High speed BJT with collector current 1A

[Dipanjan Das]

Looking for a transistor (through hole)

Amp of Current I'm dealing with: 1A
Voltage: 12V
Switching speed needed: minimum 100kHz

Will be using it in H-bridge config to construct a high frequency DC to AC inverter of 100 kHz

 

[BigBoss]

Look at www.onsemi.com
There are many biploars in that range..

 

[crutschow]

I would recommend MOSFETs with a voltage rating of at least 20V and a current rating of at least 3A for that speed requirement. There are many available in that range.

 

[schmitt trigger]

Although you can use a BJT, I'm with Crutschow that the Mosfet is a much better alternative.
Not only from the lower drive losses, but because of the absence from secondary breakdown that plague BJTs.

 

[Audioguru]

100kHz is not high speed and there are hundreds of transistors that can conduct only 1A.
Why don't you use an L293 or L298 IC?

 

[Dipanjan Das]

Why don't you use an L293 or L298 IC?

Well, it would be great if I can use that coz I've an extra L293 laying around...
Can it be that fast? Like 100kHz? I've worked with that IC before while making a small radio control car for a TECH FEST in our college..
I know it has a huge voltage range and probably a good current handling capacity with low internal capacitance.

But still in my other part of the circuit, a single NPN would be needed with similar frequency response (for making a NOT gate, cant use that big IC in my circuit coz I just need one gate of it). And I doubt that IC can really switch that fast.

 

[crutschow]

A small, fast NPN is a 2N2369.
Any digital-logic IC will work at 100kHz.

 

[Dipanjan Das]

a 2N3019 would help? I'll use L293D for H-Bridge... but for that NOT gate?
the voltage and current ratings seems to be fine..
but the switching frequency? I dont know how to get that

https://www.onsemi.com/pub_link/Collateral/2N3019-D.PDF

Thats the data sheet

 

[crutschow]

A 2N3019 is a fairly slow (10µs turn-off) switch.
A 2N2369 or a 2N5769 are both relatively fast.

 

[dick_freebird]

Your "switching frequency" depends on whether (or how
hard" you will saturate the device along with the working
temperature. A switching transistor may have a datasheet
recovery time spec. A linear one may not bother and all
bets are off. One that does offer a turnoff time spec may
only do so at one test condition, while your load variation
could well put you at different, varying points on the map.
This could just be a minor duty cycle distortion or it could
end up with a transistor that never turns off (if 1/fSW is
shorter than saturation recovery time, you will not leave
saturation before you re-pump the base full of charge).

This is why MOSFETs win anytime you are much above
50kHz and not competing on ultimate current density.

H-bridges care a whole lot about dead times and shoot-
through prevention. A very variable turnoff time which
you want to match-and-offset tightly with a not-so-variable
turnon time, does that sound like a recipe for success or
just a way to solve the surplus transistors problem, one
smoke cloud at a time?

 

[schmitt trigger]

........... does that sound like a recipe for success or
just a way to solve the surplus transistors problem, one
smoke cloud at a time?

The original poster has not mentioned exactly why he really wants to use a BJT, but your explanation sounds plausible to me.