Crossover problems

[spreemusik]

Hello,
after trying for days to make sense of this circuit, I give up. It is supposed to be a crossover (BBC design) with a crossover frequency of 1.8kHz and a low shelf. I've tried simulating it, tried replacing the transistors with opamps, but I see no way this could ever do what it is supposed to. The low shelf works, but somehow these npn pnp pairs don't. Not that I know much about transistors, and the type isn't stated either, but there seems to be no biasing, so I guess the input signal would need some dc offset. Please correct me if I'm wrong.

My plan ist to make a similar xover, using opamps but keeping the original RC combinations. Any pointers are very welcome.

Jan

 

[magicchord]

Yup, it looks like the input needs to be offset to +15volts.

 

[bcarso]

Yep. It must have assumed a nominal 1/2 of V+ bias on the preamp output feeding it.

The lower section creates a decoupled d.c. bias from the preamp via the 47k/15uF lowpass network, so that the 15k at the base becomes one of the R's in a typical maximally-flat unity-gain Sallen Key highpass filter. There is a little gain tweak with the 100nF/47k network from the input Q emitter to ground that you may not need or want.

 

[spreemusik]

Thank you very much. It is slowly starting to make sense. I'll fiddle around some more as soon as I have some time.
Greetings,
Jan