Forssell's EQ paper - some questions

[abbey road d enfer]

OK, got it.
I don't have any objection using the "Forsell" structure, but I wouldn't have any objection either if you used a string of 3-5 opamps, provided they are of reasonable performance. In the "Forsell" structure, which is in fact the standard structure used in half the graphic EQ's ever made, the main noise contributors are the so-called "gyrators". Your choice of using real inductors would largely compensate using a couple more opamps in the string. You must be aware that for the lowest band, you would need inductors of rather large value (about 10H) and that they cannot be small. I used RM14 Al1000 gapped cores, which computes to 3000 turns for 10H.

 

[Gold]

You must be aware that for the lowest band, you would need inductors of rather large value (about 10H) and that they cannot be small. I used RM14 Al1000 gapped cores, which computes to 3000 turns for 10H.

I guess that would be the main problem/expense. The size doesn't bother me. I could start with the other bands and if I like it see about getting inductors made.

 

[abbey road d enfer]

The size doesn't bother me.

That's not what I meant. I mentioned that because I've seen many times people using the small inductors used for HF decoupling and being disappointed with the results.

 

[L´Andratté]

Goddamn! A whole opamp in the signal path? That´s excessive! Look at this one:

http://1.bp.blogspot.com/-xpNLGD9xkZQ/UFXW19mvOjI/AAAAAAAAD4g/S88GTVjipms/s1600/Vermona-Mixer1222-page3.jpg

Sounds good btw... and a cheap way to get nice inductors, if you don´t mind ripping historic  : hardware.

 

[abbey road d enfer]

Goddamn! A whole opamp in the signal path? That´s excessive! Look at this one:

http://1.bp.blogspot.com/-xpNLGD9xkZQ/UFXW19mvOjI/AAAAAAAAD4g/S88GTVjipms/s1600/Vermona-Mixer1222-page3.jpg

Sounds good btw... and a cheap way to get nice inductors, if you don´t mind ripping historic  : hardware.

Don't expect much performance in terms of headroom and noise. The operating level has been deliberately toned down in order to accomodate the limited headroom. The actual maximum output level is about 1.5Vrms => +6dBu, which suggests establishing the nominal input level at -12dBu.
As I said earlier, there are generally good reasons for what may appear at simple glance to be superfluous components, provided one starts with a good design. In that particular case, the structure itself is not optimized. If the "swinging input" scheme had been applied between base and emitter, instead of collector and emitter, a good 6dB of headroom would have been gained.