2 MOSFETs Switch application

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
I was recently asked to use 2 mosfets to make a switch that turns a fan on/off via a PWM pin of an stm32 (this allows a gradual switch-on of the fan and not all at once .. here).
I cannot understand why I need two and not one mosfet.

Maybe it's because having a GPIO PWM [0 ... 3.3]V output that must drive a mosfet with Vgs(th) of 2V or more ... there is a risk of not switching the mosfet on .. and thus I need an amplification stage (to be implemented between GPIO PWM and the mosfet that switches the fan on/off).
Is this correct?

Keep in mind that I would like to use a through-hole mosfet package for better dissipation (I will mount it on a PCB) ... than perhaps other types of packages
 

ronsimpson

Joined Oct 7, 2019
2,985
must drive a mosfet with Vgs(th) of 2V or more
Look at a data sheet for the details of what "Vgs(th)" is.
1) There are often two or three numbers. Minimum/typical/max On the cover of the data sheet it might say 2V but in the details it might say 1.5/2.0/2.5 and you don't know the exact number.
2) Vgs(th) is when the part just thinks about turning on. Look at the conditions! Often the D current is 10uA. (micro A) This will not run the fan.
3) Look at the graphs of Vgs verses current or verses Rds on. For a 1A fan (start up current) you might need 4V of Gate voltage.

Look for a "logic level" MOSFET.
 

crutschow

Joined Mar 14, 2008
34,280
What is the voltage value, type (AC or DC), and current the fan requires?

For a 3.3V control signal, you would need a logic-level type MOSFET with the on-resistance specified for a Vgs of 3.3V or less.
The usually means the Vgs(max) threshold is <2V.
 

Ian0

Joined Aug 7, 2020
9,667
Here's some information on pwm control of fans:
https://www.analog.com/media/en/ana...umber-1/articles/how-to-control-fan-speed.pdf
but you would be well advised to consider @MaxHeadRoom 's sage advice in post #14 here:
https://forum.allaboutcircuits.com/threads/controlling-speed-of-12vdc-fan.169773/
I would not be surprised to find a decoupling capacitor in electronically commutated fans, and if it is present the MOSFET will cause a big current spike as the capacitor is charged with each PWM pulse.
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
Look at a data sheet for the details of what "Vgs(th)" is.
1) There are often two or three numbers. Minimum/typical/max On the cover of the data sheet it might say 2V but in the details it might say 1.5/2.0/2.5 and you don't know the exact number.
2) Vgs(th) is when the part just thinks about turning on. Look at the conditions! Often the D current is 10uA. (micro A) This will not run the fan.
3) Look at the graphs of Vgs verses current or verses Rds on. For a 1A fan (start up current) you might need 4V of Gate voltage.

Look for a "logic level" MOSFET.
When I'm looking for a logic-level MOSFET, do I just search for an N-Channel 3.3 Vgs MOSFET or does the datasheet have to explicitly say "logic level"?

I ask because I googled "logic level mosfet" and the Mouser website took me to the category "N-Channel 3.3 V MOSFET".

@crutschow
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
What is the voltage value, type (AC or DC), and current the fan requires?

For a 3.3V control signal, you would need a logic-level type MOSFET with the on-resistance specified for a Vgs of 3.3V or less.
The usually means the Vgs(max) threshold is <2V.
fan: 24V 7A
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
Here's some information on pwm control of fans:
https://www.analog.com/media/en/ana...umber-1/articles/how-to-control-fan-speed.pdf
but you would be well advised to consider @MaxHeadRoom 's sage advice in post #14 here:
https://forum.allaboutcircuits.com/threads/controlling-speed-of-12vdc-fan.169773/
I would not be surprised to find a decoupling capacitor in electronically commutated fans, and if it is present the MOSFET will cause a big current spike as the capacitor is charged with each PWM pulse.
Very interesting the guide you sent me. Thanks!

I would like to implement the low frequency PWM technique but it is still not clear to me why I would need two mosfets and not just one
 

ronsimpson

Joined Oct 7, 2019
2,985
I was looking for a MOSFET, through hole to-220 that will do 7A 30V. Many of the new parts are not included because they do not come in a to-220.
Here is a RFP15N05L rated at 15A, 50V and built for "TTL 5V" drive.
Look at the graph. At 3.0V drive the transistor tears open at 7A. It really wants 5V of Gate drive.
1667145036202.png
With 0.17 ohms RDSon it will run HOT. You need a higher current part. The surface mount parts may have lower Gate drive voltage.
 

ronsimpson

Joined Oct 7, 2019
2,985
If you give up on through hole here is a new low drive voltage part.
You can see at with 2.0V on the gate it is weel turned on to 20A. This one is built to drive from 3.0V.
SiS414
1667147185194.png
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
Where does it say you need 2 MOSFETs?
Where does it say you need 2 MOSFETs?
I think one mosfet is enough and I have already opened another thread asking for advice.
Whereas I only opened this thread because of the issue of the 2 mosfets which I don't quite understand.

We were told/advised by our employer to use two, but I still don't understand why.
 
Last edited:

Ian0

Joined Aug 7, 2020
9,667
I think one mosfet is enough and I have already opened another thread asking for advice.
Whereas I only opened this thread because of the issue of the 2 mosfets which I don't quite understand.

We were told/advised by our employer to use two, but I still don't understand why.
If we don’t know where the second one is supposed to be connected, we can‘t tell you if we think it is necessary!
 

Irving

Joined Jan 30, 2016
3,843
Whereas I only opened this thread because of the issue of the 2 mosfets which I don't quite understand.
I'd go back and ask why - potentially its being suggested to run 2 in parallel to reduce power dissipation in a single device, depending on environment. A TO-220 device without a heatsink can theoretically handle 2W dissipation. A 7A fan on at 100% duty cycle is trending towards 1 Watt, in open surroundings, but in a casing with minimal air flow it could be marginal hence the suggestion to use 2?

The Si414 SMD device is again spec'd at 2.2W max dissipation on a 1" x 1" square FR4 circuit board and at 3.3v Vgs and 7A is again at trending towards 1.1W (case temp of 70degC) which could again be seen as marginal for a single device on a crowded 2-layer board.
 

ronsimpson

Joined Oct 7, 2019
2,985
Two different ways to use two FETs.
As Irving said, parallel two.
or
Use a very small part up front to get good drive voltage. Now you have much more options for the high-power part.
Note R1, R9 is used to reduce the voltage to M2 down to 12V. Most MOSFETS can only take 20V max.
1667220787484.png
IPP065N04NG
2SK669-AC
 

Irving

Joined Jan 30, 2016
3,843
Two different ways to use two FETs.
As Irving said, parallel two.
or
Use a very small part up front to get good drive voltage. Now you have much more options for the high-power part.
Note R1, R9 is used to reduce the voltage to M2 down to 12V. Most MOSFETS can only take 20V max.
View attachment 279615
IPP065N04NG
2SK669-AC
The difficulty with that is that it inverts the logic. 'low' = ON.

An alternative, non-inverting, solution is to use a gate driver, such as TI's UCC27531 in a SOT233-6 case. It can provide gate currents up to 2.5A source & 5A sink. A small zener diode protects the gate against over-voltage if needed.

1667239702418.png
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
If you give up on through hole here is a new low drive voltage part.
You can see at with 2.0V on the gate it is weel turned on to 20A. This one is built to drive from 3.0V.
SiS414
View attachment 279579
The behaviour of the circuit is as follows:

1) for the first few seconds I want the current on the fan (drain current) of about 4A using a PWM=50% (in fact from the GPIO Vgs=3.3V/2 = 1.65V) .. this prevents an initial current spike in the fan
2) then I set PWM=100% (so Vgs = 3.3V) so the maximum current of 8A of the fan is "released"

Looking at the graph, I notice that when 2) happens the Id(max) is about 30A.. perfect!

Can you explain to me what happens in the Vds-Id graph that you attached when I'm in the condition 1) ??
In this case I have difficulty understanding the graph because there is no curve for Vgs=1.65V ... but confirm me that 4A on the fan is below the Id(max) limit with Vgs=1.65V
 

ronsimpson

Joined Oct 7, 2019
2,985
1667507587916.png
When you have 50% duty cycle 4A, it is likely 4A average, 8A peak. So the current is not really reduced.
What happens when you don't have enough drive voltage?
Using the above graph, VGS=1.5V, at near 0A there is near 0VD-S. As you apply more current, the VDS goes up. At about 3A the transistor can not keep "on" and it tears open. The voltage D-S jumps up to 2V at about 3A. Increase the current to 3.01 and it might have your full 24V D-S. This is not a good way to use the transistor.

Look at 20A for a example. AT VGS=1.5 you will never see 20A. At 2V Gate you will have 0.5V VDS. At 2.5V Gate you will have 0.3 VDS. Above 2.5V there is little advantage of more Gate Drive at 20A.
 

Ian0

Joined Aug 7, 2020
9,667
Above 2.5V there is little advantage of more Gate Drive at 20A.
True if every instance of the said MOSFET had the same Vgs(th).
For the average MOSFET, Vgs(th) can vary from 2V to 4V at 25°C, and then can vary by another half a volt with temperature.
If 2.5V is enough for the "typical" device shown on the graph, then it really needs 4V to cover the entire range of possible values at all possible temperatures.
 

ronsimpson

Joined Oct 7, 2019
2,985
Agree,
What happens is, "with a 7A load we need a 7A MOSFET". The transistor will run HOT at best. You will need extra voltage on the Gate if you are running at max. current.
The graph is for a 60A part, chosen because I know it will work at 7A with low voltage. If we ran it at 60A you need 5V on the Gate. A part like this will run cool.
 
Top