PWM Signal wire length

[Crusty]

I have a 10A schottkey diode installed. Thanks for the advice

[JCA34F]

And another thing about DC clamp meters, (I used them at work before retiring 14 yrs ago), I didn't trust their readings below about 20% of full scale, especially with "wacky" waveforms from DC motor controllers.
Newer technology has hopefully improved them.

[Crusty]

UNIT-T210E might be OK but in any case my conventional meters only go to 10A and I don't want to risk frying on of those as well as a MOSFET

[JCA34F]

If your meter are fused, they should be OK, but buy some spare fuses first (from first hand experience.)

[Crusty]

BTW thank you for taking the time to help me, it is much appreciated. This is something I took up after I retired from a totally different field. If you ever want to know about aqueous amino-plast copolymers I am your man

[JCA34F]

Ill keep it in mind.
BTW, have you ever connected the fan straight to 12V and measured the amps?

[Crusty]

No but as I using a car battery at the moment I will give it a try

[JCA34F]

Please report back, thanks.

[Crusty]

Well it has been a most interesting afternoon and things become a little clearer.

Firstly the fan when conected directly to the car battery draws 20A @12V = 240W !! so the spec quoted of 12V 120W is a gross under estimation.

I did about 20 differnt combinations of frequency and duty but posting them is not really helpful to my conclusions.

A range from 2.0 - 24.0 kHz produced satisfactory air flow (by adjusting duty) when the current draw was 3A-4A at each tested frequency. Which you would expect really.

Passive heat sink temperature varied 35°- 45°C and didn't really correlate to well with frequency, i.e. lower Hz = cooler.

Noise was pretty horrible at the lower frequencies and I could hear it up to 10kHz, higher than that it was not noticable.

At all the frquencies I used the DMM and a clamp meter the amperes correlated very well.

I think one of my heating problems was due to a dying power supply, using the car battery resulted in a significantly lower MOSFET temperature.

The cable length used 12' maximum didn't impact on heating.

I would like to thank everyone who took the time to cotribute to this posting

[JCA34F]

Keeping in mind my reply about affinity laws (post 52), does the fan have the original blade?

[Crusty]

Yes no modifications made to the fan

[Crusty]

Final post.

I decided to use an old ATX2 300W PSU to power this project, as it had a few advantages over my original 12V power supply and I had one lying around. I could use the 5V stand bye to power the Arduino and by adding a little code could switch the PSU on or off to supply the 12V to the fan as required. This is my first attempt at a Fritzing diagram, so I hope I wired the diagram correctly to reflect my actual project.

// Greenhouse Fan control Crusty Sneers

#include <PWM.h>
int fan = 9; // the pin that the fan is attached to
int duty = 50;
int32_t frequency = 13000; // frequency in Hz

#include <OneWire.h>
#include <DS18B20.h>

#define ONE_WIRE_BUS 7

OneWire oneWire(ONE_WIRE_BUS);
DS18B20 sensor(&oneWire);
const int in1 = 8;  //PSU 12V supply on

void setup()
{
  Serial.begin(9600);
  InitTimersSafe(); //initialize all timers except for 0, to save time keeping functions
  bool success = SetPinFrequencySafe(fan, frequency);
  if (success) {

    Serial.println("Success");
    sensor.begin();
    pinMode(in1, OUTPUT); // Switch  PSU ON
  }
}

void loop()
{
  sensor.requestTemperatures();

  while (!sensor.isConversionComplete());  // wait until sensor is ready
  float T1 = (sensor.getTempC());

  Serial.print("Temp: ");
  Serial.println(T1);

  if (T1 < 25.0)
  {

    pwmWrite(fan, duty - 50);
    Serial.println("Fan OFF");
    delay(5000);
  }

  if (T1 > 25.0 )
  {

    digitalWrite(in1, HIGH); // starts PSU from standby for 12V fan
    Serial.println("PSU ON");
    delay(5000); // delay for PSU to stabilise
  } else {
    digitalWrite(in1, LOW);
    Serial.println("PSU OFF");

  }

  if (T1 > 25.1 && T1 < 25.3)
  {
    pwmWrite(fan, duty + 77.5); // start fan at higher speed prevent stalling
    Serial.println("Initialising Fan");
    Serial.println("Fan 50%");
    delay(1000); // delay to get fan to speed before regulating
    
  }
  else if (T1 > 25.3 && T1 < 26.1)
  {
    pwmWrite(fan, duty);
    Serial.println("Fan 20%");
    delay(5000);
  }

  else if (T1 > 26.1 && T1 < 27.0)
  {
    pwmWrite(fan, duty + 26.5);
    Serial.println("Fan 30%");
    delay(5000);
  }

  else if (T1 > 27.1 && T1 < 28.0)
  {
    pwmWrite(fan, duty + 52);
    Serial.println("Fan 40%");
    delay(5000);
  }

  if (T1 > 28.1 )
  {
    pwmWrite(fan, duty + 77.5);
    Serial.println("Fan Maximum");
    delay(5000);
  }

}

[LarryD]

The Arduino will not turn off the MOSEFT.

We do not need Fritzing, we need a schematic.

---

An ATX Power Supply probably needs a minimum load on 5v to get the supply to work properly.

[Crusty]

I did a "proper" conversion and used a 10 ohm 10W resistor with integral heat sink clamped to the power transistors heat sink. I did try to do a schematic with Fritzing, but I couldn't get my head around the programme. I should have just used a pencil and paper and photographed it, I guess.

Thanks for taking the time to look