Logic Level N-channel MOSFET question

[qwwwizx]

Hi all,

My first post on this forum, - and I am a little rusty in electronics so my question here might be n00bish.

I am messing with pulsewidth modulation for different purposes, and for that I just changed from using normal transistors / FETs to logic-level FETs. My problem is that I thought I could just shift my NPN Darlington out with this Logic Level N-Channel FET to get proper ON / OFF pulses, using a simple IC.

https://pdf1.alldatasheet.com/datasheet-pdf/view/182048/FAIRCHILD/RFD14N05L.html

My problem is that the transister seems to be always ON. I can even disconnect the wire to the gate pin and it will still be ON. After looking in the PDF, it looks like there is a diode between source and drain. How am I suppose to use that transistor, and can it be used for this at all?

Any help will be appreciated.

 

[crutschow]

To turn off a MOSFET you need to remove the large gate charge by connecting the gate to the source with a low impedance connection. MOSFETs have a high gate capacitance (see transistor data sheet) and require a low impedance source to rapidly charge and discharge this capacitance for fast switching. Whether its a logic-level type or not has no affect on that requirement.

The diode between the drain and source is reverse biased under normal operational bias voltages so will not conduct.

 

[qwwwizx]

Thank you,

It seems to solve the problem. So far I just equipped a 10 ohm resistor and it can now shut off. I need a little more experience with this, so I might get back later on. From your signature "--Inside every little problem is a big problem trying to get out.--"

 

[jpanhalt]

If you are driving it directly with a PIC, 10Ω is a bit small: 5V/10Ω = 0.5A.

Even 100 Ω is a bit small. I suggest at least 200Ω, and you will probably find that 1000Ω will work just fine.

John

 

[qwwwizx]

I was thinking about asking that question/ on how low is "low impedance"? so I am happy you brought it up.

I will try my way with 200-1k tomorrow.

BTW: While playing around with it, I use ne555 and ne556 timers to do some PWM modulation. I use the output to drive the gate on the MOSFET. Since the 555 lags current for a real mosfets, I use the logic level gate MOSFETs. I want the transistor to be fully OFF, or fully ON (saturated in technical terms, I believe).

 

[dr pepper]

You can get purpose made ic's for driving the gates of fets, however in my experience such ic's are sensitive to board layout, you can use a pair of transistors to speed up switching of the gate, if your load isnt so great and switching frequency isnt so high you might not need them.
If you drive a mosfet gate directly from a pic micro beware of read modify write issues caused by transition lag.

Logic level mosfets are a form of fet where the junction will saturate at a lower voltage, some power fets will not saturate fully with 5v on the gate, whereas a logic level fet will, they are popular for automotive applications.

 

[crutschow]

..................................

BTW: While playing around with it, I use ne555 and ne556 timers to do some PWM modulation. I use the output to drive the gate on the MOSFET. Since the 555 lags current for a real mosfets, I use the logic level gate MOSFETs. I want the transistor to be fully OFF, or fully ON (saturated in technical terms, I believe).

A 555 has enough output push-pull current to directly drive low current MOSFETs with fairly good switching times. You don't need any resistors for that.

If you use 10V or more for the 555 power supply you can use normal gate level MOSFETs. (I don't understand what you mean by "the 555 lags current for a real mosfets.."? )

 

[qwwwizx]

crutschow:
Well, I might be wrong about it. My experience is limited, so I am still learning.

I use 5v to run my ne555, but will change it to 12v. What I mean by the messy sentence is, that I read, that a power MOSFET will require up to 1 Amp to be fully saturated at high frequency. The ne555 should be able to push around 200ma which I thought as a limit. My hope was to reach a PWM frequency between 100khz - 1Mhz, but have no idea if that is considered a high or low frequency? The ne555 might be my limit here? The best I have seen on my 555 is around 300Khz so far, but havent got a clue if its useful at that frequency.

Earlier I used a BD139 power NPN which seemed to not open completely with 2-4v at the output pin of my ne555, which made me try the logic gate transistor instead. The logic gate transistor gave me some other issues, already solved, thanks to you.

 

[ARandomOWl]

The current that you drive the gate with does not determine the current the MOSFET will pass through the Drain-source. It is the voltage on the gate that determines how much current passes through the drain-source. The more current you use to drive the gate, the faster the gate capacitance will charge and the faster the MOSFET will turn on.

 

[crutschow]

At 5V the RFD14N05L MOSFET has a gate switching charge of 25nC. 200mA drive will switch this charge in about 125ns. That's probably OK for switching frequencies up to a few hundred kHz or so. I would not try to use a 555 about a couple hundred kHz.

 

[audioguru]

The maximum output high from an NE555 that has a 5V supply is only +2.75V to +3.7V which might not be enough to fully turn on a logic-level Mosfet.

 

[qwwwizx]

I equipped a 470ohm resistor to the transistor and it still works great. Furthermore I upped the voltage to 12v. I can see a voltage drop over the transistor.

The maximum output high from an NE555 that has a 5V supply is only +2.75V to +3.7V which might not be enough to fully turn on a logic-level Mosfet.

I think some of my earlier issues has been related to these 2.75v - 3.7v. Now that I upped the ne556 voltage to 12v, everything seems to work.

 

[qwwwizx]

I got 11.5v to feed the load now, when PWM is on full duty cycle with a little fan equipped. It drops further to as low as 9.5v when I load it with 4 fans (12v/400ma). My PSU is an old ATX 350W, so I should have plenty of power. Can it be the breadboard suffering, or is it related to the transistor? I mean, I should be able to hold it on 11.5v even adding a few fans to it.

 

[ARandomOWl]

Do you have a schematic. It is impossible to advise without one.

 

[qwwwizx]

I dont know if its enough:

Its more or less like this - the 556 version:

View attachment 66223
https://www.overclockers.com/pwm-fan-controller#

R1: 1K
R2: 10K
C1: 220pF
C2: 10nF
C3: 100nF
C4: 10uF
P1: is a series of resistors starting from 1k to 150k or something where I can adjust, since I dont have a POT yet.


Mine is adjusted a bit for my purpose. I give it 12v, and my PWM out is driving my Logic Level gate N-channel transistor.

All components are using the same +12v and GND. I have a 470ohm resistor between the source and gate. The transistor is just equiped to open and close based on the PWM signal.

 

[ARandomOWl]

You missed the most important part. Where is the load connected?

 

[qwwwizx]

My first official drawing, but I think it should show how I connected it:
View attachment 66225

The Fans are the only load equipped so far.

 

[ARandomOWl]

Can you add component values to your schematic? Or at least give the value of R2.

Also a tip. Draw your most positive voltage rail at the top

 

[qwwwizx]

I added the values below the first schematic now. I should have done it from the start like that.

PS. Now updated with the pot information too.

Thank you for the tip about drawing, I was thinking a lot about how to draw it without too many lines crossing each other.

 

[audioguru]

The 470 ohm resistor from gate to source is stealing away some gate drive voltage. Remove it.
It can be 10k ohms if you want the Mosfet to be turned off when the power supply is turned off.