Good BJT for 3A operating current?

[RichardJones]

I am looking for a BJT that is linear and is temperature / drift stable at a quiscent current of 3A. NPN.

Going to be run as an emitter follower into a choke load.

Power supply voltage may be as high as 20v. So power dissapation is a must.

Any good ones out there???

[Jocko Homo]

What exactly is your definition of good? Linear? Cheap? Rugged? Fast?

Jocko

[RichardJones]

linear and fast.

and sounds good too!!

[Jason Hubbard]

Take a look at MJL3281 - low beta droop at higher junction temperatures and comes in TO247 case so should disspate heat well.

[AKSA]

Richard,

As mentioned on the joenet, use 2SC5200 (NPN) and 2SA1943 (PNP).

These are made by Toshiba, have typical beta of 100, 12A, 230V and 30MHz. They are flat packs, excellent devices, and beta flat from 100mA to 7A.

Cheers,

Hugh

[Brian Guralnick]

I personally just love the EM follower gain curves between 1 amp & 5 on On Semiconductor's MJL3281A & MJL1302A pair. They also are clean up to 30Mhz.

[RichardJones]

If I were to use the MJL3281A in a Class A output stage loaded on the emitter with a choke what is the easiest way to bias it?

Probably run from a +24v output stage.

Bias around 3A.

I want a single gain stage... voltage gain of 1... using an input tranny to step up the voltage.

Unless I use a CFP with a MOSFET and BJT.. would this minimize the Ciss?

Thanks!!!!!

[Brian Guralnick]

Originally posted by RichardJones
If I were to use the MJL3281A in a Class A output stage loaded on the emitter with a choke what is the easiest way to bias it?

Probably run from a +24v output stage.

Bias around 3A.

I want a single gain stage... voltage gain of 1... using an input tranny to step up the voltage.

Unless I use a CFP with a MOSFET and BJT.. would this minimize the Ciss?

Thanks!!!!!

1. Easy 3A bias, if you want heat, get a 100v supply, use a 29 ohm 350 watt resistor & run the transistor only in the last 24 volts. This would be a minimalist purist design.

2. However, would using s second MJL3281A as a constant current source be acceptible? For this, you would only need a powersupply with an additional 12-24v on top of your 24v output range. If so, I can draw up something simple. How about using 3 LM317 1 Amp constant current desing?

Note, without the additional voltage on your supply, your constant 3A will not be well regulated as Vbc increases to 24v.

[RichardJones]

Could you draw up something???

I am looking for a simple input stage to this.... probably a step-up tranny, would this work? like a 1:10.

How would it be connected?

I like the idea of 100v supply with the resistor loading instead of the CCS....

Any other ideas or thoughts?

[Brian Guralnick]

Originally posted by RichardJones
I like the idea of 100v supply with the resistor loading instead of the CCS....

I am in the middle of designing a balanced bipolar regulated Mosfet output version myself. The Balanced design only needs a 60V supply to match a 100v unbalanced version, and it further supresses power supply ripple noise & helps cancel out supply load due to large output signal swings.

Yours could be a nice mosfet driving transistor base design. Though, if you want a full BJT, the transistor drive would look like a darlington. I'll send a diagram tonight when I have some free time.

The best pric/watt effective resistors I've found to date:

Huntington Edgewound Resistors:

300 watts, Air cooled, Digikey ripoff price, 26.56$ cdn. (*0.64 for US$ price)

50 watts, Heatsink / Chassi mount, 1%, Digikey ripoff price for 10, 67.20$.