# What Driver Voltage to Drive a Mosfet? (SMPS app)

• posted

My power mosfet has an absolute gate spec of 30VDC. Most of the Mosfet datasheet gate behavior graphs only go to 10V.

Is there a drive voltage that'll provide low switching loss? Looks like there's a 6V to 30V range to choose from.

Mosfet Specs Gate charge 40nC Input capacitance 1500pF Absolute max gate voltage 30V Id=10amps @Vgs=7V

Circuit Specs f=600khz Id=2amps Mosfet driver can be powered up to 36V. D from BC

• posted

Do you mean by loss the energy dissipation on the Drain-Source side of the MOSFET, during the switching? It will largely depend on how fast the MOSFET goes from the "open" to "closed" state. I would start by choosing first the Vgs_1 high enough to get the Vds under load as low as you need it for your "charge" part of the cycle, and then would go on optimizing the circuit on the Mosfet drive side to get Vgs from 0 to Vgs_1 and back as fast as possible. Then simulate and validate the energy dissipation.

-- Andy

• posted

International Rectifier has a good application note on driving mosfets.

Application Note AN-937 Gate Drive Characteristics and Requirements for HEXFET Power MOSFETs

I don't have the link but a search ought to do it.

When they give you the characteristic curves, that tells you what it takes to use it as a saturated switch - what you are doing . . .

The ten volts is probably turned on fully so there's no point in carrying the graph out further.

The only things to watch for is to get enough current to the gate fast enough to charge and discharge the 1500 pf of gate capacitance - or you'll be in the linear region and creating heat. They may have a very high gate impedance - but that pesky gate capacitance requires current to keep it from spending time in the linear region during the transitions - so the driver impedance has to be low

Keep transients on the gate (and drain) subdued or they might easily exceed the 30 volt maximum and cause a failure. That usually means short wire runs from driver to gate and/or gate resistor or snubbing network. If you rely on diodes for protection - they have to be fast switching diodes.

--
----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+
• posted

Yup..it's the gate peak voltage for low drain source switching loss. I'm letting the mosfet do whatever it can do in the way of rapid current rise and fall times for reducing Pds switching loss.

Are you saying I should use the lowest gate voltage possible for sufficient drain current? That way there is less gate charging and discharging? A gate charged to say 6V can be discharged faster than say a gate charged to 20V.

D from BC

• posted

I'm using a IXDD414 mosfet driver with a impressive speedy 14A peak current rating. Probably overkill but just using overrated parts for assurance. I'll check out the app note and take care with the PCB layout for the gate trace. Thanks. D from BC

• posted

The lowest gate voltage (under the range of the working conditions and the part-to-part variations) to get Vds at the given Id (or Rds) as low as you need.

Yes. The spec that you cite "Gate charge 40nC" is valid for specific starting and ending Vgs only.

Yes.

-- Andy

• posted

Ooops..I just realized I forgot a word in my last post * * Rewrite: "I'm letting the mosfet *driver* do whatever it can do in the way of rapid current rise and fall times for reducing Pds switching loss." That probably caused some confusion...

Thanks for the info. Gonna juggle some parameters now. :) D from BC

• posted

Why such a high F? Trying to push gate driver dissipation in exchange for better...something?

I don't see any advantage over 100 or 200kHz, but that's me.

Tim

-- Deep Fryer: A very philosophical monk. Website @

• posted

I'm still learning smps design.. I'm trying out a smps design of mine at 600khz (150Watts) to reduce the size of the magnetics.

At that frequency, it probably becomes important to carefully select the mosfet driver peak gate voltage (or driver supply rail).

My actual built and tested design works great at 100khz...I just want to find out if I can push the design further. :) If it goes well at 600Khz...I hope to take the design into the Mhz next.

So..yah..the mosfet D-S switching loss is something I'm concerned with. Are you saying that power conversion above 200khz gets ugly? D from BC

• posted

Include this in the current spec of the driver. This much charge must be moved quickly to switch the part.

Include this in the current spec of the driver. This gives a bunch of extra charge you have to move. It isn't all right at the instance of switching though.

You sould also look for the capacitance from the drain to the gate. This makes a bunch more charge for you to move with the driver.

Stay way below this number.

Swing up to about 10V to stay well above this number.

• posted

That is an impressive driver IC

I found this