> This just seems too good to be true ;-] Its a cruel hoax by Fairchild who
will
> discontinue the part the day before I order ;-]
>
> I'll put you on my beer distribution list if this works out ;-]
>
> >$.22 ea @ 1K pcs.
> >pn  FDC6323L Integrated Load Switch
>
> If it works this is much cheaper than the combination of other parts ie
2222
> and power trans.
>
> My only question would be if its possible to hand solder this. I didn't
see a
> through hole model and I haven't tried to solder a sot 6 before with an
iron.


Don't be seduced by the dark side of the force, Luke, er Dave.
This part indeed sounds wonderful BUT do you need it.
Did you follow what I said about using a single pnp transistor and 1
resistor (possibly 2 resistors).
There is no need for an extra "power" transistor if you choose the part
well.
It is possible that the IC will be cheaper than a single transistor but
maybe not.
For small volumes or hand soldering the transistor is also likely to cause
you less mechanical pain.

Required:
1 x PNP TO92 transistor.
1 x 470r or so resistor
Try BC327, ZTX749 or anything you can find with similar specs.
The Zetex ZTXxxx Elines generally have better specs than almost anything
else but are not as cheap.
A BC327, which may work OK, costs about $US0.05 in any sort of volume (or
less).

Transistor emitter to 3.3v+
Transistor collector to load.
Transistor base to 470r.
Other end of 470r to PIC pin.

Done.

If you want this to turn OFF when the PIC is powered down then add a 10k
from base to 3.3v+
If you want it to turn on (and start the PIC) if the PIC is asleep (which
you probably don't) connect the 10k from base to ground.

Resistor values will need a little thinking about.
In above case with 3v3 supply the PIC provides SAY 0.5v low to turn on.
460r sees (3v3-0.5) = 2.8v.
Resistor = 2.8/470r = 6mA
PIC can easily supply this.
Transistor needs to supply 350 mA so beta required = 350/6 = 60 or more.
The BC327 will have rather more than this and most applicable Zetex
transistors will have much more than this.

Vsat should be a few tenths of a volt at worst.

I'm not sure what the attraction of using the dark side solution is but by
all means try it.

Note that both solutions do NOT address the current limit problem which I
and others mentioned.
(Unless the driver has selectable current limiting).
That's why I asked the questions about how the LEDs would go on 3.3v or
close to it.
You need to be able to answer the questions I asked to be able to actually
design this rather than just hoping it will work OK.

Tr the one transistor solution -  you may be pleasantly surprised (and maybe
not :-) ).


        Russell McMahon

.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.