simple MOSFET question

Hi,

I have a problem reading the datasheet, so I would really appreciate some help.

For example: IRF630

The maximum rating for V(gs) (gate source voltage) is +- 20V
The mosfet will turn on when V(gs)th (gate treshold voltage) is 2/3/4 (min/typ/max)

How to interpret this?

What I assume is that if V(gs) is 0-1.999V the mosfet will be closed (ds resistance infinite in theory) and if V(gs) is between 2 and 4V the mosfet will open (ds resistance <0.5R).

But, the part that I do not get, if the maximum rating for V(gs) is +-20V I assume that if V(gs) is for example 10V, the mosfet will "open" and when V(gs) goes to for example -10V the mosfet will close.

I attached the parts of the datasheet that are confusing to me.

Another thing I would like to confirm (just to know I understand this correctly), mosfet is "open" or "closed", there are no "in between states" like with transistor, when I the potential difference between gate and source is > 2V it will open, and DS resistance is "same" for 2V and for 4V, right? it does not corellate as with transistor (ok transistor correlates base current but this is just analogy)?

thanks in advance

A mosfet dose have a "in between state". Its the same as on a BJT transistor, the harder you drive the gate the more it opens. But its not linear mostly it goes very slowly and then starts going up fast. That mosfet is probably a logic level mosfet because 5V would open it almost full. Typically 10 to 15V is used to drive MOSFETs and IGBTs because you always want to open it as much as you can so you get less heat on it. Also mosfet drivers can mostly supply about 1A of current, true that a mosfet takes no current on its gate, but the gate has capacitance and you want to get the gate from 0 to full in a very short time to again lower heating.

Thanks, so the DS resistance will vary on V(gs) ..
the min/avg/max and "apsolute max" still remains confusing .. what will happen if I get V(gs) to 15V? Will it "open fully" just like with 5V but start to heat or just open or ?

This one I showed up here is just example, I'm still deciding wich one I will try (will be driving them using micro controler so 0/5V or 0/3V) but in order to decide wich one to use, I have to be able to understand the datasheet.

Well 15V will open it even more than 5V and so make it heat even less. But at 5V its supposed to be almost fully open, so its resistance wont be as low as at 15V but it should be close to it.

EDIT: Oh and these logic level mosfets are perfect for driving directly off a 5V microcontrollers pin. But i do recommend using a buffer to provide a bigger current if your switching a very large load very fast (Like PWM for a huge motor) Since the few 10s of mA that the MCU provides will make it switch a bit slow for fast PWM and so make it heat more

Someone Electro said:

Well 15V will open it even more than 5V and so make it heat even less. But at 5V its supposed to be almost fully open, so its resistance wont be as low as at 15V but it should be close to it.

Click to expand...

Thanks, I got it now.

Someone Electro said:

Since the few 10s of mA that the MCU provides will make it switch a bit slow for fast PWM and so make it heat more

Click to expand...

I assume you talk about micro controler "energising" the gate to slow at it will only source some 25mA.. I do not think it will be a problem for now, but good idea with the buffer that can source more current.

This one for example (IRF630) is rated 9A. I will be driving 2-3A max so I do not believe heat will be the problem... I generally wanted to be sure I understand the datasheet, so, the "max" is potential that will "open fully" but more potential will only help.

Thanks for the help

im_noob said:

I generally wanted to be sure I understand the datasheet, so, the "max" is potential that will "open fully" but more potential will only help.

Click to expand...

"max" potential is not "open fully". If you look at the data sheet the Vgs(th) is for a drain-source current of 250µA. For more current you need a higher gate voltage. If you look at the Id((on) and rds(on) values, they are given at a Vgs of 10V. Thus for the lowest resistance "on" operation you should have at least 10V Vgs.

ah, I missed that completely .. thanks for pointing that out ... now the picture is even clearer .. I think I will be able to read the data sheet now correctly and select appropriate device for some testing.. (might not even go fith mosfet but with fet or jfet .. still not sure about the differences but I think now I can compare data sheets with bit more knowledge)

It might help to understand that Vgs(th) is the "threshold" voltage of the MOSFET, that is, the point at which it is just starting to turn on. Because of manufacturing tolerances this can vary from 2V to 4V for the IRF630.

Most power field-effect transistors are MOSFETs. JFETs are usually lower power, small signal devices.

You could also tell what do you want to switch with the mosfet. We might be able to give sugestions what switching devide is best for the job

A good piece of advice that make selecting MOSFET easier is to look at the Rds value on the datasheet, in particular what manufacturer have quoted for the Vgs condition to get this Rds value.

For IRF630, the Vgs quoted is 10V.

So one can usually assume the IRF630 MOSFET works with 10V Vgs normally. Any voltage values lower than 10V will make it "open less" and thus results in a higher Rds value.

eblc1388, that is a good advice, thanks

At the moment, I do not plan to do anything but learn, I find that my head is full of "theory" and zero practice. I know how "field effect transistor" work, what is the "manufacturing difference" between mosfet, jfet, bjt and others, but in practice, I never ever hold anything other then bjt in my hand. So for now, I'm browsing the datasheets, web and bugging some people I know for help.. I friend promissed a demostration in few days (he is also here on forum) so I'll go and actually look at real time scope output first time in my life
So, the question is more theoretical then practical, as I still have no idea what I will switch using the MOSFET or JFET, the "led chaser" will surely not need a MOSFET and that (with some servo controll attempts) is all that I did untill now

carrying on with the theme of this thread, I have a similar problem; I'm trying to PWM some high power LEDs (at about 100-200Hz) (~3V, 800-1400mA, 1-3W) with a microcontroller, and it would be nice to just switch it straight from the uC pins. can you explain a little more about this buffer idea? what I gather from it is using two transistors, one a MOSFET to source current to the LEDs, and another type (BJT?) to source more current to the gate of the MOSFET to charge it faster. also, wouldnt it be a decent idea to get a fairly high-resistance pull down resistor on the gate of the MOSFET (say, 1MΩ) to drain the gate capacitance (which is like 50-500pF, so time constant is ~5µs)?

A switching frequency of only 100Hz to 200Hz is very low. The 25mA from a micro-controller will charge and discharge the Mosfet's capacitance fairly slowly but since the frequency is so low then the Mosfet won't heat much. The Mosfet must be a "logic level" one that is spec'd to operate well with a 5V gate voltage.

If the frequency is much higher then the Mosfet must switch quicker so it doesn't spend most of the time ramping as a linear amplifier getting hot.

Most Mosfets have a high capacitance. An ordinary IRF540 has a gate capacitance of 1960pF plus Miller capacitance. A 1M resistor will take half a cycle of 100Hz to discharge the gate capacitance. Then the Mosfet is heating all the time. When the Mosfet is heating then the load is not getting the entire supply voltage.

solis365 said:

wouldnt it be a decent idea to get a fairly high-resistance pull down resistor on the gate of the MOSFET (say, 1MΩ) to drain the gate capacitance (which is like 50-500pF, so time constant is ~5µs)?

Click to expand...

The time constant of 500pF and 1MΩ is 500µs.

oops, lost a few tens places in there...

but what is the deal with gate-discharging resistors? are they commonly used, or...?

The gate of a Mosfet has a fairly high capacitance. It must be charged and discharged quickly by a fairly high current so that the Mosfet switches on and off quickly so it doesn't get hot.

A resistor can't charge and discharge and it can't have a fairly high current.
But a gate driver can charge and discharge the gate capacitance quickly by a fairly high current.

A micro-controller can drive a logic-level Mosfet with 5V at 25mA. It might switch quick enough if the switching frequency is low so the Mosfet spends most of its time doing nothing then switching slowly occasionally.

A complimentary pair of emitter-follower transistors can provide almost any voltage at up to 800mA. Then the switching frequency can be fairly high because the Mosfet switches quickly and doesn't spend time getting hot.

A Mosfet Driver IC can provide any voltage at up to 2A to charge and discharge the gate capacitance extremely quickly so the Mosfet can switch at a high frequency without getting hot.

You need to use something that sources and sinks current to drive the gate of a mosfet, But if its a MCU pin it might be a good idea to put a high value resistor to pull it down. Since if the IO pin goes tristate its letting the gate float and it might float right in the middle of conduction making a awful lot of heat and meaby overheating it self if its a small one