mosfets vs power transistors / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

mosfets vs power transistors

1 2 3 next

Move Thread

LAN_403

IamSmooth

Sat Jan 19 2008, 02:36AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Is a mosfet just a switch, or can it be slowly "opened" like a power transistor to allow varying amounts of current through it?

Mates

Sat Jan 19 2008, 02:39AM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

IamSmooth wrote ...

Is a mosfet just a switch, or can it be slowly "opened" like a power transistor to allow varying amounts of current through it?

There are high quality sound aplifiers based on mosfets...

Bjørn

Sat Jan 19 2008, 03:00AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

Did you try the HvWiki?

"Metal Oxide Semiconductor Field Effect Transistor.
These are special types of transistors that are voltage controlled rather then current controlled as BJTs are. "

"Mosfets have a logarithmic resistance between Source and Drain as you linearly increase the gate voltage, "

The wiki still needs a lot of work but there is a lot of useful information there.

Hazmatt_(The Underdog)

Sat Jan 19 2008, 04:06AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

You can use a MOSFET as a programmable resistor in "Triode" mode. This is the region prior to the saturation voltage of the mos in the spec sheet.

IamSmooth

Sat Jan 19 2008, 06:06PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Can someone tell me if there is something wrong with this approach or if it will work?
Let's say I have a generator that can put out up to 300w that goes to a grid tie inverter. If the grid goes down, the generator will still need a load. If I heat-sink a mosfet that can handle 350w

will this be sufficient to handle the power?

kell

Sun Jan 20 2008, 02:04AM

Registered Member #142 Joined: Sat Feb 11 2006, 01:19PM
Location:
Posts: 102

You need to go into specifics to get a real answer, but I'd say you need a dummy load.
I'd put 50 watts into a mosfet (with a good heat sink).
But 350 watts? No.
According to the data sheet, it will do 350 at Tc=25 C
How is it going to stay at room temp while dissipating 350 watts!?
You'll never even get within shouting distance of 350 watts with that thing.
Think more like 50. Maybe 100, with a fan and good circulation.

Bjørn

Sun Jan 20 2008, 02:57AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

Just follow what the datasheets says and it will work, always. If it does not work then you forgot something. This transistor has a derating factor of 2.77 W/Â°C so the power you can dissapate depends on the temperature you manage to keep the transistor casing at.

Hint: The factor of 2.77 comes from the thermal resistance between the silicon junction and the transistor casing that is 0.36 Â°C/W. So at 25 deg C and 350W power the silicon junction will be at 350*0.36 = 126 + 25 = 151 Â°C if the casing is held at 25 Â°C. This is the absolute maximum temperature of the silicon junction and it will quickly be destroyed if the temperature is held higher at extended periods of time.

The heatsink will have a thermal resistance that tells you how much it will heat up at a given power, measured in Â°C/W. The interface between the transistor and the heatsink will also have a thermal resistance, that will depend on the transistor package and your ability to mount the transistor perfectly. These two combined will tell you how much the transistor casing will increase in temperature for a given power.

The idea is fine, you just need to dimension everything correctly. You can offload some of the power into a dummy load of some sort.

Slava

Sun Jan 20 2008, 03:35AM

Registered Member #518 Joined: Tue Feb 13 2007, 05:20AM
Location: New York
Posts: 168

mosfets are usually better because they dont lose any power in the gate, they are also pretty awesome because of their sensitivity and can hold a static charge in the gate if you dont bleed it off.

IamSmooth

Sun Jan 20 2008, 05:01AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

How does one program the resistance? I've tried to read on this, but I am missing it.

I conducted an experiment with a transistor and mosfet. I connected the NPN transistor's collector to V+ and the emitter to ground. I had a 3.6v zener in series with a 150 ohm resistor connected to the base. A current meter read the current through the collector. I slowly increased V+ on my power supply. Once I exceeded a few volts the collector began to pass current. As I increased V+ more current passed; when I stopped the current held steady. No problem here.

I then connected a N-mosfet's drain to V+ and the source to ground. I put two 150 ohm resistors in series and connected the mid-point to the gate. I increased V+ and at a certain voltage the current meter detected current going from the drain to the source. Even though I stopped increasing V+ the current was still increasing. I had to turn the power supply off. Why did this happen? Shouldn't the current remain fixed if the gate voltage is steady?

NOTE: I noticed that when I put a 5 ohm load in between the source and the ground the current didn't seem to run away. As the MOSFET heated up the current slowly increased, but not like before.

Bjørn

Sun Jan 20 2008, 06:03AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

The resistance changes with temperature among other things. MOSFETs have generally a positive temperature coefficient, so the resistance goes up when they get warmer. Since power MOSFETs can be very complex devices there are conditions at low currents compared to the maximum current where they have a negative temperature coefficient.

The datasheet of your transistor should give you some curves that shows the expected transfer functions at a few selected voltages and currents.

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob