'Power Switch' type of transister (FET?) recommendation needed.

[toodles]

I'm putting together a little interface board to connect the controller PCB from a Xbox 360 arcade stick to my microcontroller. The digital inputs on the Xbox 360 do not use a common ground, so I am using a collection of four 74HC4066N analog switch IC's to connect the two traces for each digital input of up, down, left, right, and all of the buttons.

What I'd like to do is be able to control whether the Xbox PCB is getting any power, so the power would be disconnected except when I know for sure that I need to use the Xbox controller. I have a digital output from the microcontroller available for exactly this purpose. Because the only connections to this board are from the outputs of the 4066N's and ground, I think I can cut the +V to the Xbox360 controller and have it power down without problem; I dont think there is any leakage current from the 4066N's that will bleed into the board like some digital output lines might do with unpowered CMOS devices.

What I need is a recommendation of a simple part like a transistor that can be activated by the digital output line of a PIC to allow current to pass when the line is high (+5v). The current is probably in the area of <100mA when running normally, but something that could easily handle 200mA is prefered. There is no kind of switching frequency to be concerned with; needing it switched more than once per minute is extremely unlikely.

Can anyone recommend a part that could do the trick? Can anyone recommend a way to do it with regular 3906/3904 transistors? Lastly, how can I figure out the best resistor to use between the digital IO line of the PIC and the base pin of the transistor? (Im extremely naive about analog circuits)

 

[Boncuk]

Hi doodles,

you might use a BUZ11. It saturates at a gate voltage of 3.5V already and can stand some amps. At 100 to 200mA you might omit a heat sink. Connect the base via a 100Ohms resistor to your MCU output and GND to the source (N-Channel MosFet). Then connect the load between VCC and the drain. Package is TO220.

Boncuk

 

[toodles]

I did some digging and found a few 2n2222's in my box, so I will likely go with that, but thank you very much for the suggestion of using the BUZ11's.

One thing I keep seeing when using transistors as a switch to control higher loads such as LEDs is that they seem to all use the switch between the load and ground:
+V-----LED-----(CBE)----GND with CBE being the transistor.
Is this better or prefered over something like:
+V----(CBE)----LED----GND
?

 

[audioguru]

If an NPN transistor is used with the load between its collector and the positive supply then when the transistor is turned on the load gets nearly the full supply voltage.

But if the NPN transistor is used with the load connected from the emitter to ground then when the transistor is turned on the load gets a voltage loss of about 1V to 4V.

 

[Hero999]

That's true but you can eliminate that problem by making the base voltage 1.2V higher than the collector voltage.

 

[audioguru]

That's true but you can eliminate that problem by making the base voltage 1.2V higher than the collector voltage.

How do you make the base of an emitter-follower 1.2V higher than the supply voltage?
Connect the load to the collector where it belongs.

 

[Hero999]

By using another separate supply that's 1.2V higher of course.

 

[Boncuk]

By using another separate supply that's 1.2V higher of course.

Many ways lead to Rome. Why does it have to be the most expensive, unprofessional and ineffective way?

Boncuk

 

[Styx]

Many ways lead to Rome. Why does it have to be the most expensive, unprofessional and ineffective way?

Boncuk

please explain why it is unprofessional? all my gate-drives take the gate potential of an IGBT to 615V (w.r.t. OV of the DC-link) ie above the collector potential

"unprofessional" is a very bad term to use here as well as ineffective (cause it would work). expensive however I do agree with, but Hero999 wasn't suggesting something that *should* be implement, but a means to how it could be implemented

There are many situations where you would want a high-side drive.

 

[Hero999]

I assume that he used such words because English probably isn't his first language.

I was obviously just suggesting a way to do it, I wasn't suggesting that he did it. In this case it's far easier to use a PNP transistor but if this was a MOSFET or IBGT driving a high current load like a motor, it would be far more effective to use the higher voltage power supply as the p-type equivalents of these devices have worse characteristics and are also more expensive.