I came across this configuration while trying to understand the remote temp sensing ability of the TPS4811.

It seems as if the diode pin voltage is responsible for turning on the BJT. V(BE) being >= 0.7V. The same is also controlling the Collector current & Base current.

How exactly is this working to enable remote temperature sensing ability of this IC?


Or rather how is it able to sense temperature?


2 Answers 2


The terminal behaviour of a diode can be approximately represented by the equation

$$I_\mathrm{D} \approx I_\mathrm{S}\mathrm{e}^{V_\mathrm{D}/V_T},\tag{1}$$

where \$I_\mathrm{D}\$ is the diode current, \$I_\mathrm{S}\$ is the reverse saturation current and \$V_T = k_\mathrm{B}T/e\$ is the thermal voltage, being \$k_\mathrm{B}\$ the Boltzmann constant, \$e\$ the elementary charge, and \$T\$ the thermodynamic temperature. The above relationship holds for \$I_\mathrm{D}\gg I_\mathrm{S}\$.

If we periodically switch the current \$I_\mathrm{D}\$ between two values \$I_1\$ and \$I_2\$, we obtain

$$\frac{I_2}{I_1} \approx \mathrm{e}^\frac{V_2-V_1}{V_T},$$

where \$V_1\$ and \$V_2\$ are the diode voltages measured with the two currents. From the above equation, solving for \$T\$,

$$T \approx \frac{e(V_2-V_1)}{k_\mathrm{B}\ln(I_2/I_1)}.$$

The main idea of this arrangement is that it allows to measure the temperature \$T\$ cancelling the dependence on the diode parameter \$I_\mathrm{S}\$, which is difficult to characterise and is also temperature-dependent. Note that the diode is not used as an ON/OFF device.

Now, why using a diode-connected BJT instead of a proper diode? The reason is that the characteristic of a diode-connected BJT follows the model (1) much better than a diode. Hence, the temperature measurement is more accurate.

In the TPS4811x, \$I_1\approx 10\,\mu\mathrm{A}\$ and \$I_2\approx 160\,\mu\mathrm{A}\$.

Other details on this type of measurement can be found in the following document:

Texas Instruments, Optimizing Remote Temperature Sensor Design, SBOA173A Application Report, 2019.

  • 2
    \$\begingroup\$ Good and comprehensive answer. \$\endgroup\$
    – LvW
    Commented May 1, 2023 at 12:44

The forward voltage of a PN junction is temperature dependent. By driving a constant current through the junction the temperature can be calculated by measuring the forward voltage.

The transistor contains two PN junctions but only one is required so the second one is short-circuited.

  • \$\begingroup\$ Would there have been any difference if we used a simple diode itself. On the surface level in both the cases the V(f) would be the same. \$\endgroup\$
    – Harkirat
    Commented May 1, 2023 at 9:35
  • \$\begingroup\$ I have read about this in the past but don't remember. There may be a difference in the junction design / function for a diode and for an amplifiying transistor. The transistor may be easier to mount with good thermal contact. \$\endgroup\$
    – Transistor
    Commented May 1, 2023 at 9:43
  • 2
    \$\begingroup\$ @Harkirat If I remember correctly, BJTs generally have junctions that act more like an ideal diode (ideality factor closer to 1) than typical discrete diodes, so they're preferred for anything like this that relies on the exact turn-on characteristics. \$\endgroup\$
    – Hearth
    Commented May 2, 2023 at 2:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.