0
\$\begingroup\$

I'm dealing with an N-Mosfet switch circuit to control, through PWM, the temperature of a heater resistor. The mosfet is a IRFS7430-7PPBF. The load is driven at 12V and has a power rating of 80W. Since I want to drive the load resistor with pure unfiltered PWM, I chose a mosfet with a very low RDSon and I didn't care much about the PWM signal carrier frequency. I'm driving the gate with 5V logic signal and I can provide current in excess of 20 mA. I tried to set different values of PWM frequency to drive the gate of the mosfet. At frequencies in range of several tens of thousands of herts (62.5 KHz) the mosfet gets very hot already at 70% duty cycle (3-4 A). Gradually decreasing frequency helps the mosfet to run very cool until no temperature difference can be felt at around 200 Hz. I explain this because of the switching losses in the mosfet which are very high due to a high internal capacitance. However the mosfet is getting far hotter than I initially calculated. Why is this happening? Is it normal? I'm fine running the mosfet at very low frequencies since I want also to be low in EMI emissions. The strange thing that I noticed is that the benchtop laboratory power supply which I am using to power the circuit starts behaving in a very strange way at low PWM frequencies. Even if I am far beyond the power and current limits of the power supply, it starts continuously changing the displayed value of voltage and current, as it couldn't hold those values. Then the C.C. and C.V. (current control and voltage control) lamps start blinking. These behaviour does not happen at high PWM frequencies. I tried to investigate whether a filter capacitor across the power terminals would help but it does not seem to produce any effect. I'm afraid that the circuit is producing some dangerous disturbances back to the power supply. How could I solve this?

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • 1
    \$\begingroup\$ Do you have a dummy ground, a 1K ohm or 10K ohm resistor from point 'PWM' to ground? The power supply has no load to stabilize it. The mosfet is not a load, except maybe a few 1,000pf. The mosfet is not a dc load, so you need to add one. This is not a complete answer because I do not know the layout of your circuit. \$\endgroup\$ – Sparky256 Apr 26 '16 at 17:55
  • \$\begingroup\$ What is your actual question? \$\endgroup\$ – Andy aka Apr 26 '16 at 17:56
  • \$\begingroup\$ You do not need a Flyback diode unless your load is an inductor or transformer. \$\endgroup\$ – Sparky256 Apr 26 '16 at 18:04
  • \$\begingroup\$ Where is your PWM signal coming from? What level is it? This FET has a threshold voltage of up to 3.9V, so not a good logic-level FET. Also due to high gate capacitance a good FET driver would be necessary to run it at higher PWM frequencies. \$\endgroup\$ – John D Apr 26 '16 at 18:42
  • \$\begingroup\$ I edited my question with some extra details. The pwm is coming straight from a microcontroller with 5V logic, which can provide current in excess of 20mA. I do have a flyback diode because the load might not be pure resistive since I am using a specific resistor. \$\endgroup\$ – Francesco Apr 26 '16 at 20:19
1
\$\begingroup\$

You probably have two separate problems.

First, you do not specify your PWM amplitude or your gate drive capability. From the data sheet the input capacitance is about 14,000 pF. If your PWM logic has a current limit of, let's say, 10 mA, a PWM edge will take 3 or 4 usec to transition from 0 to 5 volts. At 62 kHz, that represents a total period of about .3 to .5 second per second when the FET is not fully on or off (62000 x 2 x 4 us = 496 msec). So it's no wonder the FET gets hot. Try driving your gate with a dedicated gate driver. Simply because I'm familiar with them, I like to use the MAXIM 4426/4427 drivers, but there are plenty of alternatives.

Second, at very low switching frequencies, you may well be getting interference with your supply's control loop. These typically have kHz or less bandwidth, and if you hit them just right you may be getting into a resonant condition. In this case, you are best off using an inductive low-pass filter on your heater.

\$\endgroup\$
  • \$\begingroup\$ Is it dangerous to drive the mosfet directly from the microcontroller at 5V logic? My MC can provide current in up to 20mA. I actually want to drive the heater at low frequency to limit EMI and avoid complexity. \$\endgroup\$ – Francesco Apr 26 '16 at 20:23
  • \$\begingroup\$ It's not dangerous in the usual sense. The problem is that the gate voltage may not be adequate to drive the FET full on. \$\endgroup\$ – WhatRoughBeast Apr 26 '16 at 21:17
  • \$\begingroup\$ What about the power supply? Is there a risk to have it damaged? What if I use an external power adapter? Can I trust that my circuit want break the adapter with stressy spikes and resonant frequencies? What should I take care about? Thanks. \$\endgroup\$ – Francesco Apr 27 '16 at 7:48
1
\$\begingroup\$

You have to choose a different MOSFET.

When it comes to PWM applications, too "big" a MOSFET leads to unnecessarily inefficient switching. Your application is to switch up to around 7A. The IRFS740 is a 500A MOSFET. The gate capacitance is such that there is no way an uC can directly drive it to switch efficiently at 10's of KHz. I think you should be able to find a MOSFET that can be driven by uC at up to a few 10's of KHz effectively (although not necessarily most efficiently) when balanced against conduction loss due to Rds-on. I am guessing the sweet spot could be at 2 to 3 times the current, so you could start your search around a 15-20A MOSFET.

Also, don't over spec the voltage. That just leads to higher capacitance and higher Rds-on with almost no side benefit.

If you look at the IRFS740 datasheet, pay attention to figure 1. That shows a very steep rise of Rds-on at around Vgs of 5V. That means the IRFS740 is not going to give you reliable results when driven by no more than 5V. You need a MOSFET that has Vgs-th probably no higher than 2.5V. (The datasheet Vgs-th is typically specified at very low conduction current.)

\$\endgroup\$
  • \$\begingroup\$ I actually want the mosfet to run at low frequency since I need low EMI. I wanted just to investigate why was the mosfet behaving that way at high frequencies. On the other hand I need the mosfet to be fully saturated at 5V. \$\endgroup\$ – Francesco Apr 27 '16 at 7:45
0
\$\begingroup\$

Do you have a dummy ground, a 1K ohm or 10K ohm resistor from point 'PWM' to ground? The power supply/pwm has no load to stabilize it. The mosfet is not a load, except maybe a few 1,000pf.
The mosfet is not a dc load, so you need to add one(the resistor). You must have a capacitor from your Vcc terminal to the mosfet ground, with leads as short as possible. A value of 470uF should be enough. I am not sure this is a complete answer because I do not know your circuit layout.
Do NOT insert a voltage or PWM signal into the mosfet >12p-p, or the built-in zener will clamp the signal and may get hot. If getting hot, this mosfet should be mounted to at least a small heatsink. I understand your experimenting and this is not to be a final product.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.