PWM Signal wire length

Other Hardware General Electronics
1

For a couple of year I have been using an Arduino controlled 18" fan to ventilate my greenhouse. It is quite simple using using a temperature sensor and a N-channel MOSFET. The fan is driven at 13kHz with variable duty based on temperature and draws 4 amps @ 12v. The arduino is is currently about 12 feet from the fan with the MOSFET placed quite close.

The MOSFET has a passive heat sink which get rather warm (no suprises there !). A more elegant solution would be to mount the MOSFET/heatsink in the air stream from the fan. The arduino needs to be in its current location in order to get the mean temperature of the greenhouse and operate some other sensors.

Should I use a shielded cable to extend the PWM signal ?

2

it can't hurt although if you have a good cable 12 feet is not that far.
You could have multiple pairs in the cable and ground the ones you don't use.
don't mix the 12V fan power supply with the signal.

3

It shouldn't if it's a logic level mosfet.
Which Arduino (3.3volt or 5volt logic).
Do you have a kickback diode across the fan, and which one (should be Schottky).
Leo..

4

This may come from a too high frequency or too slow rise/fall times caused by the wire length or in detail from a poor choice of MOSFET.

5

If it's a logic-level FET being directly PWM'd by an arduino it's likely to warm up. The lower gate voltage that allows direct 5V drive means higher gate capacitance which in turn means slower turn on/turn off times.

Slowing down the PWM frequency should reduce the heating.

2 Likes
6

A couple of such MOSFETs exist, with different characteristics.

7

I use a P40NF10L logic level with 100A Schottky diode across the fan

8

Warm isn’t a problem.

Always show us a good schematic of your proposed circuit.
Show us a good image of your ‘actual’ wiring.

9

When I first made the project I tested the fan using a programmable PWM generator. The 13kHz I ended up with was a compromise between power consumption and noise (growling). If I drop the frequency I will put a capacitor in the circuit to slow the rise time down of the pwm

10

With climate change it is… throwing away energy…

11

Humm

1200×628 58 KB

1 Like
12

:innocent: …

13

Not much to draw really.
Project has been running for 2 years now. The power leads from the battery to the fan are 8 gauge and are 8ft. conductors are seperated by 4". The distance from the adruino to the MOSFET gate is about 2", at the moment

14

Then it shouldn’t take too long to draw one.

We really have no idea how the Arduino output connects to resistors or their values.

Hopefully you have no capacitor in the gate circuit.

1 Like
15

The data sheet says 0.036 Ohm @ 5V Vgs meaning 150mW @ 4A. This is not causing any heat. 3.3V should be sufficient as well. Please check your circuit, best by use of a scope.

EDIT: calculating current and forgot to convert into power :frowning:

Doing so can kill your MOSFET. The analog area should be avoided because it can produce much heat.

1 Like
16

To the fan or Arduino?

17

Which has a loss of 0.5watt with 4Amp DC with 5volt drive.
It should get rather warm without heatsink, but not so with.

Must be switching losses when the fet gets hot with heatsink.
Why did you change from default to 13kHz PWM.
Leo..

18

I wasn't aware of a default frequency for this type of fan. In its original use it is not PWM controlled.

Fan
Fan768×596 76.8 KB

19

That's in a static state. Also need to look at the resistance as that Vgs is switching between 0-5V.

20

I was referring to the default PWM frequency of an Arduino pin.
Which is 500Hz on pin 3, 9, 10, 11 and 1kHz on pin 5, 6 for an Uno R3, Nano v3.
13kHz is a strange choice, and has higher switching losses (heat) than a lower frequency.

Please post a hand-drawn representation of the circuit, with resistor/diode values and wire lengths.
Leo..