MOSFET Failure

Hi all,

I’m running an Arduino PWM with FQP30N06L Mosfet to control a proportional hydraulic solenoid valve. My circuit diagram is attached.

The machine has been working fine for 2 days (working around 12 hours per day) on our potato planter. However, this morning when we fired it up the pressure was non-adjustable and at the top of its range. On further investigation, the MOSFET seems to have failed. It is constantly open allowing full power to the coil. The Arduino board and PWM is functioning accordingly. Code is below:

const int analogInPin = A0;  // Analog input pin that the potentiometer is attached to
const int analogOutPin = 3; // Analog output pin that the LED is attached to

int sensorValue = 0;        // value read from the pot
int outputValue = 0;        // value output to the PWM (analog out)

void setup() {
  // initialize serial communications at 9600 bps:
  Serial.begin(9600);
  //Pin 3 - 122 Hz PWM / dither frequency
  TCCR2B = (TCCR2B & 0b11111000) | 0x06;
}

void loop() {
  // read the analog in value:
  sensorValue = analogRead(analogInPin);            
  // map it to the range of the analog out:
  outputValue = map(sensorValue, 0, 1023, 200, 255);  
  // change the analog out value:
  analogWrite(analogOutPin, outputValue);          

  // print the results to the serial monitor:
  Serial.print("sensor = " );                      
  Serial.print(sensorValue);      
  Serial.print("\t output = ");      
  Serial.println(outputValue);  
           
}

Could the FQP30N06L have over heated? What would cause failure of this kind? I might add, you couldn’t touch the mosfet after a few minutes of operation, it was that hot. However, I’m only handling 1.5 amps so I cant understand why this heat would be generated?

Any help appreciated as we have a lot of spuds to plant.

Thanks,
Jim

Coil.pdf (39.6 KB)

Electro-prop valve.pdf (59.4 KB)

FQP30N06L.pdf (726 KB)

This doesn't look to be a logic-level FET, so if you weren't driving hard enough (i.e. with only 5 volts instead of 10), it would overheat.

Edit: Forget that - are you sure it's an L device?

Did it run hot when it was working normally?

Have you replaced it? Did it run hot?

Weedpharma

Hi,
I had to look twice about the MOSFET being logic, but this datasheet from sparkfun confirms it.

When I have been controlling hydraulic valves with PWM, I find that using a switching device that is built to do that specific jobs works the best.
I have used the BTS621L1 from Siemens and it works very well, it is logic level input and is designed to drive PWM coils.
This unit has two drivers in it and is a high-side driver,which is how most hydraulic electronics prefer to operate, rather than low-side like your config.

Hope this helps… Tom… :slight_smile:

BTS621L1_SmartFET_Siemens.pdf (184 KB)

Hi,
You have set the PWM up to 122Hz, that is fine if it works, but it is not the dither frequency.
Dither is another property of PWM hydraulic drive profile.

If your valve is working okay, then don't bother trying to use dither.

I see you are only working in the top 200 to 255 duty, so the lack of dither is not important.

The spec sheet mentions it in case you have valve shudder or starting from low duty cycle problems.

Can you post a picture of your layout please? So we can see how your gnd and power circuits are layed out.
You seem to have setup your system well, just needs some reliability.

Thanks ..Tom.. :slight_smile:

Could it be a problem with back-EMV from the coil?
Or does the internal diode in the FET handle that?

Thanks for the replies chaps.

Whandall, that is a good point. Would it be worth fitting an external diode?

When the setup was working without any issues the FET was always extremely hot. Let me put it this way, if I was to touch it then I'd have known about it, any moisture of my fingers with hiss as it burns off.

TomGeorge thanks for the comment. I originally controlled this valve with a BTS621L1 high-side driver. Initially it worked well, but after 4 or 5 minutes the solenoid would begin to jump erratically causing massive pressure spikes which left the machine unusable. On trying the FQP mosfet and switching low-side this problem went away and my controller worked flawlessly up until now. A PWM frequency of 122Hz was used with both the original BTS and FQP. I am working on super-imposing a PWM dither frequency, but I need to sort this problem first.

The tractors still out working away from the yard now, but with manual adjustment using the original hand operated valve. When its back later I'll try and get a few pictures.

Surely the FQP shouldn't get hot enough to burn?

When the setup was working without any issues the FET was always extremely hot.

Then it was too hot.

Surely the FQP shouldn’t get hot enough to burn?

Why not? All power switching components can get too hot, even without design / implementation mistakes.

Have you got a schematic of how you wired it up?

Adding a diode directly to the coil to short the back-EMV would be a good idea IMHO.
Whether the FET could be killed by such spikes, I don't know.

The body of the MOSFET is not particularly large so it does not take a lot of power to get it hot.

With PWM there are many transitions between on and off. This time is where the R changes from low to high while there is a current flowing. This is what generates the heat. The faster the transition, the lower the heat.

The frequency determines the number of transitions through the R change. Lower F may assist.

Also the charge and discharge time of the gate capacitor determines the length of time of the transition. You may be able to speed up the discharge by having a R from the gate to ground.

Weedpharma

You may be able to speed up the discharge by having a R from the gate to ground.

I don't think any resistor between gate and ground would be any improvement over the current sinking capacity of an Arduino PWM pin.

Grumpy_Mike:
I don't think any resistor between gate and ground would be any improvement over the current sinking capacity of an Arduino PWM pin.

Correct if there is a direct connection from pin to gate. If there is a series R of any real amount (k's) it may make a difference but probably negligible as suggested.

Weedpharma

The two things that will cause it to overheat are either not driving it all the way into saturation or spending too long in the linear portion of the switching (assuming the Rds is spec'd low enough etc.). Without having a scope to look at it, it is a tough call as to which (or possibly both) are the issue, but it should NOT be running that hot. It is also a good idea as mentioned to have the flyback diode across the coil and not rely on the diode in the FET to handle the back emf spikes. It always surprised me how much capacitance is associated with the FET gate - one tends to think in terms of "it doesn't take any current to turn it on unlike a bipolar transistor, what is the down side?" but that gate capacitance can kill you - especially at faster switching speeds (that is why there are all the special drivers circuits out there).

If there is a series R of any real amount (k’s) it may make a difference

Well if there is then that in itself can cause the problem with the turn on time being longer. That is why I, and many others, are asking for a schematic.

Heatsink on the fet is also an option - heatsinks are cheap, and high power fets are frequently mounted on heatsinks.

Or you could just need a beefier fet! What current does the device you're driving pull? You generally don't want to push fets close to their limits.

In any case, I would definitely install a diode to clamp the back EMF - always do that for inductive loads.

I've already attached a circuit diagram and coil spec sheet in my original post guys. It really is a basic circuit as you can see from the schematic.

From what I understand the FQP30N06L is logic level as the noted gate voltage supply in the datasheet says 5v. However I am still not convinced that I am driving this right or the mosfet is even right? If I increase the PWM frequency to say 900Hz the FET gets even hotter much quicker. This leading me to think that its something to do with switching losses and too higher a transition period as suggested by weedpharma and GPSMikey. However I am no expert on this, so I'm sure someone will be able to advise me further.

From the datasheet the FQP should be well up to the job with 32amp 60volt capability. RDS(on) is listed as 35mOhms with a VGS of 10v or 5v. I'm using the 5v from the PWM on a 12v circuit pulling no more than 1.5 amps. Like I say though, I'm no expert on whether this MOSFET is suitable.

I do have an Oscilloscope, but to be honest I'm unsure of what to look for.

Thanks,
Jim

Look for a shallow slope on the gate

Hi,
You have to realise you are driving an inductor, not a resistor, have you checked the device I suggested?

Tom... :slight_smile:

Jim_cliff:
I've already attached a circuit diagram and coil spec sheet in my original post guys. It really is a basic circuit as you can see from the schematic.

From what I understand the FQP30N06L is logic level as the noted gate voltage supply in the datasheet says 5v. However I am still not convinced that I am driving this right or the mosfet is even right? If I increase the PWM frequency to say 900Hz the FET gets even hotter much quicker. This leading me to think that its something to do with switching losses and too higher a transition period as suggested by weedpharma and GPSMikey. However I am no expert on this, so I'm sure someone will be able to advise me further.

From the datasheet the FQP should be well up to the job with 32amp 60volt capability. RDS(on) is listed as 35mOhms with a VGS of 10v or 5v. I'm using the 5v from the PWM on a 12v circuit pulling no more than 1.5 amps. Like I say though, I'm no expert on whether this MOSFET is suitable.

I do have an Oscilloscope, but to be honest I'm unsure of what to look for.

Thanks,
Jim

Put a fast Schottky diode from Source to Drain. Cathode (banded side) to Drain. Every time the PWM cycle turns off, all of the energy stored in the coil is being dissipated in the body diode of your MOSFET. That diode is not very efficient. so it heats up. Use something like a 1N5822 (3A 40V Schottky Fast Recovery Diode).
Also drive the Gate harder. The Gate on a MOSFET acts as a capacitor. You have to charge it up and discharge it. The faster you can drive it to it's saturation region, the more efficient it can pass current. They specify the capacity 32A at 10V gate. But, your coil wants the most current when you turn it on (gate voltage lowest, least efficient) .

Try this circuit:

It uses 2 transistors(Q1,Q2) to form a totem pole driver. C1 supplies the current for the 'ON' switch (Q2) and Q1 drives the 'OFF' signal. The two Schottky diodes redirect the 'flyback' current away from the MOSFET.

Chuck.

Hi,
A lower frequency may help, I'm surprised the data sheets don't suggest a PWM frequency.

Tom..... :slight_smile: