2
\$\begingroup\$

I am experimenting with optically controlling BJTs. In my current project, I take transistors in metal cans and use a diamond saw lubricated with isopropyl alcohol to slice the top off. This generates very little debris and presumably doesn't scratch the silicon die inside.

After the cutting, I hooked up the transistor and was able to use light to control the transistor, and it seemed to operate the same as any ordinary phototransistor.

So here's my question: Will a transistor opened in this manner degrade over time? Are there problems associated with exposing it to air or dust? Or is it passivated or otherwise protected in a robust way?

\$\endgroup\$
  • 3
    \$\begingroup\$ Just curious: I can see doing this once as an educational experiment, but if you're concerned about reliability etc. why not just get a phototransistor? They're not expensive and they are readily available. \$\endgroup\$ – John D Sep 15 '17 at 14:08
  • 1
    \$\begingroup\$ I know that sawn wafers degrade over time (a couple of years is still ok, I believe). This degradation is prevented once the dies are packaged. Now, if you remove the packing, chances are that you're left with something as degradable as the original sawn wafer dies. \$\endgroup\$ – dim lost faith in SE Sep 15 '17 at 14:16
  • \$\begingroup\$ @dim Do you know the mechanism of the degradation? \$\endgroup\$ – Colin Marcus Sep 15 '17 at 17:00
  • \$\begingroup\$ @colin no, and that's why I only made a comment, not an answer. I know that because my company had a big stock of sawn dies of obsolete components and they were struggling to find a way to sell them before they were unusable. \$\endgroup\$ – dim lost faith in SE Sep 15 '17 at 18:00
  • \$\begingroup\$ Very likely they degrade because of moisture. You can "preserve" sawn dies by "baking" them I believe (that removes the moisture) and then storing them in such a way that moisture cannot penetrate into the dies. I know that there are companies that can take care of this for you, at a cost of course. So it's not trivial to store dies (be they sawn or not) but it can be done. \$\endgroup\$ – Bimpelrekkie Sep 15 '17 at 20:51
1
\$\begingroup\$

Will a transistor opened in this manner degrade over time?

As far as I know and from personal experience: no.

There is always a "passivation" layer on top of any semiconductor device to prevent moisture from entering the PN junction. I'm not talking about moisture as when the device gets wet but moisture that is always in the air.

Without that layer the devices would be vulnerable immediately after production. Often packaging is done at a different location so the devices need some basic protection. Also not all packaging (metal/plastic/ceramic) will provide enough of a barrier to prevent moisture from entering.

I have opened the housing of a 2N3055 transistor (it is big one) years ago and it is still working fine.

\$\endgroup\$
  • \$\begingroup\$ hmm... a sample of one is not a good measure. If it's for your hobby project it may be fine, but I would never event think about doing so in a product. Also, package failure, according to some sources, is the #1 aging failure. \$\endgroup\$ – Trevor_G Sep 15 '17 at 14:03
  • 3
    \$\begingroup\$ @Trevor Agreed, this is for experimentation. If you want to use an opened device in a (mass produced) product then I would say: don't, buy a proper photo transistor or such. \$\endgroup\$ – Bimpelrekkie Sep 15 '17 at 14:07
  • 1
    \$\begingroup\$ @Bimpelrekkie I have an application that requires access to the base terminal and also particular parameters (either the 3904 or 2369), and I have found it difficult to find a suitable phototransistor. \$\endgroup\$ – Colin Marcus Sep 15 '17 at 17:03
  • \$\begingroup\$ Sightly different, but back in the day the OC71 used to come in glass packages that were painted black to prevent them acting as photo-transistors, there was also a (much more expensive) photo transistor version (The OCP71) and for a while it was possible to strip the paint to convert one to the other. They eventually changed to an opaque fill inside the tube, possibly at least in part to prevent this \$\endgroup\$ – Dan Mills Sep 17 '17 at 17:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.