[SI-LIST] Re: Temperature vs. performance in different type of processes.
- From: "Volk, Andrew M" <andrew.m.volk@xxxxxxxxx>
- To: si-list@xxxxxxxxxxxxx
- Date: Fri, 13 Jul 2001 17:09:58 -0700
In CMOS, the carrier mobility decreases with increasing temperature, weakens
the transistor and slows the speed. At the same time, the transistor
threshold is decreasing with increasing temperature. That will increase
gate drive to the transistor for a given supply and will speed up the logic
slightly. For logic which is seeing full supply swings, the mobility effect
will dominate by a significant margin. If the swings to the circuit are
small and close to the threshold voltage, then you can get a case where the
threshold voltage effect dominates and the worst case will be low
temperature.
Andrew Volk
Intel Corp.
-----Original Message-----
From: Ingraham, Andrew [mailto:Andrew.Ingraham@xxxxxxxxxx]
Sent: Friday, July 13, 2001 4:40 PM
To: 'si-list@xxxxxxxxxxxxx'
Subject: [SI-LIST] Re: Temperature vs. performance in different type of
processes.
> "CMOS,Bipolar and BI-CMOS processes each have unique characteristics that
> determine these
> performance conditions." said in the article.
>
> Can somebody shed some light on it? Why do they have different temperature
> defined for maximum and minimum conditions?
>
It is a function of the physics of bipolar vs. MOS transistors.
For bipolars, recall that the current through a semiconductor junction at a
given voltage increases as the temperature increases. That's just how they
work. Bipolar transistors are "stronger" (conduct more) and generally
switch faster when they get hot. (This was what made linear bipolar amps
susceptible to "thermal runaway.")
I don't recall off-hand the physics that determine temperature dependency
for MOSFETs, but the end result is that they switch faster at lower
temperatures. Cryogenic cooling is used on some faster CMOS computers. You
can even buy "souped up" computers with a built-in refrigeration unit
attached to the CPU.
> What is for BI-CMOS?
That depends on the individual Bi-CMOS circuit. It could go either way.
Heck, I suppose it could be nonlinear with the fastest (or slowest) point in
the middle of the temperature range. (There's no one way to design a
circuit and call it "Bi-CMOS.")
Regards,
Andy
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
