Roman Blacks SMPS

[Mosaic]

H has anyone tried making this:
With average 2n2222/2n3906 transistors?

https://www.romanblack.com/smps/smps.htm

I suppose there must be some compromises such as power handling?

Anything else?

 

[ronv]

I think it is optimized for the high gain of the BC327 so you might need to tweek it a little as far as base current goes. The efficiency would then go down a little. Maybe Mr. RB will come around.

 

[()blivion]

I have simulated and built the above. You can use other transistors but yes, the efficiency and power suffer a lot. That unit is a buck converter, i.e. drops from a higher voltage to a stabilized lower one. He has another one that boosts power as well. You can improve on the boost converter version by moving the zener regulator to turn off Q2 and get better efficiency with only a small drop in regulation accuracy. There is also my "Scavenger" boost converter in the circuits section of these forms. You can see that *HERE*

 

[Mr RB]

H has anyone tried making this:
With average 2n2222/2n3906 transistors?

https://www.romanblack.com/smps/smps.htm

I suppose there must be some compromises such as power handling?

Anything else?

As RonV said the BC337 has better properties, mainly gain and especially a low Vce saturation voltage when it is turned on.

Yes people have built the SMPS using any small signal transistor like the 2N2222, which has pretty good gain. The low saturation voltage is mainly important when the circuit is providing large currents like >200mA. And also for trying to get high efficiency >90%.

In a lot of cases people use the SMPS to power a 5v microcontroller circuit etc that might only need 10-30mA or so, in that case 2n2222 will work fine.

 

[Timescope]

Please note that the 2N3906 maximum collector current is 200mA. BC327, BC337 and 2N2222 are 800mA.

Timescope