Possible Affect of Lower Gate to Source Threshold Voltage for an Inverter Application

Thread Starter

mohtishamali

Joined Jul 9, 2021
55
Good Day,

I have a question regarding gate to source threshold voltage of MOSFET. We were using MOSFET AUIRFP4568 and it has to be replaced now with IXTH150N15X4 in our 1 KW 24V 40A Inverter Application. As per datasheet IXTP MOSFET has lower VGSth as compared to AUIRF and at higher tempratures, the IXTP device VGSth is decreasing more rapidly as compared to AUIRFP. This makes this device more vulnerable against self turn on. However, this assessment is done only on the basis of datasheet information. What can be done in order to make concreate opinion about diffrence between VGSth of both MOSFETs. What kind of tests you will suggest to be carried out in lab in order to have a clear picture.
Below is information provided in datasheet and will wait for your valuable feedback.
Thanks in advance
Test conditions: ID=4mA
For IXTP device VGSth @ 25 °C is 3.5V & VGSth @ 120 °C is 2.45V
Down to 30% of initial value hence decreasing rapidly
For Infenion device VGSth @ 25 °C is 4.47V & VGSth @ 120 °C is 3.60V
Down to 20% of initial value hence less rapid decrease in VGSth as compared to IXTP MOSFET
1670226705716.png1670226732723.png
 

Alec_t

Joined Sep 17, 2013
13,231
If your inverter output is ~40A and the FET has a robust gate drive Vgs of at least 10V, why are you concerned about the threshold where the test current is only 4mA?
 

BobTPH

Joined Jun 5, 2013
6,080
If there is a lot of switching loss then a lower threshold would aggravate the problem by having the transistors in saturation mode longer.
 

Thread Starter

mohtishamali

Joined Jul 9, 2021
55
If your inverter output is ~40A and the FET has a robust gate drive Vgs of at least 10V, why are you concerned about the threshold where the test current is only 4mA?
Because this inverter has very rough and tough application so we have to make sure this aspect is fully addressed. Can you comment which tests can be performed in order to investigate this furthur?
thanks
 

crutschow

Joined Mar 14, 2008
31,126
If there is a lot of switching loss then a lower threshold would aggravate the problem by having the transistors in saturation mode longer.
I do not understand your reason for that assumption.
It's how fast the gate voltage goes through the threshold region (saturation region) into the fully-on linear region, not what the threshold is.
If anything, it would seem that for a lower threshold voltage, and the same gate voltage signal transition, the lower-threshold device would switch faster than the high-threshold device.
 

BobTPH

Joined Jun 5, 2013
6,080
I do not understand your reason for that assumption.
It's how fast the gate voltage goes through the threshold region (saturation region) into the fully-on linear region, not what the threshold is.
If anything, it would seem that for a lower threshold voltage, and the same gate voltage signal transition, the lower-threshold device would switch faster than the high-threshold device.
You are correct, I did nor think that through. I was thinking of the on voltage staying the same.
 

Thread Starter

mohtishamali

Joined Jul 9, 2021
55
I do not understand your reason for that assumption.
It's how fast the gate voltage goes through the threshold region (saturation region) into the fully-on linear region, not what the threshold is.
If anything, it would seem that for a lower threshold voltage, and the same gate voltage signal transition, the lower-threshold device would switch faster than the high-threshold device.
Thanks for your valuable comment.
For an application where the MOSFET will operate at higher temperatures from 85Cdeg to 100Cdeg, this self turning on phenomenon becomes relevant. Since new device's VGSth increases more rapidly with increase in junction temperature thats why the idea is to investigate it further in the lab
 

Alec_t

Joined Sep 17, 2013
13,231
thats why the idea is to investigate it further in the lab
The curves shown relate to junction temperature, not case temperature. Does your lab have access to the FET junctions, or will you be assuming some constant thermal resistance between case and junction?
For an application where the MOSFET will operate at higher temperatures from 85Cdeg to 100Cdeg, this self turning on phenomenon becomes relevant.
Are those junction temperatures or case temperatures?
For gates driven hard by an active component I'm surprised that self turning on can occur. Just how are you driving the gates?
 

Thread Starter

mohtishamali

Joined Jul 9, 2021
55
The curves shown relate to junction temperature, not case temperature. Does your lab have access to the FET junctions, or will you be assuming some constant thermal resistance between case and junction?

Are those junction temperatures or case temperatures?
For gates driven hard by an active component I'm surprised that self turning on can occur. Just how are you driving the gates?
I am talking about case temperature. we are using this gate driver from infenion to derive the mosfet. AUIRS21814S
 
Top