MOSFET vs. Transistor

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

MOSFET vs. Transistor

Post by jonjonr6 »

I have a basic understanding of a power transistor and what it does, but can someone explain the differences between a power transistor and a power mosfet.<p>Can a mosfet increase voltage output?
If so, can I input 13V DC and get amplification upwards of 10x?<p>Thanks.
IRONMAN
Posts: 33
Joined: Fri Mar 28, 2003 1:01 am
Location: TEXAS
Contact:

Re: MOSFET vs. Transistor

Post by IRONMAN »

The differences are:
A mosfet is a voltage controlled device, as opposed to bipolar transistor which is a current controlled device.
A mosfet has an insulated gate(MOS: metal oxide semiconductor), therfore REALLY high input resistance.
Also they have a very low resistance when conducting fully, making them great for power applications.
The main draw back is that since the gates insulating layer is so thin, they are very sensitive to high voltage spikes(static electricity has been known to fry them on occasion. Also has relativley high junction capacitance(thats part of the principal of operation).<p>As far as increasing voltage output, an amplifier simply takes a voltage or power source and modulates it with an input signal, I.E. you have an input of say one volt RMS, you apply it to an amp that varies conduction in some manner from a power source of say 12 volts d.c., and you wind up with a reproduction of the original signal only magnified by some factor, say five.<p>Hope that helps a little.
rosborne
Posts: 73
Joined: Mon Jun 23, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by rosborne »

Power Mosfets could be used as Voltage Amplifiers, but the primary usage as I understand it is to control (switch) current. MOSFETs make excellent switches allowing for interesting things like pulse width modulation or PWM. The MOSFET can be switched on and off rapidly to produce a speed controller function for a DC motor.
The active region of a MOSFET is so steep that they easily saturate or cut off. So they are difficult to bias in the active region, necessary for amplification.
If you want some of the advantages of both, check out IGBTs (Insulated Gate Bipolar Transistors).
-Rick
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by jonjonr6 »

OK, I think I understand somewhat.
But that leads to the next question, How would you increase a DC input voltage (say 12v) to DC high voltage output (say 120v). There are lots of diagrams for AC to DC, but I have yet to find a DC-DC design that goes much above 30vdc.
Do transformers only work with AC input?<p>Thanks
Jon
rosborne
Posts: 73
Joined: Mon Jun 23, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by rosborne »

Transformers only work with AC. The voltage in an inductor is defined by a first order differential equation as L*di/dt. Conceptually, the voltage across an inductor is equal to the inductance (which is defined my the material and it's geometry) times the change in current over the change in time. When a DC current source is instantaneously connected to an inductor, the inductor resists (not to be confused with resistance) the current that is pouring into it. The inductor resists to the extent that it's physical properties allow. As it resists, a magnetic field is generated around the inductor. If you remove the current source, the inductor will again resist the change in current by supplying the current from it's collapsing magnetic field. If two inductors are near each other the magnetic field in one will cause a current in the other and a transformer is born. The transformer depends on the continual expanding and collapsing magnetic field of a primary and an induced positive and negative current in the secondary. That's why AC only for transformers. DC to DC amplification depends on the concept of a charge pump, which in turn depends on fast switching of charged capacitive elements to create an apparent higher DC voltage. In general you must start with a higher AC voltage than your DC requirement and regulate.
-Rick
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by jonjonr6 »

OK, I understand (to my limited extent).
So if I have 10-14vDC input, and need +12 (~10A), +5 (~20A), +3.3 (~20A), -12 (1A), -5.5 (1A), I should use maybe a voltage regulator or zener to drop the voltage down to a stable input voltage (maybe 5v), then a charge pump to increase that voltage back up (maybe 15v to stabilize the supply), then regulate the votlages out as needed above, and use some power transformers at the output to get the amperage rating I need for each voltage rail. Of course, the negative voltages will need inverting before regulating. <p>And this would be easier than trying to boost 10-14vDC input to the +60/-80 I originally needed to feed my power supply.<p>Thanks,
Jon
kermit5
Posts: 1
Joined: Mon Jun 23, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by kermit5 »

You can't just apply a DC voltage to a transistor's input and get an amplified DC voltage at its output.<p>In terms of using them as part of a power supply, think of a bipolar power transistor as a variable resistor and a power fet as a switch, connected to a power supply. In a bipolar transistor, an input signal causes its resistance to vary between low and high so the output current follows the input signal (at a greater range). <p>In the powerfet, the resistance is quickly switched between practically infinity and zero so the current switches between zero and some high value that that must be limited by external resistance. This is a basic of switching type power supplies. It's desirable to make the switch between low and high as rapid as possible to avoid heat buildup. <p>It is this pulsing in output current/voltage that is rectified and filtered to produce a DC voltage. Voltage mulipliers can be used to raise the output to a value higher than the power supply voltage to the transistor.<p>The main point is, a transistor doesn't simply amplify some small DC input voltage to some higher DC output voltage. <p>W
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by jonjonr6 »

OK, now I'm alittle confused.
Trying to simplify this as much as I can...
A transistor is nothing more than a gate. The collector is like the input. The emitter is like the output. And the base is like the control. If you put, say 10v on the collector, and used a say 5v regulator or zener on the base, then you'd have 5 volts at the emitter, but at a higher current handling than a regulator or zener can pass.
Is this not basically correct?
IRONMAN
Posts: 33
Joined: Fri Mar 28, 2003 1:01 am
Location: TEXAS
Contact:

Re: MOSFET vs. Transistor

Post by IRONMAN »

What you're trying to do is use a transistor in a series pass configuration to amplify the current capability of a regulator?
I use to do that a lot, but I'm sure there are high current regulators on the market(why re-invent the wheel?), I originally thought you wanted to use mosfet in some kind of amplifier, to be honest a series pass bipolar would be better in a linear regulator, but I think I would look at specs for some of the existing regulators first.
One more thing; you do realize that in order for a regulator to pass a certain amount of current, your source(I.E. transformer, rectifiers,etc..), have to be able to supply as much without dropping to such a low voltage level that the regulator wont function?<p>Good luck.
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by jonjonr6 »

So they do make regulators that will pass god solid 5A or more current.
I haven't had any luck finding such a part.
Assuming I find the regulators I need, would I need to up the source voltage using the charge pump to feed the regulators?
I found a design to do this. Looks like I could increase the voltage up to 17v or so, but the MAX1044 only takes up to 10.5v in, and I could peak at 14.4v.
http://www.geofex.com/circuits/+9_to_33.htm
IRONMAN
Posts: 33
Joined: Fri Mar 28, 2003 1:01 am
Location: TEXAS
Contact:

Re: MOSFET vs. Transistor

Post by IRONMAN »

Question: what are you using as a source? I mean what are you getting your raw voltage from?
Sounds like you're wanting to use a car battery, is this correct?
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by jonjonr6 »

The goal is to install a computer in my car. Yes, I know that 12vDC ATX power supplies are available commercially, but the cheapest is $130, and specs out at 14A and 10A on the 3.3v and 5v rails. It's not strong enough to power much more than an ITX or PII system.
I originally wanted to modify an existing ATX power supply. I've seen a website where this has been done, but the guy re-wound some transformers to make it work with his specific p/s.
So, yes, a car battery, and a car alternator.
The pitfalls are obvious, car electrical system is noisey and unstable. The voltages can dip pretty low during cranking, and peak at 14.4 (or more if the alternator's regulator fails). I do have a 1 Farad capacitor to help out. I also plan to use a dual battery isolator to isolate the capacitor from the rest of the car (to it won't drop as much during cranking). I wan tot use standby mode on the PC, but I'm not certain as to how much current is drawn in standby, anything above .1A and the battery's taking a beating overnight.
Yeah, I know this is somewhat nuts, but it has been done successfully.
Thanks
IRONMAN
Posts: 33
Joined: Fri Mar 28, 2003 1:01 am
Location: TEXAS
Contact:

Re: MOSFET vs. Transistor

Post by IRONMAN »

Ouch!!
To be honest what you're wanting to do is both complicated and impractical, I'm not trying to be smart, it's just that what you require would be a tremendous amount of work if you built everything yourself, for instance you need voltages that are higher than the supply & you also need a negative out-put? this would mean making a switching power supply & winding your own toroid coils(large expensive coils).<p> Now there is a much cheaper simpler way to do it: power inverters, they are so commonplace anymore you can get 'em cheap, they automatically regulate the out-put @ 120 VAC, and very simple to install(hook up to battery cables & plug computer in.).<p>If you can't find any reasonable, let me know and I'll do a little research.
But trust me, I'm pretty knowledgable in the field of electronics and the only reason I'd build a power supply like you'd need is if I were REALLY bored.<p>Good luck.
rshayes
Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by rshayes »

There may be some confusion between active devices and the way that they are being used. Both power transistors and MOSFETs can be used either as linear amplifiers and as switches. Power MOSFETs act in a similar fashion to power vacuum tubes, which were used for nearly half a century in both linear and switching applications. Power transistors replaced power tubes for many applications, and power MOSFETs have replaced power transistors for many of these applications.<p>Power transistors used as switches can have very low on voltages when used as a saturated switch. The most extreme case that I know of was a germanium power transistor that was specified with a collector-emitter voltage drop of 50 millivolts with a collector current of 50 amps. This was designed for power converters powered by sea water batteries. The catch is that the collector breakdown voltage of these parts was about 5 volts and that the base drive needed was about 10 percent of the collector current.<p>Low saturation voltages can also be obtained with MOSFETs, but this is done by increasing the die area. Eventually the die becomes large enough to need a larger package. The next problem comes when the die becomes large enough that thermal stress will fracture it unless special mounting methods are used.<p>Early switching power supplies used transistors originally designed for horizontal output transistors in television sets. These could be made to work with rather complex drive circuits that worked around their limitations.<p>These were replaced by power MOSFETs when the price of the devices dropped to a reasonable level (Some of the early devices cost over $80 apiece). Power MOSFETs are easier to drive and do not suffer from stroage time, which makes them much easier to use. The die size is substantially larger than that needed for a bipolar transistor capable of carrying the same current, but for power levels up to a few hundred watts this is not a significant problem. For higher power levels, a combination field effect-biploar device (IGBJT) is used.<p>From your other posting on power supplies, I assume that you are interested in using a 12 volt power source. With such a low source voltage, power MOSFETs capable of switching the currents required may be expensive enough to make bipolar devices attractive again. Some forms of switching supply result in peak currents 2 to 4 times the average current. Power levels in the 150 watt range may require switching currents in the 50 amp range in the power converter. This may be a problem no matter which switching device is used. Parasitic inductances and resistances start to become problems for both power transistors and power MOSFETs at high current levels.
jonjonr6
Posts: 113
Joined: Mon Jun 09, 2003 1:01 am
Contact:

Re: MOSFET vs. Transistor

Post by jonjonr6 »

Hmmm, I think I need more knowledge about power transistors and mosfets. There are alot of terms used (both here and while looking at component specs), and I don't know what they all mean. Obviously they have significant importance.
What is the dropout voltage?<p>And what about this regulator...
http://www.mskennedy.com/pdf/5010rh.pdf<p>Two of these paralleled and I've got the current handling I need.
Keep in mind, my typical voltage would be 13.8 (car running).
I do have a DC/AC power invertor, but I'm trying to avoid using it. It draws .5A just being hooked up to power (off state).
Keep in mind, I have an old system board to play with, so a crude p/s is fine, even if it kills the old board eventually.
thanks
Post Reply

Who is online

Users browsing this forum: No registered users and 27 guests