strip to push on it though.

Unfortunately this silver-based paint doesn’t seem up to it; attempting to clean up the traces led to most of it flaking off, in spite of the boasts on the packet, and at this thickness, what’s left doesn’t seem to conduct worth a damn. To be fair, it’s several years past its shelf life, but I’m not sure why silver should stop being silver. It’s pretty gluggy, but I mixed it as well as I could inside the pen with a poker, which I used to apply the stuff, since I had terrible results with the pen tip back in the day…

Anyone have any ideas how I can make this work? I actually found a NOS DBX-100 on eBay, but the buttons are toast and the batteries (there are two) have probably leaked inside it like with the DBC-62 I bought.

I bought it anyway, since it was a reasonable price, and at worst I can put my working module inside it, scrape off the perished rubber and replace it with an exorbitantly priced decal somebody in the UK makes. But maybe the NOS module will work with fresh batteries, and I can get two of these units running…

1 Like

Hi Kimmo.

Oh sick one!

What is the dimensions of the missing keys?
Maybe a super small very low voltage e-paper resistive touch screen?
What is the chip inside the watch?
What kind of things could it drive?

1 Like

Not at home right now, but IIRC the 4x4 grid is something like 30x18mm…

Adding something which requires power and its own control circuitry seems way too difficult; I’d be flat out just figuring out the circuit between the buttons and the connector; there are two sheets with traces running all over, sort of spot-welded where they electrically connect. I’ll see if I can get a decent pic of it later.

I’m hoping somebody here can offer some suggestions as to how to reverse-engineer or at least repair this plastic film based circuitry…

The chip inside the watch is your typical digital watch situation; something proprietary that’s just buried under a blob of epoxy.

1 Like

Here’s the circuit for the contacts, with the little ribbon which separated from it (the connector ribbon is part of one of the layers on the newer button pad). There’s another layer, not pictured, which just has sixteen ovals which bridge the contacts when pressed.

Wow! We’re getting super deep into the rabbit hole now :smiley:

I’m not sure if it’s helpful at all, but if your flex PCB is no longer viable and you need a new one… you could design one in eg. KiCad as a flex PCB and have it manufactured by JLCPCB, PCBWay etc who all do flex PCB manufacturing.

Mad respect for the restoration attempt. Best of luck :call_me_hand: and please keep us updated

2 Likes

This was some super fiddly soldering; that wire is the tiny 80 conductor IDE cable. The idea was to have the wires hanging out of the hole on the front, but I couldn’t assemble it like this. Luckily it runs with only one battery, so I soldered them onto the other side of the board, hanging out of where the bottom battery goes.

Pretty sure I’ve figured out the circuit for the buttons; just have to put together a matrix of sixteen buttons to test it.

Once I’ve got that verified, I’ll go ahead and get hold of one of those PCB design programs and come up with something…

The original circuit is two layers with a ribbon connector stuck on, which seems like it might pose a problem… Maybe I’ll have to re-jig it so the connector is part of one of the layers, which will be tricky. And there are a bunch of vias joining the two layers; can the PCB fab guys manage that?

1 Like

Hm. I seem to have hit a wall.

The buttons all work; I’m able to get each of them to do something. Problem is, when I went to figure out how to set the damn thing, something weird happened.

The top left button on the pad, according to the text on the crystal, is also ADJ… when I press it on the main screen, the time is replaced by THIS WEEK, another function depicted on the crystal. So far, so good; presumably I just have to long-press it.

But when I do that, THIS WEEK just remains on the screen, and the watch is frozen. The only way to get anything to change from there, is to pop the battery. Bad sign. Even more disturbing: I couldn’t figure out what the two upper side buttons were supposed to do, which seemed a bit silly given some of the much fiddlier front buttons have four or even more functions each… Until I noticed the seconds pause while I held them, and also the same went for the lower right side button which switches between 12/24h. And when released, the time didn’t jump forwards to catch up; it remained delayed.

Uh-oh. That’s very bad.

I am surprised the thing could work so much and yet be cactus. The calculator works but I can’t set the time. Guess I could power it up at the stroke of midnight on new year’s eve…

Dammit!

Now I really hope the NOS one I’ve bought hasn’t been ruined by leaking batteries.

1 Like

Okay, some good news - I was reluctant to short the AC pad because I wasn’t sure what to short it to, but after watching a couple of repair vids I gather the done thing is to short it to battery negative. After doing that, the freezing and pausing behaviour is gone, thankfully…

Although I still can’t figure out how to adjust it or what the upper two side buttons are for. The NOS watch comes with a manual though, which will no doubt be very helpful.

This is interesting… Six jumpers on the left of the PCB; three of them bridged. Wonder what they do… Language, maybe?

Look at all those test points!

1 Like

Hi Kimmo

Probably for “bed of nails” testing which was an automated process popular years ago for large production runs. The “nails” being spring loaded pogo pins with points which contacted each test point. A dedicated computer would run tests between selected points or maybe between every pin and every other pin with Pass / Fail parameters. Similar to “signature analysis”. Both these testing procedures carried out with the DUT un powered.
Cheers Bob

1 Like

Figured as much about the test points… This PCB/IC is the most complex I’ve seen in a watch. QC would be more necessary I bet.

Very intrigued about the jumpers… I wish I could find someone who knows a lot specifically about this stuff.

1 Like

Forgot that I was able to download a PDF of it. Turns out there’s nothing wrong with my 261 module :smiley:

4 Likes

The NOS one rocked up!

And it works! Old batteries with only 0.01V hadn’t leaked!

I read somewhere the screen bleed is due to poor sealing letting air in; I have a vacuum pump, so I’ll see if it’s possible to fix this. In the meantime, my old screen only has a tiny bit of bleed in a couple of corners.

As for the buttons, I’d probably need to get a custom flex PCB made to fix the old one, but this one just needs a pad from a later model; there are some NOS ones around… I’ll probably just leave this one on until it’s completely toast, which might be a while, since I don’t need to use my nails; the tip of my index finger is pretty firm and has a fairly small radius…

2 Likes

Great transplant @Kimmo :smiley: what a relief the battery didn’t toast the whole thing.
Vintage gear is so nice… I miss the days when everything seemed to have a custom VFD or LCD :drooling_face:

1 Like

Tell you what, it was fiddly AF. The screws are all tiny little slot jobs, which is a nightmare come reassembly. Philips are so much easier to deal with at this scale… And the extra tiny spring which contacts the cut-down piezo behind the keypad was a total bastard. Took me more than five minutes to get it into its hole!

3 Likes