NPN transistor with lowest Vbe?

[fizz_it]

Hi - I am looking for an NPN transistor with the lowest Vbe I can get.

Any suggestions for specific transistors? (rather than a material class i.e. GaAs)

I only have 3.1 volts on the collector and the emitter voltage has to be above 2.5 volts to get a circuit to trip. So 3.1v - 0.7v = 2.4 (too low) So a standard BJT won't work. I don't have much PCB space so I can't add a boost circuit of any sort

This will operate only as a switch. And it is a low power application.


Help! Thanks

 

[waht]

A germanium transistor has Vbe of about 0.2 volts. I'm not sure if they are used anymore.

you could also just put an op-amp with variable gain to amplify voltage to trip the transistor.

 

[skeptic2]

All silicon transistors have about 0.7 Vbe. Try germanium transistors with about 0.3 Vbe.

http://www.nteinc.com/Web_pgs/Germanium.html

 

[fizz_it]

skeptic2 - Thanks a 1x10^6

I am ordering the BFU725F from NXP

 

[es1]

So 3.1v - 0.7v = 2.4 (too low) So a standard BJT won't work. I don't have much PCB space so I can't add a boost circuit of any sort

This will operate only as a switch. And it is a low power application.
Any suggestions for specific transistors?

I see you already ordered the transistor but another way to go was with a PNP. So the transistor would be normally on then shut off in the 2.5V situation.

Of course this would invert the logic of the switch. If that was unacceptable an inverter could bring it back.

 

[Phrak]

Hi - I am looking for an NPN transistor with the lowest Vbe I can get.

chances are, you need to rethink your design.

 

[MATLABdude]

chances are, you need to rethink your design.

I'd suggest going with a basic BJT inverter:
http://courseware.ee.calpoly.edu/~dbraun/courses/ee307/F03/13/02_13_PhilippeGonzaga.html

With the output current known, choose the base and collector resistances such that the BJT operates in saturation or cut-off, and you'll have almost the full operating voltage as the 'high' logic level, and only the BJT saturation voltage (usually 0.2V or lower) as the 'low' logic level.

If you can't have an inverted signal, you can cascade two of these stages together (though you'll have to run through the design calculations again for the output stage). This is something you wouldn't be able to do with simple base followers, as you're losing that 0.7V every time!

If you were attempting to actually power something (instead of just sending an output to another piece of logic) you can often use a FET as the output stage (and usually achieve a lower "0")