MOSFET vs. Transistor

[Bernius1]

Just my $.02 ; In either case, a transistor used as a switch is full-on, and probably beyond its amplification range. Re: power supply, don't PC's use +5v AND -5v ?

[Will]

JonJon,
I will try to help you with basics - it seems to me that what much of the posters is saying to you is over your head (Most of it is over mine too) I think it would help you if you could assimilate the essence of dissipated power and then, by looking at the max power rating of a device you would be able to tell whether or not you were going to blow it up.
An ordinary transistor is a current amplifying device - it ha smany characteristics which describe it and one of the most importwsnt is 'hfe' (HFE lower case) which means 'Hybrid Forward Emitter ' characteristic. What it specifically means is the ratio of the collector current to the base current when the transistor is connected in 'Common Emitter' configuration i.e. the base is the input connection, the collector is the oup[ut connection and the emitter is common to both (Usually Ground) Thus, if the you inject base currents of 0.75, 1.0 1.25 mA into a transistor with an hfe of 40 into a transistor working within it's active range range then the collector currents would be, respectively, 30, 40 and 50 mA. Since the base to emitter resistance to current flow is approximately that of a diode i.e. 0.5 - 0.7 volts then, unless you put a resistance in the base line to restrict the base current to that which you desire then you may just blow the junction because of overcurrent/overheating. For instance, if you just applied 12 volts to the base-emitter junction then this would mean a net voltage of (12 - 0.5 - 0.7) applied to only the leads which would certainly result in the 'Blue Smoke Syndrome'
If you are using such a device in a switching mode then, as long as you turn it onmfully a quickly a minimum of juntion heat will be generated and you might get away with a relatively high load. HOWEVER - Supposing your transistor were rated at 1.0 Watt then. if it was turned almost On or almost OFF with a 10 volt collector supply and a 200 ohm resistance between the collector and the 12 volt supply positive (Then you would be reading your amplified voltage at the collector connection) then, with 10 mA or 50 mA flowing the drop across the transistor would be either 10 or 2 volts i.e. in each case (0.01* 10 or 0.05*2)so that the power dissipated in the transistor would be 100 mill-Watt. If you transistor were rated at 150 mW then you would be OK - Suppose then that you wanted a 6.0 volt (collector) output ? The collector volts would then be 6.0 volts and the current would be 30 mA. The dissipated power would then be 180 mW which, in a 150 mW transistor would, once again, result in the Blue Smoke Syndrome.
I hope this helps. MOSFET devices are not current amplifiers but the criteria for assessing volts dissipated across them at various intermediate current loads still applied except that you then multiply the square of the load current by the device ON resistance.
The power dissipation figures in the example above are in accordance with the 'Maximum Power Theorem' Which states basically that the maximum power dissipation occurs in a DC circuit when source and load resistance are exactly equal (200 ohm) i.e. as with 6 volt drop across each as above and assuming negligible power supply internal resistance. The statement of the maximum power theorem for AC circuits is slightly different

[Will]

MISTAKE ! MISTAKE ! The transistor rating above should be 100 milli-Watt - not 1.0 Watt

[Will]

JonJon,
I forgot to ask you - can you post the website to which you referred in your post about a website where someone was messing with an ATX power supply ? I have a real desire to get inside of a computer power supply for general interest purposes but can never find any information on the circuits thereof

[jonjonr6]

I'm confused again. But I guess what I was misunderstanding is that I need the transistor turned either On or Off, not partially on, or else the heat will kill it. Even with a beefy heatsink? We're talking a large aluminum heatsink if needed.
And what about the high power regulators, like the ones available at mskennedy?<p>Anyways, here's a couple webpages I found...<p>http://www.acs.comcen.com.au/atxps.html<p>http://www.formfactors.org/developer/sp ... xspecs.htm<p>I have an ATX power supply schematic somewhere, but I don't remember where I downloaded it ffrom.

[Will]

That's OK - Having the transistor turned always OFF or ON - You still calculate it's power dissipation as stated and if it is within the allowable then you are OK If you have to use a heatsink then you must look at the the cpmbined specs of the heatsink and the transistor. If you were going to use bipolar transistors then you would probably supply the base current to the larger transistor (Say a 2N 3055 or similar) with a small transistor (Say a 2N 2222) - If you were going to use a MOSFET then you need to design the switch on/switch off configuration in accordance with the specs, calculate the power dissipation using the Ron resistance per the spec and design the switch circuit in accordance with some of the advice already given you.
Have fun

[Will]

That's OK - Having the transistor turned always OFF or ON - You still calculate it's power dissipation as stated and if it is within the allowable then you are OK If you have to use a heatsink then you must look at the the cpmbined specs of the heatsink and the transistor. If you were going to use bipolar transistors then you would probably supply the base current to the larger transistor (Say a 2N 3055 or similar) with a small transistor (Say a 2N 2222) - If you were going to use a MOSFET then you need to design the switch on/switch off configuration in accordance with the specs, calculate the power dissipation using the Ron resistance per the spec and design the switch circuit in accordance with some of the advice already given you.
Have fun