Using a tactile switch rated for 50mA to drive the base could someone suggest a good power transistor to run 12A?

Also, can typical PCBs handle that kind of current?

 

Yes, a PCB can handle 12A, just keep the traces short and wide. I'd recommend a MOSFET rather than a BJT. What voltage levels do you have in mind?

 

Around 3-4V.
I was considering mosfet use, but I don't want to limit the current at all when the switch is on. From what I've understood the mosfet has a pretty wide gate to source voltage range to cover in order to fully enhance. But I do like the idea of extremely low (virtually zero) control current that mosfets provide. If there is one that is fully enhanced at 3V or more and fully cut off at 2.5V that would be perfect.

 

Around 3-4V.
I was considering mosfet use, but I don't want to limit the current at all when the switch is on. From what I've understood the mosfet has a pretty wide gate to source voltage range to cover in order to fully enhance. But I do like the idea of extremely low (virtually zero) control current that mosfets provide. If there is one that is fully enhanced at 3V or more and fully cut off at 2.5V that would be perfect.

There are MOSFETs that are fully on at 3V, but if you are driving with a switch, why would you need it to be off at 2.5V?

 

In this a BJT has much the same problems as a transistor, if you are using it in digitial modes I mean. You need to figure a way to feed it 0V for a low, and 5V for a high, or compensate somehow. I've been to use LEDs and zeners to absorb extra voltage (a white LED resembles a 3.6V zener), but I doubt this would have good high frequency characteristics.

 

It's nothing to do with high frequency...it's just a slow manual tactile switch.

The reason for cutoff at 2.5V is to act as a low voltage limiter so if the voltage is less than that it doesn't come on at all. Not as important as being set up so that it is full at on 3V.

 

Yes, but you mentioned that was why you didn't like MOSFETs. Anything analog will screw with either transistor. Because of this limitation some feeder circuitry would be required by either. Check this thread out...

http://forum.allaboutcircuits.com/showthread.php?t=22027

 

I just don't want to have to go up to 6 or 10 volts to fully enhance. As long as there's no current limiting at 3V that's fine. A 0.5 V ramp up is perfect. Completely off at 2.5V, starting to come on from 2.5 to 3.0 V and fully on at anything more than 3.0V

 

With a BJT, to get a DC current gain of 250+ you need a Darlington pair. Expect Vce at saturation to be around 2V. This reduces the voltage available to the load. At 12A the transistor will dissipate 25W and need a heatsink.

I'd recommend using a low-voltage MOSFET driver, like the Micrel MIC5014/5015 and an N-channel FET. You can achieve a circuit that is more compact and much more efficient. It will be more consistent over a range of inputs and loads. It gives you a very wide range of transistors to use; A cheap, common transistor may actually save you the cost of the driver, versus an exotic transistor with precise parameters.

 

This (MIC5014) appears to be pretty handy. It comes on at 3V which is right on. Thanks for that.
If I'm reading the specs correctly, it's supplying about 4V extra at 3V so if I find a regular logic power MOSFET that is fully enhanced at 5V I should be ok.

 

Fairchild FDS6574A has Rds(on)=9milliohms with Vgs=1.8V. There are many spec'ed at Vgs=2.5V.

 

Thanks. One properly chosen component doing a job that would otherwise require multiple components. Perfect.