Measuring Power MOSFET VDS

[HariHerekar]

Measuring Power Mosfet VDS
Just looking for a rudimentary explanation.

Was reading about how junction temperature (Tj) measurements are done for a Power MOSFET under power cycling conditions.
As per my understanding, a thermal chamber is used to heat the entire DUT to a uniform temperature, after which a test current is passed through the DUT and the corresponding VDS is noted down. Essentially what we create is a table of VDS vs Temperature for all values of temperature.

Now when we wish to measure the Tj during power cycling, we note the VDS, compare it with our previous table and infer the Tj.

This makes sense to me so far, however I do not understand how a test current produces a voltage drop, in cases when the channel hasn't been formed yet. For example, when the DUT is being cooled, the VG applied is negative, so as to completely close the channel and prevent any conduction i.e. heating from occurring. In such a case for Tj measurement, how will a test current produce a voltage drop ? Does a power mosfet conduct small currents for VDS>0 even when the gate voltage is negative ?

Please feel free to correct my understanding of things and I would appreciate if someone could explain how we measure the VDS using a test current for a Power MOSFET

 

[dick_freebird]

Seems like a messy approach and one that probably will
embed error.

Vds (really Vdb) can be used to measure temp (at the D-B
junction, not necessarily same as other locations) if you
put the body diode forward-biased at a spec current
(same as any other PN diode). Harder to do in-circuit
and know where every bit of the current goes.

Measuring channel would be a real waste of time but
the diode is pretty consistent.

 

[Easy peasy]

If you apply Vds maximum - you will always get a leakage current - even for Vds = 0

 

[KlausST]

Hi,

Was reading about how junction temperature (Tj) measurements

Please give a link to this document, so we can read what you have read and discuss about.

***
I´ve worked at a company for semiconductor testing.
For testing temperatre they (low frequency pulsed) the DUT. In times when FET is OFF (V_GS = 0) they applied a current limited (1mA..10mA) negative voltage to DS, so that the internal body diode becomes conductive.
The diode gives rather precise die temperature readings with a ratio of about 2mV/°C.

Klaus

added: Just saw that dick_freebird also mentioned this method.

If you apply Vds maximum - you will always get a leakage current - even for Vds = 0

I can´t imagine how it works.

Klaus

 

[HariHerekar]

Hi all, appears that I my understanding of the test method was flawed.
The test being described is of the body diode present in the MOSFET, not of the VDS.

Apologies for misleading you all, I myself was confused about the process.

Although, am I correctly describing it now ?

My understanding now is that the VSD ( of the body diode ) is measured using a small test current, after a heating cycle of the test is completed.
This VDS value, will allow us to know what temperature the Tj was at.

 

[D.A.(Tony)Stewart]

-2 mV/'C @ 1.3 mA

https://www.ti.com/lit/an/sboa277a/sboa277a.pdf?ts=1724207806793

Use that with a CC source err. sink. to drain.
Measure Vf quickly.

--- Updated Aug 21, 2024 ---

Consider this; Calibrate at room temp , 0'C with ice water.