0
\$\begingroup\$

In my project, I would be using NMOS' and it would be very handy for me to have a bulk terminal. Since my choices of NMOS parts don't have bulk terminal, I was wondering if I could somehow decapsulate it (such as the tested method of dipping the epoxy package in nitric acid, etc) and attach a terminal to the bulk.

This really isn't an IC and I'm guessing that the bulk itself is quite thick (having been rated 24V). So, what I'm hoping is that there are no wires that are too thin and fragile that would break. Would this work?

How about ordering a bunch (to take care of overhead) from manufacturers that has not been potted. All they have to do is take it from the assembly line without much problem, right?

\$\endgroup\$
14
  • 1
    \$\begingroup\$ Why would it be very handy to have a bulk terminal? (Part of my answer would be: Exposing the die of a transistor can pose several problems (light sensitivity) and I think I've seen some sort of bonding wires even on high power IGBT-Modules (rated for >1000V and 100A). So probably not going to work.) \$\endgroup\$ – Arsenal Aug 25 '16 at 14:11
  • 1
    \$\begingroup\$ How about just ordering them before they are encapsulated? \$\endgroup\$ – Majin_Boo Aug 25 '16 at 14:15
  • 2
    \$\begingroup\$ Assuming that you do manage to 'de-capsulate' one without breaking it and that light-sensitivity is either not a problem or worked around, how would you actually "attach a terminal to the bulk"? Its not like you can just solder a wire on ... \$\endgroup\$ – brhans Aug 25 '16 at 14:22
  • 2
    \$\begingroup\$ @Arsenal SemiDice is relatively reasonable for parts, but even they have a decent minimum order. \$\endgroup\$ – W5VO Aug 25 '16 at 14:25
  • 4
    \$\begingroup\$ I'd really like to stress what @Arsenal's first comment said: This looks like the infamous X/Y problem: You want to do something specific, and came up with a method. However, everyone else thinks your method is really strange, but does their best to help you. However, in the end, you can't be helped, because your approach was a dead end and you forgot to explain what you wanted to do in the first place. That way, you end up wasting your energy :( so please, what do you want to do, overall? \$\endgroup\$ – Marcus Müller Aug 25 '16 at 15:28
2
\$\begingroup\$

This probably isn't going to work on many devices, as in most devices you already have access to all the terminals. Certainly on power MOSFETs, which are vertical devices, there is no bulk terminal. The substrate/bulk contact, if it exists is commonly tied to source, as otherwise you would end up with weird behavior when using the transistor as intended. You'll find most devices already make electrical contact to the back of the die for the drain.

Note that I'm not even addressing the decapsulation aspect because you couldn't make the connection you want starting from a bare die part.

I have seen some devices where the bulk terminal is not implicitly/explicitly shorted and exposed to a pin. Advanced Linear Devices makes parts like this, such as the ALD1107. However, it is very uncommon to find parts where the bulk terminal is not internally connected to the source.

\$\endgroup\$
1
  • \$\begingroup\$ Looks like there's a part of your answer missing! \$\endgroup\$ – AndrejaKo Aug 25 '16 at 15:46

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.