Question about diode temperature and forward voltage

[Jonathan Kirwan]

Jon,

Great. I've never seen Isat in that form. I'll do some work in matlab
and see if it works out. I've checked the spice results and it is about
what you said. Thank you for taking the time to delve into this.

No problem. I was also bugged in the same way. Luckily, I happen to
have "Modeling the Bipolar Transistor" by Ian Getreu (Tektronix, at
the time) written circa 1979-1980. It's an excellent book on the BJT
and covers detailed procedures in the last half of it for deriving the
various parameter values from specific instrumentation and measurement
procedures he discusses. That last half is actually the more valuable
part of it, as I've not found any other single source discussing very
specific means that one can actually set up on their bench for
measuring BJT parameters up through Gummel-Poon. But it also includes
this Isat equation, too, in the very first few pages.

Jon

 

[Phil Allison]

PHYSICS is the study of PHYSICAL PHENOMENA, such as the operating
characteristics of a p-n junction. That's where the diode equation
comes from. To say that the diode equation has nothing to do with the
physical operation of the device is incorrect.

** Learn to read - IDIOT !!



....... Phil

 

[Phil Allison]

PHYSICS is the study of PHYSICAL PHENOMENA, such as the operating
characteristics of a p-n junction. That's where the diode equation
comes from. To say that the diode equation has nothing to do with the
physical operation of the device is incorrect.

** Learn to READ - you Google Groping IDIOT.

AND DO NOT TOP POST !!!!



....... Phil

 

[G. Schindler]

How do you get If/Is < 1? *Is* is REALLY small, isn't it (pico amps)?

Bob

You are right about Is being really small .... obviously I was mistaken
about trying to make If < Is. I had to go back to the books and recheck
the formulas and the constants. When I did this, I found/realized that
'q' the charge on an an electron is negative (which makes sense). This
factor corrects the direction of the slope for you. As many later
contributors have shown the formula you started with is an approximation
based on some assumptions. This is not a bad thing, it all depends on
the level of accuracy you desire and the accuracy of the
characterization for the diode in question.

BTW ... Somewhere in this thread someone suggests that 'k' (Boltzmans
Constant) is negative but that is not shown in my textbooks.

Good luck.

 

[Hans Lohninger]

I'm trying to understand how a diode can be used to detect changes in
temperature. According to all the literature I've read, a diode's
forward voltage, Vf, falls by 2 mV for every 1 degree C increase in
temperature. This doesn't seem to agree with the diode equation, which
suggests that Vf should increase with temperature.

Vf = k*T/q * ln( If / Is)

What am I missing?

Here's a page showing (in simplified form) how to estimate the temperature
coefficient from the diode equation:

http://www.vias.org/feee/sensor_temp_diode.html

Regards,

Hans

 

[Ban]

Here's a page showing (in simplified form) how to estimate the
temperature coefficient from the diode equation:

http://www.vias.org/feee/sensor_temp_diode.html

Regards,

Hans

If you really wrote that, you should understand it first. Not only is the
calculation plainly wrong, you also guessed this 10^-12A=Is. T means
absolute temperature and is around 300K, one degree more it is 301K, That
makes your coefficient positive.
In case of diodes made from materials with high energy gap, the reverse
biasing current Is cannot be calculated from the diode equation, due to
recombination of charge carriers.
Is is the reverse leakage current and proportional to T^alpha*e(-q*Vg /2kT)
alpha is around 3, Vg=1.124V for silicon

 

[Phil Allison]

"Ban"

If you really wrote that, you should understand it first. Not only is the
calculation plainly wrong, you also guessed this 10^-12A=Is.

** Seems to be a classic example of a FAKE calculation.

The author knew what the *answer * ought to be and contrived his estimates
values to get it.

See what happens if you make the innocent change of I test to 1mA and I sat
to 1 uA ?

VOILA - the diode's tempco is suddenly 3 times smaller !!






........ Phil

 

[Ban]

** Seems to be a classic example of a FAKE calculation.

The author knew what the *answer * ought to be and contrived his
estimates values to get it.

Guess who is the idiot who made up this page?(look in the credits on that
site) Even after reading the replies here he still recommends it. This
really shows his utter ignorance and stupidity. Well the web is full of
these fake specialists, reminds me of an Australian site though...