At a moderate bias voltage of the diode one can expresses the diode equation in the form

I= I0 exp V/nVt , n is the ideality factor and Vt is the thermal voltage kT/q, kT is the thermal energy and q is the electronic charge.

Taking the ln of both side one gets

V=nVt ln I/I0,

I0= I00 T3 exp -Eg/ Vt,

One can inject constant current in the diode and measure the voltage of the diode as a function of Temperature.

Only the T^3 term is the main cause of the week nonlinearity of the thermometer.

The characteristics can be linearized numerically.

Better than the diode one uses a transistor with short circuit between the collector base junction.

Linearization can be also accomplished by using to identical diodes one is a reference at certain temperature and same current. Then one can calibrate the difference of the voltage to measure the temperature difference.

You can follow the diode thermometer in the book at the link: Book Electronic Devices

Best wishes.

At a moderate bias voltage of the diode one can expresses the diode equation in the form

I= I0 exp V/nVt , n is the ideality factor and Vt is the thermal voltage kT/q, kT is the thermal energy and q is the electronic charge.

Taking the ln of both side one gets

V=nVt ln I/I0,

I0= I00 T3 exp -Eg/ Vt,

One can inject constant current in the diode and measure the voltage of the diode as a function of Temperature.

Only the T^3 term is the main cause of the week nonlinearity of the thermometer.

The characteristics can be linearized numerically.

Better than the diode one uses a transistor with short circuit between the collector base junction.

Linearization can be also accomplished by using to identical diodes one is a reference at certain temperature and same current. Then one can calibrate the difference of the voltage to measure the temperature difference.

You can follow the diode thermometer in the book at the link: Book Electronic Devices

Best wishes.

Dear Professor!

I sincerely thank you for such a scientifically significant answer!

I am more engaged in technical electronics and in practice choosing the optimal junction mixture I built a thermometer based on a factory diode and a transistor with sufficient accuracy of 0.1-0.2 ° C and using a linearization circuit in the feedback of the conversion circuit, I achieved linearity in a wide range (-20 - 100 ° C).

But I want to find out whether the technological and technical effects of factory elements in laboratory conditions are possible, after which it will be possible to significantly improve these indicators?

Dear Islam,

would you like to relate the temperature with voltage of the transistor at a given current and physical and technological parameters of the device?

As i said the best diode that can be used is the biploar transistor shortcircuit collector current.

In this case the collector current can be expressed by

IC= [q D ni^2 exp V/Vt]/ NA WB with D is the diffusion coefficient,

Na the acceptor concentration in the base and WB is the base width ni is the intrinsic concentration.

This equation is given in the link in my first post.

Best wishes

Dear Abdelhalim Zekry!

I want to note that thanks to the discussion with you, I recognized one of the leading scientists in the field of Electronics. This is very gratifying to me. My regards!

As you have noted in fact, a bipolar transistor turned on in diode mode, according to the circuit, as you emphasize, gives better results. And this I actually investigated before in practice. I am sure that your advice and comments will allow you to get the best results.

Regards and best wishes, Islam.

Please read the book

Microelectronics

by

Grabel

from

Northeastern University

Boston, MA

Dear Ziad Sobih! I know the general and standard principles for using the pn- junction as a sensor. And I have sufficient practical experience in exploring it as a temperature sensor. I want to find some non-standard methods.

Thank you for your answer!