Teardown and failure investigation of optocouplers.

[ResR]

Well I came across a very interesting failure, I found 4 used 1979 vintage optoisolators ILD-74 (and a lot of 5V logic chips, mostly SN74LS series), I unmounted it from their sockets inside some kind of apparatus, that didn't had any covers left, I was planning to use those chips in a project, but those dual optocouplers had one half of the optoisolation units faulty and other half not. So I took a knife to scrape the share line onto the chip and used heavy duty wire cutters (the toolbox ones) to literally cut the chip to half, the working half facing down onto the bed (one half of one chip took off and never landed). Been careful to separate the working half of the optoisolators , I took the faulty part and cut it half in a center to reveal two semiconductors and a rubbery substance between them. I accidentally ripped the bonding wires off from the first time, so I cut other way second time just to remove the top 1/3 of the chip from the another faulty half f the chip to reveal the semiconductors with the rubbery substance intact. I measured the both sides and the phototransistor seemed working, also the IR-led didn't had any light output visible for the phone camera.

[SeanB]

Those old IR emitters did have a failure mode where they lost emission. The temporary solution ( often worked for a long time) was to up the current through them. This was a big issue with those used in older HP laser printers where they would get intermittent paper feed issues that eventually showed as a jam with no paper present. Solution was to solder a 470R resistor across the existing resistor as you otherwise had to take the unit apart to get the board out. Would cure it for a long time, and when it occurred again you needed to strip it to replace the switch.