Hi Highly Charged Elements,

I'm looking for a high speed (>= 500 kHz) MOSFET H-Bridge, for PWM operation at 500kHz or faster. About 1A of current should be fine enough as I need to deliver just 1-2W of power (say max 5W). It will drive a trafo which will finally drive a piezo, that being the purpose. (The modulating signal would be band-limited to under 70kHz if you care to know). Simple PWM input, no reversal of "direction" is needed, would be nice to have output disable to a high-Z state independently of the PWM input. So half bridges would be just fine.

With that in mind, I was looking for TI parts but couldn't find any beyond 200 kHz. So far my searches have turned up only ROHM BD63576 and her cousins BD65496 BD65491 etc which are good upto 500 kHz -- see ROHM Brushed DC motor driver BD63576nux [https://www.rohm.com/products/motor-actuator-drivers/dc-brush-motor/bd63576nux-product]. (These are for Brushed DC motor control, by the way).

Looking for any suggestions or alternative ideas from you all. Thanks!

 

You don't say what voltage you need, but if it's less than 18V can you use a MOSFET gate driver IC.
After all, a piezo is a capacitor just like a MOSFET gate, and probably has a similar amount of capacitance.

 

You can check Infineon's motor driver ICs, such as the BTS7960B.

 

Here's a stout FET-Gate-Driver that will do what You need ...........
.
.
.

.

 

Simple PWM input, no reversal of "direction" is needed,

So why an H bridge? The normal arrangement would two low side switches driving both sides of a center tapped transformer.

 

You can check Infineon's motor driver ICs, such as the BTS7960B.

Looked up the datasheet, it says "PWM capability of up to 25 kHz". Thanks anyway.

 

So why an H bridge? The normal arrangement would be a half bridge driving both sides of a center tapped transformer.

True! I was also saying that -- a half-bridge could be used. And keeping DC net = 0 to avoid core troubles etc.

 

You don't say what voltage you need, but if it's less than 18V can you use a MOSFET gate driver IC.
After all, a piezo is a capacitor just like a MOSFET gate, and probably has a similar amount of capacitance.

The trafo hasn't been picked, so for a few watts, under 18V is feasible. True that the piezo is like a capacitor, but near its resonant frequency (where we will operate it) the game changes and it will be ready to carry a few watts across (series resonance, with moderate Q values). But the reflected impedance at the trafo primary would be about a thousand times smaller (assuming a 1:30-ish trafo). For approx 3W I guess something like 0.5A would have to flow, I'll have to see if I can spot a FET gate driver that can sink/source that much. Thanks for the idea in any case.

 

Here's a stout FET-Gate-Driver that will do what You need ...........
.
.
.
View attachment 307256
.

Yes this indeed is a beast with 30A and fast switching times. Wow!

 

True! I was also saying that -- a half-bridge could be used. And keeping DC net = 0 to avoid core troubles etc.

Actually, you don’t need even a half bridge with a enter tapped transformer, two low side switches is enough.

If you are not using a center tapped transformer, you need the H bridge because you do need to reverse the current.

 

Actually, you don’t need even a half bridge with a enter tapped transformer, two low side switches is enough.

If you are not using a center tapped transformer, you need the H bridge because you do need to reverse the current.

Ah yes...you caught me there

 

Yes this indeed is a beast with 30A and fast switching times. Wow!

.
And, it's available with an Enable-Input.
And, it comes in a TO-220-5 Package so it can dissipate substantial amounts of Heat,
this will come in handy when things don't go exactly to plan, avoiding the dreaded "Blue-Smoke".
But, the DC Current handling is actually only ~8-Amps,

A Piezo-Element can have much more Capacitance than a MOSFET-Gate,
this may require substantial amounts of Current under certain circumstances.
I would consider installing a Resistor on each Driver to limit the maximum Current,
or, this could better be accomplished by Regulating-down the Power-Supply-Voltage instead.
They can be paralleled by using Output-Resistors,
but at that point, I would bet that you'll smoke some Piezo-Elements, they do generate HEAT.
.
.
.