Hello


I made a circuit that turns on/off the fan according to the temperature and controls the FAN speed(PWM)


and this is my PCB design




this is 2 layer board and both GND plane

the FAN Speed Control motor is 3A and the Comp FAN motor requires 4 A power

before this circuit, there was no C6, C7, C8 so my mosfet was burn. (when the circuit is running, it momentarily rises to over 8A)

There is no C6,C7,C8 on the PCB under test, so I wired the parts

and it's working now, but C7 and C8 are a little warm

1. How should I place these capacitors? Did I draw the PCB correctly?

2. Can i replace multiple capacitors with one large capacity capacitor? (one 1000uF/50V or two 470uF/50V)

 

Hi,
A few questions and comments about your shematic.
What ist the porpuse of RV1 (especially pin 4 and 5)?
Your schematic symbol of U4 is messed. It is very difficult to understand if the symbol is wrong.

Can you explain what you try to do with the different parts?

C7 and C8 should not get warm.

Which mosfet did get burned?

 

The LM393 outputs require pull-up resistors.

 

RV1 4,5 is switch (switch potentiometer)
RV1 is controlling PWM duty

U4 symbol is wrong, i corrected it now

FDS9435A (Comp Controller) get burned

 

R19 is pull-up resistor. schematic symbol of U4 is wrong sorry my mistake

 

You show R24 going to output of the LM393, that won't work. The LM393 is an open collector output.

 

Sorry for the late reply i changed the circuit a little bit
FET changed to relay







C6, C7 is still warm the temperature of the MOSFET (Q2) is about 40°C i think that's why

Will increasing the value of C6, C7 reduce temperature?

 

Will increasing the value of C6, C7 reduce temperature?

No but increasing the gate drive to Q2 should. What is the Vcc voltage for the NE555?
What's up with D2 and D3? 20 amp stud mounted rectifiers shown reversed bias in the VR & BTN circuit.

 

now NE555 VCC is 5V,
should I add 12v regulator to supply 12v to NE555?

D2 and D3 are used for locus switches. I drew it on the PCB, but I won't actually

 

now NE555 VCC is 5V,
should I add 12v regulator to supply 12v to NE555?
D2 and D3 are used for locus switches. I drew it on the PCB, but I won't actually

Maybe not but adding a 1K resistor from pin3 to Vcc. will increase the drive to appx 5 volts. Specs for Q2 show 3.2mOhm for a Vgs of 4.5 volts.
What are locus switches?

 

this is a rokcer switch

I have rewritten the circuit diagram to make it easier to read



What I'm making is an AC controller. something like this

 

Have you tried adding the 1K resistor to pin3 of the NE555 to increase drive to Q2?

 

Do you mean to increase the R3 value?
but doesn't it take less time to charge the gate driver and make the heat worse?

 

Have you tried adding the 1K resistor to pin3 of the NE555 to increase drive to Q2?

i tested NE555 12V but same. fet gets heat

 

Put the scope on pin2 of the mosfet Q2 and see what that looks like.
Why such a high PWM frequency for the Fan motor? Seems like a couple hundred hz should be sufficient.

 

Do you mean to increase the R3 value?
but doesn't it take less time to charge the gate driver and make the heat worse?

No, sghioto means add a 1k resistor between pins 3 and 8 of U4. You can also remove R10 to get a little more drive to Q2's Gate. And no, the less time it takes to charge the gate, the less time it takes to switch, so the less heat dissipated in the MOSFET!

 

No, @sghioto means add a 1k resistor between pins 3 and 8 of U4.

Thanks for that comment but something is still off if increasing the 555 Vcc to 12 volts did not improve the heating problem.

 

Put the scope on pin2 of the mosfet Q2 and see what that looks like.
Why such a high PWM frequency for the Fan motor? Seems like a couple hundred hz should be sufficient.

This is the result pin2 of the mosfet Q2
There seems to be a problem Is it a diode problem?
noise is heard from the motor when the frequency is low

No, sghioto means add a 1k resistor between pins 3 and 8 of U4. You can also remove R10 to get a little more drive to Q2's Gate. And no, the less time it takes to charge the gate, the less time it takes to switch, so the less heat dissipated in the MOSFET!

I got the same result, something else seems to be the cause.

 

\

1. Scope -> Q2(Mosfet's Pin2) Drain
2. Scope -> Q2(Mosfet's Pin1) Base

The gate charging time seems to be long, so i reduced the gate resistance(22R) and removed the pull-down resistance
but I don't think it's resolved

 

i tested PWM Contoller module





i got same result
does this scope waveform mean normal operation? (yellow waveform from FET's Drain)

[Edit]
oh it's very interesting
I opened the actual product


and it's waveform like this


and this controller works very good and there is no heat at all when the 5A fan is operated(about 32°C )