Terry,
The solution I posted will work OK for slow to moderate speeds, but it's not terribly efficient as there will always be current flowing through all of the resistors when Q1 is on. It isn't suitable for high speed operation, as the gate discharge time for the MOSFET will be too slow through the 1k resistor.
If you would please start your own thread with a link back to this one, and state your requirements as best you can, we'll see what we can do to help you. This will avoid getting this thread confused; as it "belongs" to gibby_z.
Bill,
Apparently, they want to switch the BuckPucks on from the high side. Don't know why they feel they have to do it that way. If they could let the BuckPuck and LED grounds float, an N-channel logic-level MOSFET would work just fine to provide both the LED and BuckPuck ground.
As we know now by experimenting, the BuckPuck V- and LED- go straight through the BuckPuck, so you really only need to connect one of them.
I just bought some P-channels with the same case style. Still can't find any logic levels P-channels locally though. I suspect this is what they would need.
Logic level P-ch power MOSFETs would help out only over a pretty narrow Vcc range; maybe 4.5v to 5.5v - with an exception.
If Vcc were below that range, you'd need a negative supply to fully turn on the MOSFET, as Vgs needs to get to -4.5v in many cases, however some will be on at -3.3v or so.
If Vcc were higher than around 5.5v (ballpark), the uC wouldn't be able to get Vgs to 0v - unless the uC's supply were riding the positive rail using a negative regulator.