Power mosfet gate charge vs. Gate-source-voltage charateristic

[bryanurbe]

Hello To All,


Hope this message finds everyone well. I have the following question:


The Qg Vs. Vgs characteristic provided for many power MOSFETs is often limited to a peak Vgs of 10V and it is given for some assumed test conditions (i.e. Vds = 300, Id = 55A, & Ig = 10mA - please see attached datasheet for IXFN110N60P3). That being said, my questions are:


1) How should I determine my desired gate charge if I want to operate the devise at a different set of values than those assumed by the manufacturer in generating the Qg Vs. Vgs characteristic curve.


2) If I want to apply a Vgs of 15V (i.e. to achieve minimum Rds on), is it acceptable to obtain the required gate charge by extrapolating the given Qg Vs. Vgs characteristic?


Thanks in advance for any assistance you may be able to provide.


Regards,



Bryan

 

[rumpfy]

1. The gate charge is relatively fixed by the gate input capacitance. With a higher gate voltage, the gate charge is increased. The data sheet seems to imply that gate drive voltages of more that 10 volt are not necessary in order to drive the device into full conduction. It is the switching losses which generally dominate the device dissipation. The gate charge is the charge that must be removed in a VERY SHORT TIME in order to keep the device dissipation low. Rapid removal of gate charge requires a high peak current and this may be the limitation of the device.

2. From the above, I'm not sure you need to raise the gate voltage. I'm also not sure that increasing the gate voltage will reduce Rds on. This is a high speed device so you need to concentrate on the switching effects as well as the DC effects.

Others may a better take on this.

hope this helps.

 

[MrAl]

Hi,

If you are unsure then another way to handle this would be to forget about the gate charge concept and go with the direct analysis knowing the capacitances present in the MOSFET device. It's been a long long time since i had to do this myself so i'd have to look it up again, but basically it means doing an analysis of the MOSFET as a switch with various capacitances, and finding out the effect of the gate current on the switching speed as a result of having those capacitances present.

 

[MrAl]

Hello again,

It dawned on me that it may be simpler for you to measure the gate charge requirement for your application than to actually try to calculate it.
You'd have to set up your device to switch with the required drain voltage and load, and measure the gate to source voltage or maybe the drain voltage. If the time it takes to switch the drain voltage to near zero is t1 and we start applying a gate current Ig at time t0, then the total gate charge is:
Q=(t1-t0)*Ig

When looking at the gate voltage, it will first ramp up past Vth (due to the gate source capacitance) then it will level off (due to the gate to drain capacitance and falling drain voltage) then later it will start to ramp up again. The time t1 above will then be the time when it starts to ramp up for the second time.

So the drain voltage falls more or less as a ramp, and the gate voltage rises as a ramp, then flattens out, then rises as a ramp a second time. Either the start of the second gate ramp or the end of the drain ramp is the time we call t1.

Either of these might be the best way unless you can get a spice model and try to get the right values for that model too.

 

[bryanurbe]

rumpfy and MrAl,

Thank you for your input.

Bryan