No, Iss would only change if you had a point-contact diode, or some other odd geometry, and the depletion region varied; for a planar transistor, it's a device constant. Vt, of course, is proportional to Kelvin temperature.
Define 'close proximity'; on the same die? If things nearby are dissipating power, they make temperature gradients.
So, use a quad transistor, and use one for logarithm, one for thermometer, and two for heaters.
-
=A0 =A0 =A0 =A0 =A0
=A0 =A0 =A0 I_ec
g(to
=3D20C,
.com
I had your problem 10 years ago in a video processor. Solved it by using a MAT-04 quad transistor on a common substrate. 1 transistor is the temperature sensor, 2 used as heaters and the 4th is the actual log or anti-log transistor. Also used an opamp to control the heaters. Temperature is held at 69 C within less than 1/10 a degree. Even had an idiot light that shows red during warmup (about 5 seconds) and turns green at temp. Voltage reference came from an ADI AD-588.
G=B2
understand --
aamp,
nA.
Have you considered the possibilities of reverse diode connected?
??-))
--=20
diode=20
the=20
Thank you much for dumping gaggle gropes.
reward"
understand
I_e
is
you
what
but
future.
temperature
temperature
known
Though it will work, i expect that suggested quad transistor solution(s) will work seriously better.
?-)
I wonder if the string could oscillate. Sort of like traffic-jam dynamics.
John
--
=A0 =A0 =A0 =A0 =A0
=A0 =A0 =A0 I_ec
ng
(to
f s s=3D20C,
y rage
some
=A0 =A0...Jim Thompson
=A0 =A0| =A0 =A0mens =A0 =A0 |
=A0 | =A0 =A0 et =A0 =A0 =A0|
=A0|
=A0 =A0 =A0 |
Yes, that is how they sense die temp in CPUs. Oh, don't get a big head because I'm agreeing with you.
You mean connect base to emitter and run the transistor upside down?
Running a BJT upside down gets you a different BJT, one with a breakdown voltage of about 6 volts or so, a beta of around 5, and a really really low saturation voltage. The low V_CEsat is really the only reason to do that, and in the diode connection the transistor never saturates, so there's no benefit that I can see.
I've never measured the f_T of an inverted transistor, though--has anyone here done that?
Cheers
Phil Hobbs
Hmm, interesting possibility--I don't know the answer.
Cheers
Phil Hobbs
See US 5654662.
Best regards, Spehro Pefhany
VCEsat is an inverse function of reverse beta, so a transistor operated in inverse mode has a reverse beta of the forward device ;-) Main use is switching. There used to be symmetric transistors called "INCH" (IIRC) for low on-drop in both directions. ...Jim Thompson
[On the Road, in New York]
How about the high frequency characteristics? Would a cascode behave better or worse with an inverted BJT as the upper transistor?
Best regards, Spehro Pefhany
I don't know. Interesting question. I _think_ it might actually be wider bandwidth, but VAR (output impedance slope... someone explain that to Larkin :-) might take away any improvements. ...Jim Thompson
[On the Road, in New York]
Zetex has some transistors with rater high beta in the inverted mode
- and still low Vsat..
Cool, thanks. In a current source application, the small capacitance helps a lot, and the crummy beta doesn't hurt much. I'll toss that into the bag of tricks.
Cheers
Phil Hobbs
Have something to add? Share your thoughts — no account required.
Ask the community — no account required