SP C1 Notch EQ MOD

[kingkorg]

Hi guys. I decided to try what Ruud suggested in this post.

https://groupdiy.com/index.php?topic=52519.0

Most of the time the sibilance is caused by a sharp peak around 8 KHz.
Increasing the values of C20 and C21 causes a HF roll-off, but the peak will remain.
Did somebody ever experiment with a series L-C circuit, connected between the drain and the source of the FET?
(Tuned to the peak frequency, maybe with an extra resistor in parallel with the C to 'tune' the Q-factor?)

So i used a chinese transformer i had as an inductor,  and 4.7 nF in series to match my most sibilant range. Additionally i used standard mod source to ground with 10nF for range above 10KHZ. This is switchable, and i used trim pot of 100k in series with inductor to adjust attenuation of notch cut.

Do you guys think this is ok, and is it ok to use that transformer as an inductor? It sounds awesome. I removed exactly what was bothering me.

 

[kingkorg]

Frequency response of the circuit. with and without mods. Attenuation of notch EQ is 5 db

 

[kingkorg]

Inside

 

[kingkorg]

Modded audio sample
https://app.box.com/s/nzxbsdlvx3qn1h37yzw4rbennuxtmb01

Flat audio sample
https://app.box.com/s/w33calj889ctnd2jxpz2lijobz019q9y

 

[kingkorg]

This would be the mod + switches and a pot

 

[ln76d]

Nice  8)

I'm starting thinking Matija, that you are my brother from another mother
With the next schoeps topology, i will try it for chinese k47 2-4kHz bump.
Some time ago (there's topic somwhere) i had similar idea to use notch, but for chinese u87 with emitter follower.
Main idea was to cut 500Hz.

 

[RuudNL]

Additionally i used standard mod source to ground with 10nF for range above 10KHZ.

I suppose you mean: "drain to ground"?
(Source to ground would boost HF)

 

[kingkorg]

Nice  8)

I'm starting thinking Matija, that you are my brother from another mother
With the next schoeps topology, i will try it for chinese k47 2-4kHz bump.
Some time ago (there's topic somwhere) i had similar idea to use notch, but for chinese u87 with emitter follower.
Main idea was to cut 500Hz.

Well anything is possible, your tweaks work like charm for me as well  ;D ;D ;D

I agree with 300hz-500hz range, I EQ it out on 95% of material. I always found it a bit weird that most of us (including me) like to call mics with that range accentuated ''rich'' and ''warm'' sounding, when soloed, and then just get rid of it as soon as something should be mixed. OK, not everything, guitars tend to benefit from that range.

As i am not familiar with LC circuits, and cant seem to find anything on the internet, is there a way to cut even more than 5db? It is enough, but if its worth doing, why not overdo it? What is the impact of capacitance/inductance ratio? 

For example, 0.5 nF and 1200mH give 6497hz resonant peak, but 2nF and 300mH give 6497hz as well.
I use this guy here:
http://www.deephaven.co.uk/lc.html

I suppose you mean: "drain to ground"?
(Source to ground would boost HF)

I will have to double check that, but yes, most probably, i get cut anyways. Thanks for your inspirational post Ruud, i always thought something like this would need much more tweaking, but it's quite easy, and more useful than most LPF mods, as there is mostly one  or two peaks in the high end, and most tweaks tend to get rid of everything, and then it emerges back once you boost high end in mix. Kind of chasing your own tail.

I need this mostly for live applications, in studio you can do almost anything in post, but in live applications i like to spare my available eq bands (mostly 3 or 4) for creative work, not use all of my parametrics  for removing stuff i dont like in mics.  This helps incredibly!!!

 

[RuudNL]

In an ideal situation you would have the same impedance for the L and the C at the required frequency.
Of course a big L and a small C will give resonance at a certain frequency, as will a small L and a big C, but the 'Q' won't be as good as a well dimensioned L-C circuit.

Xl = 2 * Pi * f * L

Xc = 1 / 2 * Pi * f * C

I suppose with a well tuned L-C combination, you could get a 'dip' of even 10 dB, maybe even more!
(Disadvantage: the higher the 'Q', the smaller the frequency range that is attenuated!)