Why do semiconductors have a minimum temperature rating? / General Chatting / Forums

Forums

4hv.org :: Forums :: General Chatting

« Previous topic | Next topic »

Why do semiconductors have a minimum temperature rating?

1 2 next

Move Thread

LAN_403

Inducktion

Thu Dec 08 2011, 09:06PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

I understand that semiconductors have a max temp rating for obvious reasons, but what causes a transistor/mosfet/igbt to fail at extreme low temps?

Dr. Drone

Fri Dec 09 2011, 03:01AM

Registered Member #290 Joined: Mon Mar 06 2006, 08:24PM
Location:
Posts: 1673

magnet18

Fri Dec 09 2011, 05:20AM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

So, the commercial parts were exactly the same as the military and space, but just weren't tested??

Goodchild

Fri Dec 09 2011, 06:19AM

Registered Member #2292 Joined: Fri Aug 14 2009, 05:33PM
Location: The Wild West AKA Arizona
Posts: 795

I knew it!!!! Radio Shack (or Radio Shit as I like to call them) dose get the bottom of the barrel

"Hey some white smoke came out of this one, but it's still working a little, what should we do with it?"

"Oh put that one in the Radio Shack bin"

Although minimum temp is an interesting topic, maybe it makes the part fail in the same way as heat? Heat makes the die expand and fracture leading to bad things. Maybe cold can do the same, as the die cools it contracts and eventually causes it to crack and fracture.

Just how I see it in my mind at least. I hope some can shed some real truth on the subject, because I would also like to know.

Eric

hsieh

Fri Dec 09 2011, 08:05AM

Registered Member #1412 Joined: Thu Mar 27 2008, 04:07PM
Location: Taipei Taiwan
Posts: 278

I found this

several months ago.

If I have time,I would pour some liquid nitrogen on some of my circuit and see what will happen.

Mattski

Fri Dec 09 2011, 09:40AM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

Low temperature could cause mechanical failure of the chip or packaging due to uneven thermal contraction of the various components: die, slug, wirebonds, plastic package, etc. or due to embrittlement of some materials.

Semiconductors also behave differently if the temperature is low enough that the dopant materials which turn plain silicon into n-type and p-type are not fully ionized, so the n-type regions are not n-type and the p-type regions are not p-type. The n-type dopants come with an extra electron but this doesn't actually enter the semiconductor without a little bit of energy due to heat. A similar thing happens with p-type dopants, but they "accept" an electron instead from the silicon, but a bit of thermal energy is needed so that the electron can jump from it's spot into that p-type dopant atom. If it gets so cold that this doesn't happen then the transistor will stop working entirely. I don't have any books near me that would tell me at what temperature this happens with common silicon dopants unfortunately.

Steve Conner

Fri Dec 09 2011, 09:59AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

According to this paper, some MOSFETs perform much better when dunked in liquid nitrogen.

But others "do not work at cryogenic temperatures".

The main problem, as mentioned, is that no manufacturer tests them at such extreme temperatures, so they don't know what'll happen. Murphy's Law of electronics says that any unspecified parameter will do exactly what it needs to in order to ruin your day. For example, work beautifully in prototypes only to change on the first batch and bankrupt you.

In production, you never rely on any parameter that isn't specified in the datasheet, unless you're buying enough parts to have the manufacturer select them specially for you.

Dr. Drone

Fri Dec 09 2011, 04:33PM

Registered Member #290 Joined: Mon Mar 06 2006, 08:24PM
Location:
Posts: 1673

AndrewM

Fri Dec 09 2011, 06:46PM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

magnet18 wrote ...

So, the commercial parts were exactly the same as the military and space, but just weren't tested??

Yes, in the space industry testing is king. We usually ballpark for any unit that 1/3 of the purchase price accounts for parts and materials, 1/3 for touch labor, and 1/3 is testing. At the component level that is surely skewed further in the direction of labor and testing.

If a part is hard to find and we're in a pinch we'll buy "regular" parts and up-screen them ourselves - its often the same stuff, the only difference being the degree to which you've validated that the parts and the lot overall meet spec and contain no latent defects.

EDIT: Let me add that in some cases there are substantive differences, but not always. For example tin and cadmium are verboten for space use - these material are popular platings on leads and connectors for some conventional parts.

Inducktion

Thu Dec 15 2011, 07:28PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

So basically it's just a reassurance when it comes to testing for military and space grade stuff?

To me, it sounds like it makes more sense to just have the military test them themselves, rather than pay 20 bucks or more for a military grade IRFP250.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob