Pultec HF Q - can we reduce gain as Q sharpens? - and now LF question too . . .

Hi,


  so my next quandry, With a pultec, as you sharpen the Q, the boost nearly doubles. Can anyone suggest if it is possible to reduce the gain as the Q sharpens, so gain remains constant at peak?

 I am wanting to use a switch for Bandwidth as well as Gain/Cut, for recall-ability. So with a two deck switch, what can I change to achieve same gain at sharper Q?


  Am I barking up a gum-tree?



   Kindest regards,



      ANdyP

Hey again,

Your hi boost is a 10K pot which you will be building using resistors on a switch.  The 2K2 variable resistor is in line with this filter section and as you vary this resistor you are varying the Q of the filter but also allowing more or less of the filtered audio back to the output. How about the following? 

- replace the 10k hi boost pot with a dual deck rotary switch (say a 12 position)
-- On this switch build two versions of your 10K pot - with the necessary different gain values for each Q setting.
-- both of these decks are fed from the original input to the 10K pot (coming down from the input transformer)

- replace the 2K2 variable resistor with a 3 pole dual throw switch.
-- The 1st deck is varying the inline resistor to alter Q (say going between a 500ohm choice and 2K option) - this replace the original 2K2
-- The 2nd deck is switching between the two wiper options on the rotary decks
-- The 3rd deck is switching between the output of each resistor ladder and feeding it on to the rest of the circuit

A schematic would be clearer but I don't have the ability to pull one together tonight...

All that said if I get around to building a version of this idea I will abandon the variable q and choose one value.  I think there will be other interaction issues at play here.

Cheers,
Ruairi

This should be easy with relays and logic....

  . . . relays and logic . . . . . oh dear . . . . . !



     

HI,


  I guess I could use a multideck switch, or even use a separate gain switch for each Q setting. Real estate wouldn't be a problem . . .


    I'll build a pultec, and see roughly what q settings §I use.( it's been a few years, but iirc, Q was normally sett to full wide)

strangeandbouncy said:

HI,


  I guess I could use a multideck switch, or even use a separate gain switch for each Q setting. Real estate wouldn't be a problem . . .

Click to expand...

or a separate deck for each Q setting...

   yup. like I said!


  these look reasonable for $10 - guess 4 Q settings would be a start


  http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=260622716352&ssPageName=STRK:MEWAX:IT


    Kindest regards,



       ANdyP

how about calculating the Q steps so that they have the same gain difference as your boost rotary?
saves a lot of work but you'll need to operate with two hands...
....one hand increasing the Q with 1 step while the other hand is DEcreasing the boost with 1 step...

Hi Radiance,


  yes I could do that, but I want is an eq that I can adjust the Q independantly, like I could with virtually every other eq I've come across! I guess it is a case of familiarity. 


    Kindest regards,


    ANdyP

But you can adjust the Q independently the way I describe. 
Only thing  is that if you want equal gain you'll have the adjust the (gain) boost as well...
...or am I missing something?

No, you're not missing anything,


. . . I just want to adjust one knob for each!


      so demanding, me . . .



    ANdyP

HI Guys,



    also, I am very familiar with Q sEttings on various eq's, but wonder if it is possible to explain to me how to work out roughly from the curves here what the range of Q is for the HF boost of a pultec? Also, what would these figures be expressed as "bandwidth" in 8ves?( the other way of expressing this)


  before anyone comes down my road, I have been searching, but without fruit so far . . . .



    kindest regards,


    ANdyP

Hi Andy,

Wiki gives this definition for Q...

http://en.wikipedia.org/wiki/Q-factor

About half way down is "Physical interpretation of Q"

The width (bandwidth) of the resonance is given by

where f0 is the resonant frequency, and Δf, the bandwidth, is the width of the range of frequencies for which the energy is at least half its peak value.

Click to expand...

As I understand it, this means you need to divide the resonant frequency by the peak width at half height.

Edit - got 1/x error first time....

For your fuzzy picture above, the peak at 3 KHz on the left looks about 8 to 9 KHz at half height, and the right about 4.5 KHz. So rearranging the equation gives Q = resonant frequency / bandwidth, which would give Q's of 3/9 and 3/4.5 , so about 0.33 and 0.66 respectively.

Common sense check - Q goes up as the filter gets tighter, and if I push those values into T-racks, I get curves similar to yours!

I have to say that I find the log scale distracting when counting the frequency bandwidth. I'm sure someone will shout if I'm barking up the wrong trousers.

Also of interest...
http://en.wikipedia.org/wiki/Full_width_at_half_maximum

HI,


  I didn't use the term factor i my search . . .  Thank you!


    that all makes sense. I had always assumed that it was wider than that on the wide setting. Perhaps because it's so smooth in application?  Maybe I only need 2 settings on my switchable, constant gain version . . . .


    that'd keep my costs down a lot!


  kindest regards,


        ANdyP

I cocked up the maths first time around - better get some coffee!

The octave bandwidth - again it should be at half height. So from 3 KHz, one octave down will be half - 1.5KHz, and octave up will be double - 6 KHz. So your curve on the right looks to be about two octaves at half maximum height.

to ruairioflaherty,


  Hi Ruairi. You mentioned choosing just two q settings. Having played around with the bombfactory Pultec EQP1a Plugin, I am inclined to agree. For one thing, I can build it with cheepo Lorlin switches. Your values of 500R and 2K, are they based on your preferences, or did you choose them out of the air? I am just curious to get someone 's take on it who actually masters for his living . . .


    KIndest regards,



      ANdyP

strangeandbouncy said:

Your values of 500R and 2K, are they based on your preferences, or did you choose them out of the air?

Click to expand...

Hi Andy,

Totally out of the air!  I think you'll probably want the two settings to be something like "fairly broad" and "less broad" - precision work will be done with another tool.  If I could actually accurately spec a 2 position Q switch for a build like this I'd probably be aiming for something like 0.6 for the board setting and around 1 or 1.2  for the narrower one.

As this will be a switch a good idea might be to socket it so you can slide in and out a few values until you find something you like.

I haven't built one of these but I think that they are best kept simple, once you get beyond a certain level of complexity to achieve cool stuff you are better going with a Sontec/Porter style design which are designed from the ground up for more control.

Cheers,
Ruairi

Hi Ruairi,


  Yep, I am wanting a broadstrokes type of action . . . K>I>S>S . . . .


    Thanks!


    AndyP

Hi Guys,
\


    Another question for you (so sorry, but my curiosity is aroused!), Does anyone think it is possible to have a separate switch for lo boost and cut frequencies, like the is at the  Hi end? I know that there will be some interaction between the filters, but it would be really great to be able to add a little 100Hz and gently roll at 20Hz. I know also that I can quite easily suck it and see when I start building, but I haven't started yet, and I would love to know if anyone would think it was a bad idea before I try.


    Kindest regards,



      ANdyP

strangeandbouncy said:

    Another question for you (so sorry, but my curiosity is aroused!), Does anyone think it is possible to have a separate switch for lo boost and cut frequencies, like the is at the  Hi end?

Click to expand...

yes, possible. Steffen did this in his stereo solid state version found here...
http://www.diygallery.de/DIYsites/eq/Pultecschemo.gif


I've build this version myself and I did not notice any strange interactions when boost and cut were set to different frequencies...