How to adjust passive peaking filter Q?

[NewYorkDave]

DC specs of the inductor to be important?

Of course they are. Ref. my earlier comments. Any series resistance (and that includes inductor DCR) will have an effect on boost/cut amplitude and bandwidth. You can, of course, scale the input and feedback resistors around the op-amp to suit the inductance and DCR of the coils you're going to use. Buffering the signal before the EQ helps to alleviate the need to keep Rin and Rf at a high value. High-value Rin means higher-value chokes in the tank circuits.

Adding series resistance is not the way to go for Q control in this circuit, regardless of where you put the resistor in the L-C-R series circuit. Its main effect is to limit boost/cut, with its effect on the Q being secondary.

 

[Svart]

So i guess the answer is KISS and the Q adjustment is out of the question.

I'm scrounging up some multi-deck switches for this build now.

Also wilco's sample system is very generous. I don't know if they will actually send me all the samples I requested but their distribution system is lacking. I went to every single distributor's website and there are no easy ways to order parts without going through the quoting system and so forth. I would have gladly purchased these parts had there been an online order form or something of that nature.

 

[TomWaterman]

[quote author="Svart"]That's cool. Do you have any more information on the values of C and L that you used?[/quote]

I tried looking and I can't track down my calculations but did find these two rough pics, this one has Forssell gyrators instead of inductors and is in no way complete...I tried a version with inductors - think I calculated it the same way you have.

http://myweb.tiscali.co.uk/ivyterrace/7-BandGraphic.GIF
http://myweb.tiscali.co.uk/ivyterrace/Curves.GIF

Its something I'd like to get back into when I have the time...as always.
Good luck with your EQ.

-Tom

 

[Svart]

still working on this as time allows.

I have moved on to the LF sections now. I've noticed a serious lack in information on inductor based LF filters and deriving frequency from inductance. Most information quickly explains this and moves on to a lengthy and detailed explaination of a RC LF filter. what gives?

I trust NYD's statements about series R in the LC circuit for modifications of Q but of the information I have read none have stated that there is a serious change in the amplitude of the peak. How drastic is this?

I have parts coming for a build/test as well but I am interested in the meantime.

 

[TomWaterman]

I think to notice any real broadening of the peak you have to add so much resistance as to quite drastically reduce your amplitude.

When I was tweaking the Calrec eq in simulation - OK not the same type of filter but...I noticed it with a swinging input filter too, say you had +15dB of boost, changing the Q even by the slightest amount would reduce boost to maybe 8dB or so..

The series R value on each band also affects max boost/cut overall IIRC.

Forgive me if I'm not following - no sleep for days...

-Tom

 

[NewYorkDave]

[quote author="Svart"]what gives?[/quote]

Mainstream electronics thinking relegates L-C filters for low frequencies to the "obsolete" dustbin and therefore only talks about it in passing.

How drastic is this?

C'mon, man, you know the answer--just think about it a little harder.

The swinging-input EQ is essentially a voltage divider acting upon the inputs of a differential amplifier. In "cut", it's dividing the input signal; in "boost", it's dividing the feedback signal.

Suppose you have a voltage divider: two resistors, a series leg and a shunt leg. What happens to the output voltage if you add resistance to the series leg?

The only real difference between this simple analogy and the problem at hand is that our shunt leg is reactive, not purely resistive.

As Tom found, you can simulate the circuit (on PC or on paper) with various amounts of R in series with the L-C circuit and see this for yourself.

 

[TomWaterman]

Good answer Dave! :thumb:
-T