[Help] IRF540 Heating up on Piezo Atomizer

Thread Starter

Boplogz

Joined Oct 17, 2022
22
Hi everyone,

I'm seeking advice on how to lower the temperature of my MOSFET in driving the Piezo humidifier.

I have the circuit of my driver as shown below.

1679032440101.png

The input of the MOSFET is 5.20 Vp-p Sine wave at 1.7Mhz.
The Piezo has 1800pF Static Capacity and <2.5 Resonant Impedance.

When the piezo is not connected in the circuit, the power supply shows that the circuit draws 0.08A. But when the piezo is connected the bench power supply current increases to 0.82A and the MOSFET is heating up (when you touch it, you will be likely burn your finger).

I tried changing the input on MOSFET and changing the inductor to 220uH and 470uH but no effect. Still the MOSFET is heating up.

Is the heat that high normal or not?

Hoping for simple solution to lower the temperature of the MOSFET.


Thank you.
 

Attachments

Ian0

Joined Aug 7, 2020
10,276
You are currently operating it in the linear mode in Class-A - it WILL get hot!
You need at least 6V to turn an IRF540 fully on. You circuit also suggests you might be driving it with a sinewave.
 

Thread Starter

Boplogz

Joined Oct 17, 2022
22
You are currently operating it in the linear mode in Class-A - it WILL get hot!
You need at least 6V to turn an IRF540 fully on. You circuit also suggests you might be driving it with a sinewave.
I see, I will try to increase the input to 7V and will give feedback. By the way, if not much, would you mind explaining about the sine wave input? It is not good or bad idea?


[EDIT - UPDATE]

I tried varying the input voltage and at 10Vpp input and still works, I mean it humidify. However, the temperature of the MOSFET is still hot.
 
Last edited:

StefanZe

Joined Nov 6, 2019
191
A sine wave will operate the MOSFET in the linear region for a longer time compared to a square wave. Thats because the gate voltage rises a lot slower to the threshold voltage in a sine wave.

There could also be another problem. The gate has a capacitance that your driver needs to charge and discharge with 1.7MHz. This will need a lot of current from your gate driver. How do you drive the gate?
 

Thread Starter

Boplogz

Joined Oct 17, 2022
22
A sine wave will operate the MOSFET in the linear region for a longer time compared to a square wave. Thats because the gate voltage rises a lot slower to the threshold voltage in a sine wave.

There could also be another problem. The gate has a capacitance that your driver needs to charge and discharge with 1.7MHz. This will need a lot of current from your gate driver. How do you drive the gate?
I see, I will try to work on Square Wave.

I used TC4427 Driver chip.
 

StefanZe

Joined Nov 6, 2019
191
The TC4427 input should not be driven by a sine wave. Drive the chip with a 5V square wave (0V low, 5V high).
Use a supply voltage of 10-15V for the TC4427.
 

Ian0

Joined Aug 7, 2020
10,276
The TC4427 input should not be driven by a sine wave. Drive the chip with a 5V square wave (0V low, 5V high).
Use a supply voltage of 10-15V for the TC4427.
A TC4427 has a Schmitt trigger input. Driving it with a sinewave is not a problem (maybe not the greatest idea, but not a problem as the output will still be a squarewave).
Also check that your inductor is not saturating, and that you are not generating so much voltage in the drain of the IRF540 that it is avalanching.
 

StefanZe

Joined Nov 6, 2019
191
A TC4427 has a Schmitt trigger input. Driving it with a sinewave is not a problem (maybe not the greatest idea, but not a problem as the output will still be a squarewave).
I know. You still should not do it and it will mess up the duty cycle.
 

Thread Starter

Boplogz

Joined Oct 17, 2022
22
Thank you for your inputs. I will try to do the square wave and will feed back as soon Im able to update my circuit.

I will try to go back and learn inductor saturation and the mosfet avalanche.

Thank you once again.
 

Ian0

Joined Aug 7, 2020
10,276
Thank you for your inputs. I will try to do the square wave and will feed back as soon Im able to update my circuit.

I will try to go back and learn inductor saturation and the mosfet avalanche.

Thank you once again.
If you look at the Drain voltage on a scope, and it looks “flat topped” at just above 100V, then it has avalanched.

If you look at the source current (you’ll need a shunt resistor) and it doesn’t look like the simulation. It should be either be triangular, or semi-sinusoidal (as it is a resonant circuit) But If some peaks start to appear it’s saturating.
Another clue if it saturating - the inductor gets hot.
 

Thread Starter

Boplogz

Joined Oct 17, 2022
22
If you look at the Drain voltage on a scope, and it looks “flat topped” at just above 100V, then it has avalanched.

If you look at the source current (you’ll need a shunt resistor) and it doesn’t look like the simulation. It should be either be triangular, or semi-sinusoidal (as it is a resonant circuit) But If some peaks start to appear it’s saturating.
Another clue if it saturating - the inductor gets hot.
Uhmm if say for example, it is saturating. What changes should I will make?
 

Thread Starter

Boplogz

Joined Oct 17, 2022
22
Wind it for lower flux density (use a core with more cross sectional area)
Hi, an apology for the late reply. I was having a hard time changing it to square wave.
I change it and still heating up.

I also verified that it is not avalanche and saturated. Maybe I ll try lowering the inductor to 1uH.
 
Top