It may help to give more specific circuit details.
1)What sort of current are you trying to get the PN2222 to pass?
2) What sort of current limiting have you used?

3) If you have a large output capacitor and or conditions of light load current this could stay charged up at 12V when the input power goes off. Have you got a reverse blocking diode to protect the regulator and the PN2222 transistor?

Besides BrianG's excellent points, have you checked the temperature of the 2222 while it is running?

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

The current I am trying to pass is typically 114ma, and there is not an output capacitor.  Below is a schematic of my cirucuit pasted in a from a simulator, perhaps this might help.

I have not noticed excessive heat from the transitor.

The 1k pot is used to "tweak" the output voltage to 12VDC.

R3 represents the 360 ohm load that is normally on the circuit.  There is also a 1W load that is not shown in the simulator.



Thanks for your help.

A 1W load on a 1/2W transistor?  hmmm.

If your input is really regulated well enough you might try burning up your power in carbon instead of silicon.  Even a 27 Ohm 1 W resistor in series with the collector of Q1 might help.  Up to 50 Ohms might be considered.

12V x 0.114A = 1.37W.

The PN2222 can dissipate a maximum of 625mW according to Fairchild's datasheet. You need a transistor capable of dissipating more power, and it might need a small heatsink, depending on the package. The Zetex ZTX453 is one of numerous possibilities - TO92 package, no heatsink required, voltage / current / power ratings are all ok.

----------------------------------
  I don't suffer from insanity. I enjoy it...

Thanks to everyone for their help.  I had just found a ST STSA1805.  TO-92 package. It's max p is 1.1W.  

Still too small.

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

Ok, ZTX849 has a pratical power dissipation of 1.58W

Much betta!  But you still need a heat sink.  Make sure you can get one of those that just pops on top of your choice.

http://www.aavidthermalloy.com/cgi-bin/standarda.pl?PgNum=1&thermalresist=40.5&HeightMax=7&HeightMin=2&;SortBy=TR&NumPerPage=all&Device=5&x=33&y=13

Note: eng-tips will not link to this correctly because of the "?".  So you will need to cut and paste.

Otherwise it will still toast.  1 Watt on a package that size in free air translates to about 300F, that's just the physics of heat dissipation.

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

Dont know how much room you have but
what about reworking it to use something
with a TO220 case like a TIP3055?

The board is small, I really don't have room for the To220 package.

Stanfi, if as you say "there is no output capacitor" this is not good practice. Without a bypass capacitor it is quite possible that your regulator / transistor setup is oscillating at a very high frequency i.e. in the MHz region? Have you checked the 12V rail for noise with a 'scope, not just measured the volts with a meter?

If present, such oscillation can cause lots of power to be dissipated in the transistor.

There are a lot of things you can get away with for a while, but this is an inherently bad design.  Unless your high load lasts for only a couple seconds the supply will eventually fail from heat.  Have you ever measured the temperature of this transistor?  If this design was to have a precision 12.00V power source, it is likely to have a very large droop when the high current occurs.  For the same space you might as well go with a three terminal regulator.  No room for a TO-220 indicates there are no other heat sinking options.  A 24V supply is generally much higher in voltage making things worse.