"Never switching off" amp-is it possible in reality?

[darkfenriz]

hi everybody!
is it possible to make an amp with output bjts never seitching off by a current source at its base?

This one simulates well, with overbiasing like 350-500mA neither of power bjts switches off. But is it real in terms of: equal and known beta, thermal stability..... and so on?

I think componets selection would be crucial, maybe MJE21193/4 ?

any comments welcome!!!!

best regards

[Mr Evil]

It's possible, but from your schematic it looks like all the current is being pushed directly into the bases of the output transistors. That will make it totally dependant on their hfe and thus unworkable in real life.

What you want to do is control the Vbe of each transistor individually. In your circuit you could accomplish that by removing C7 and splitting R50 and connecting the midpoint to the output. Then a fixed current through the now two resistors will enforce a minimum Vbe and thus a minimum current.

[Eva]

This concept is right provided linear enough output devices, but note that the drivers are working in class B

[darkfenriz]

Originally posted by Eva
...note that the drivers are working in class B

yes of course. I don't find it a problem though.

Thanks

[lumanauw]

Can you trick the cct more upstream, like around Q54-55? Maybe a resistor there to the rails from bases gives the same effect?

[Nelson Pass]

You can reduce this trick to the switching of a pair of
small high speed diodes, but you'll have to go elsewhere
to have no switching elements in the gain path.

And why bother? Get a nice efficient driver with a big magnet
and drive the hell out of it with 10 watts of Class A.

[Steven]

darkfenriz,

I don't think this will work. You're just pulling the drivers in reverse. And maybe the drivers are better in switching than the output transistors, but still...

Something completely different: I noticed that you have added a capacitor to your current sources, probably for decoupling, but that is not what they will do. In this way you just increase the noise of the current source and add power supply noise to the output. This current source is working with feedback and the capacitor is removing the AC feedback path, making the current very sensitive to power supply noise.

Steven

[ingrast]

You may want to check a design I proposed in case you have not.

It could easily be adapted to bjt's in EF configuration, main advantage is feedback loop that stabilizes quiescent current independently of device and temperature variations.

If you have trouble figuring out how to adapt, let me know, see if can help.

Rodolfo

[thanh]

In your circuit you could accomplish that by removing C7 and splitting R50 and connecting the midpoint to the output. Then a fixed current through the now two resistors will enforce a minimum Vbe and thus a minimum current.

i think too.
darkfenriz! do you try to do it?

[darkfenriz]

not yet