# How to improve mosfet switching timing?

I am a hobbyist and I am not a native so bear with me please. I was looking for a fast mosfet and by looking at some datasheets I noticed that switching time is not a constant parameter like this one:

Which is a good thing. After a little bit of digging I found out that it has to do with the mosfet internal capacitance. In the picture above, the rise of gate voltage reduced the rise/fall time and also the delay time. At that point I am confused and I need to know if it is possible to tweak those parameters individually? What capacitance should I calculate GS GD DS (time=R(ohm)×C(Farad))? And how to calculate the gate voltage (a lot of mosfets are rated for up to 20V but statistics is only up to 10V).

edit: As suggested in @Neil answer i have found that this mosfet ISZ0703NLS has a low charge gate and low rds(ON) hope you can help me guys with examples thank you.

Edit:
I have found this document expaining a very imporatant parameter that a lot of member may want to know about and that they always refer to as "non linearity" hope it will help; source

• Please cite the source for this image -- this is required by rules, but also allows readers to check and verify your source material, or see other context provided by it if needed. Commented May 28 at 16:54
• @TimWilliams the image is just an example the serie is up left. Commented May 28 at 18:01

The way to improve the switch timing of any individual MOSFET is to put as many amps as you can into or out of the gate terminal, to change the gate charge quickly. There are many 'FET gate driver' ICs available, which claim to be able to source and sink 4 A, using a supply voltage of 10 to 12 V.

If you are looking at several FETs, then the gate charge is a bit more fundamental than the capacitance figures, look for devices with a lower gate charge, they will be easier to drive fast. Capacitances are only effective figures, based on the fact that the voltage and charge changes, so their ratio is a capacitance. But it's very non-linear, see the switching versus gate charge graphs that appear in every data sheet.

That particular FET has a max VGSthresh of 2.5 V, so the quoted gate voltages are 4.5 V, the lowest you would expect to be able to use to turn it on properly, and 10 V, above which the RDSon doesn't improve much more. You can drive the gate to 20 V, but then you have to be careful about accidentally exceeding that with transients, a 10 V design drive gives you a lot more margin for safety.

It's difficult to design to a particular switching time from the data sheet, as you can see the figures are for one particular set of currents and voltages. You need to do your own bench testing under your conditions well before you finally select any FETs.

• The first paragraph is for switching ON and OFF i guess this that mean it is not possible to individually control them? For as much as i can amps thing, i have read somewhere that it has somethine to do with frequency example and correct me if I am wrong for 1MHz 1 nC => 1000000nC = 1Amp. Commented May 28 at 18:11
• I mean 1 mA sorry. Commented May 28 at 19:13
• Look at the total gate charge curve on the datasheet. Pick the voltage you need to get the RDSon you want. (E.g. 4.5 V). Read the charge required from the curve, and then supply that charge in the desired amount of time you need to switch. For your example FET it looks like about 16 nC to get to 4.5 V Vgs. If you want to do that in 1 ns just supply 16 A for the 1 ns and you're good to go. Unless the real world gets in the way. Commented May 28 at 20:13
• @Tintin I think you're confusing several concepts here. Charge in coulombs is not capacitance. There are many practical considerations as well like gate resistance and gate and layout inductance that can make fast switching difficult. But go ahead and experiment, building stuff is a good way to learn. Commented May 29 at 16:05
• @Tintin Comments on this site aren't really for a lot of back and forth discussion. You can start another question with one clearly worded topic and one question and we can start there. Also, using scientific or engineering notation along with correct units will go a long way toward making your question legible. Commented May 29 at 19:19