MOSFETs H-Bridge 3.3V to 36V level shift

[valdorf]

Hello everyone,

currently I am designing a circuit that can shift a voltage level from 3.3V 200kHz PWM signal to 36V PWM (an ultrasonic transducer). This ultrasonic transducer consumes only 10-20mA max at 36V.
I do not know if this is best solution, but I think about using H-Bridge like I found one DMHC10H170SFJ https://www.diodes.com/assets/Datasheets/DMHC10H170SFJ.pdf which has small surface mount package, low Rds(ON), more than enough Drain-Source Voltage.

As I understand correctly, I should convert 3.3V to at least 10V for Vgs? The problem is that I do not have additional power supply like 10V, but only 3.3V, 5V and 36V. What would be solution to control this H-Bridge with 3.3V PWM input?

 

[betwixt]

You actually need more voltage than just Vgs. The bottom two MOSFETS need Vgs above ground but the top two need Vgs above the center of the bridge.
Look at bridge driver ICs, such as the IR2100 series to see how it is done using a bootstrap capacitor to raise the top MOSFET gate drive voltage.

Brian.

 

[austinmariam]

Hello everyone,

currently I am designing a circuit that can shift a voltage level from 3.3V 200kHz PWM signal to 36V PWM (an ultrasonic transducer). This ultrasonic transducer consumes only 10-20mA max at 36V.
I do not know if this is best solution, but I think about using H-Bridge like I found one DMHC10H170SFJ https://www.diodes.com/assets/Datasheets/DMHC10H170SFJ.pdf which has small surface mount package, low Rds(ON), more than enough Drain-Source Voltage.

As I understand correctly, I should convert 3.3V to at least 10V for Vgs? The problem is that I do not have additional power supply like 10V, but only 3.3V, 5V and 36V. What would be solution to control this H-Bridge with 3.3V PWM input?

You’re right, the MOSFETs need a proper gate drive voltage. Since you don’t have a 10V supply, you could use a gate driver IC that can take 3.3V logic input and boost it to the required gate voltage. Something like the IR2184 (for half-bridge) or a dedicated full-bridge driver powered from your 36V rail could work. Also, check if a bootstrap circuit can help generate the needed voltage for the high-side MOSFETs.

 

[KlausST]

Hi,

is it a logic level shift?
Or do you also need the power to be converted?

In one case a logic level controlled analog multiplexer would do.
Or - in case you don´t need push-pul driving: a simple open_collector or open_drain solution may be possible.

In the other case a transformer would not only convert from 3.3V to 36V but also the power is converted.

Klaus

 

[valdorf]

Hi,

is it a logic level shift?
Or do you also need the power to be converted?

In one case a logic level controlled analog multiplexer would do.
Or - in case you don´t need push-pul driving: a simple open_collector or open_drain solution may be possible.

In the other case a transformer would not only convert from 3.3V to 36V but also the power is converted.

Klaus

I want to drive my transducer at 36V, from 3.3V 200kHz MCU PWM. My transducer consumes only 10-20mA at 36V, so I really dont need high power circuit. If it is possible, I would like to drive with a single NMOS. After googling, I found there are transistors will low gate thershold voltage like https://assets.nexperia.com/documents/data-sheet/PMV37ENEA.pdf Maybe it is possible to drive it directly from my MCU 3.3V PWM signal? Or for example to use gate driver such as https://www.ti.com/lit/ds/symlink/ucc27511.pdf with Vcc 5V to output mosfet gate 5V PWM?

 

[danadakk]

Take a look at this, the section on high side gate drive -

https://www.ti.com/lit/ml/slua618a/slua618a.pdf

Direct drive, low side, from MCU with huge Cgate at 200 Khz not advised. You need
drivers that can handle the large C to get fast (relative term) switching, and to get the
boosted Vgs needed for high side fets, as mentioned in prior threads. You can parallel
several MCU outputs to get greater drive, but there is a tradeoff dumping all that
current thru MCU internal power rails.....external driver good....


Regards, Dana.

 

[valdorf]

Take a look at this, the section on high side gate drive -

https://www.ti.com/lit/ml/slua618a/slua618a.pdf

Direct drive, low side, from MCU with huge Cgate at 200 Khz not advised. You need
drivers that can handle the large C to get fast (relative term) switching, and to get the
boosted Vgs needed for high side fets, as mentioned in prior threads. You can parallel
several MCU outputs to get greater drive, but there is a tradeoff dumping all that
current thru MCU internal power rails.....external driver good....


Regards, Dana.

What do you think about combination of N-Mosfet PMV37ENEA and gate drive UCC27511. A 3.3V PWM 200kHz signal from mcu is connected to gate drive which powered from 5V supply. Gate drive now outputs 5V amplified pwm signal to mosfet gate. This would be enough for switching 36V for my transducer?

 

[danadakk]

I thought you wanted 36V operation.....?

 

[valdorf]

I thought you wanted 36V operation.....?

View attachment 197320

Yes, for N-Mosfet i want 36V operation, but for Mosfet gate drive UCC27511, I think it is enough Vcc 5V?

 

[BradtheRad]

Simple illustration using 36V supply to turn on high-side N-device. Shut the same device Off by pulling its bias low via another transistor. The incoming clock consists of 1V pulses, but after all this is just a simulation demo.

 

[KlausST]

Hi,

@Brad: What´s the benefit of HIGH side switching compared to LOW side switching?
I see higher effort in HIGH side switching.

LOW side hardware: A single N-CH logic level MOSFET. (No need for a series resistor, since a microcontroller output is already current limited.
(for sure one may add a resistor and maybe also some MOSFET protection circuitry ..)

Klaus

--- Updated Feb 14, 2025 ---

added:

@valdorf:
In post#5 you did not clarify any of my doubts. You just repeated what we already knew.

 

[BradtheRad]

Brad- What´s the benefit of HIGH side switching compared to LOW side switching?
I see higher effort in HIGH side switching.

LOW side hardware: A single N-CH logic level MOSFET. (No need for a series resistor, since a microcontroller output is already current limited.
(for sure one may add a resistor and maybe also some MOSFET protection circuitry ..)

Klaus

Earlier in this thread the question came up about an H-bridge and having to to with using enhancement Nmos at the high side. (Possibly to use identical devices throughout.)

I should have drawn the entire H-bridge in post #10. The simulation below is the H-bridge showing volt levels between the Nmos. The volt levels alternate close to either supply rail. To achieve this the transistors must be entirely turned On-and-Off.

Changing a few transistors to P-type would simplify operation. Or using a dedicated gate driver. In fact the load is light and might even be compatible with a class A amplifier.

Link which runs above schematic in Falstad's animated interactive simulator:

tinyurl.com/26yzqb9l

 

[Easy peasy]

The IR2110 allows 3v3 inputs and can level shift to well above 36V for an half bridge of mosfets

https://www.infineon.com/dgdl/Infineon-IR2110-DataSheet-v01_00-EN.pdf?fileId=5546d462533600a4015355c80333167e