I took the plunge and performed “open diode surgery” on my laser module.
I was able to scrape out all the epoxy by solely mechanical means. All you need is a sharp tool (I used a pair of fine point Hakko tweezers for most of it), and a lot of care and patience. I worked out tiny chips at a time. I found the epoxy around the wires formed a natural protective barrier that helped me avoid chaffing into them. Once all the epoxy was removed, I found pink foam stuffed in which I assume was to prevent the epoxy leaking inside, and maybe also to act as a heat insulating barrier.
With the epoxy and foam removed, I could slide the logic board most of the way out. But the end connected to the interior module got snagged at first, preventing me from removing it completely. I was going to cut away the end of the cylinder to try and get better access (you can see the beginning of the cut), but that turned out to be unnecessary. All you need to do is heat the outer metal cylinder a bit. A few seconds with my cutting tool provided sufficient heat for it to expand enough that the contents dropped out easily. I’m sure a heat gun or hairdryer would have done the trick as well (it didn’t take much expansion).
Here are some closeups of the extracted module:
The board was wrapped in heat shrink tubing which was easily cut away.
I took some measurements, and my circuit schematic a few posts back seems correct, except I missed a connection between the LDA pins. Here’s an updated schematic (thanks @Marshall_Banana for the datasheet on the Lasorb IC):
R1 was measured (in-circuit) as 120Ω, R2 as 985kΩ. A continuity test with probes on each end of D1 measured 542mV (black probe on black wire, red probe on blue wire) and nothing in the reverse direction. U1 measured as 653mV with the black probe at LDA and red at GND, and 356mV in the reverse direction. Although I believe the yellow wire is for chassis ground, there was no continuity between it and the brass barrel of the module.
Before any further disassembly, I gave the module a final test with a bench supply to make sure nothing got damaged so far. It still worked.
Then I desoldered the old laser diode from the board:
One of the photodiode wires snapped off during handling. It was easily repaired.
There are two lenses in the module. One in front of the laser diode, and one at the emitting end. You can just see the bottom edge of the former in the photo below (by peering through the gap between the photodiode and the back edge of the cutout). I think both are stuck with adhesive (glue? epoxy?). There look to be threads on the inner surface of the barrel. There’s what looks like a beamsplitter behind the emitter lens, which I presume sends some of the light to a photodiode. There’s a blue sticker or coating on top of the beamsplitter.
I think the original laser diode may have had a photodiode built into it, but if so it wasn’t being used (I think the third lead was cut).
Unfortunately I hit an obstacle trying to get the diode itself out from the back of the module. It looks like it was epoxied in place around the edges, and I suspect the fit is fairly tight. I tried heat, but was wary of overdoing it (at the time I thought the lenses were acrylic, which in retrospect was stupid, they must be glass).
Since I didn’t care about recovering the old diode, I decided to drill it out. By this point my patience was wearing, and I knew I probably ought to have set it aside to continue another day.
I was careful to make sure my drill bit didn’t enter deep enough to hit the lends behind the diode, but I guess my bit wasn’t sharp enough as the drilling force caused the diode itself to deform and crack the lens behind it. I got the diode out, but with the damaged lens the module is useless.
Until I can find a replacement module (or new lens?) that’s the end of DizzyRuff :-(.
I’m open to suggestions for resurrecting this printer if anyone has any! (Otherwise, I guess I’ll be doing a firesale on Form 1+ parts and still-shrink-wrapped resin tanks).
