"Corrective EQ"

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

RuudNL

Well-known member
Joined
Apr 26, 2009
Messages
3,106
Location
Haule / The Netherlands
Did somebody on this forum ever try one of the "Corrective EQ mod kits" as offered by Microphone Parts?
I am just curious if it really will cure the over bright top end of cheaper large diaphragm microphones.
Also interested in the type of modification. Bacause there is not very much amplification in the electronics, the only way to attenuate the top end is using a capacitor, I think...
 
I'm going to bump this because I as well am looking for a way to roll off highs in a project mk319 that had a bad capsule and I used the rk67 as a replacement. I'm sure there is a way to roll off some highs without killing your signal our making it more distorted. On a side note I was wondering if anybody knows of anything else in the microphone I should change out to accommodate the capsule change as I'm sure there are a pretty big difference in the specs between these two capsules. Thanks!
 
Not sure if those are the Jim Williams kits you mean, but if Jim designed them, better believe they work.  8)

We could invite him over, but ......
 
The problem with most cheap LDC's is that there is a boost around 8 KHz.
You can attenuate the high frequencies, to make it sound better, but if you measure the frequency response, you will find that you have flattened the 8 KHz bump, but also have lost some real top end.
Most cheap (Chinese) microphones use a K67-ish capsule, without frequency compensation.
The best solution to really improve such a microphone, is to replace the capsule with a K-47 type.

If you like to keep the original capsule, increasing the values of the capacitors on the bases of the PNP transistors (usually a Shoeps circuit is used) will improve the audible exaggerated high end.

IMHO it is not necessary to replace 22 (or so) components in the circuit.
The biggest improvement is the capacitor between the capsule and the gate of the FET, the FET itself (and the biasing!) and the output transistors. (Paired and selected for noise.)
 
Recently I have been experimenting with some 'Corrective EQ' capacitor values.
First I tried the solution suggested by 'microphone guru' Jim Williams.
He suggests to add two capacitors of 5600 pF (3300 pF was also mentioned) between the base and ground of the PNP transistors in the 'Schoeps' circuit.
In my test configuration, the signal drop at 10 KHz was barely noticable, and even at 15 KHz. the drop in signal was less than 1 dB.
IMHO this is not strange, since the impedances at those points are rather low: the drain and source resistors are 2200 ohms and the FET impedance is in parallel with the resistors. (OK, the source impedance is lower than the drain impedance, but anyway the impedances are low.)
At the moment I connect a 18 nF capacitor between the drain and the source of the FET, the output drops 2.5 dB at 10 KHz.
So this would mean that two capacitors of 36 nF (if this value exists) between base and ground of the PNP transistors would produce the same result. By the way: I use a 2SK170 FET.
Who has more experience with 'corrective EQ' and what are the values used in practice?
It seems the capacitors should be more in the region of 56000 pF than 5600 pF to produce an audible result!
 
An easy way to do this without modding your mic is to place a 10kOhm potentiometer and a capacitor in series between xlr pins 2 and 3 of the mic output. I did this to the power supply for one of my tube mic projects (see pictured), but you could also make a box with an xlr in and out, and tone control that does the same thing so you can use it for any mic.

Interestingly, this idea came from the Vox AC30 cut control. The potentiometer setting determines the amount of signal attenuation. The value of the capacitor determines the frequencies that are cut. This is a low-pass filter, so small capacitor values will cut only the high frequencies. As you increase the value of the capacitor you will start cutting mids, and finally you'll cut the whole signal if the capacitor value is too large. I chose 0.2uF as the capacitor value that worked with my particular mic, but I'm sure that you'd want to experiment.

The potentiometer is a 10kOhm Alps, available at RadioShack. It has a switch built in, so you can take the circuit completely out of the signal path by turning it all the way counterclockwise until it clicks.
 

Attachments

  • cut_control.jpeg
    cut_control.jpeg
    594.1 KB · Views: 47
Mmm... Your solution might work in some cases, but it will cause a heavy capacitive load.
Especially since the output impedance of the microphone is already very low.
Just for testing, the HF equalizer on my mixer is an easier solution!
 
There's a much more elegant circuit in my MicBuilders Directory under 'ChinaMod+U87hybrid'.

Some of Jim's recommendations are really yucky.
 
As chance would have it, i was the one that "opened up that can of worms" on micbuilders  ;D

Specifically because i got my hands on a cheap-ish Alctron HSMC001 (aka. Apex 435), which, from what i've read online, has a K67-type capsule (with the HF boost "by design") coupled with the flat Schoeps-style circuit.

The latest revision can be found here:

http://f1.grp.yahoofs.com/v1/ZPMZUq1tBlZuWywpqQ2DhPRVcgKXvnmKKfpUHbSnmMFRbzZhfQAKegsh-EGwPTJ5IcEsOo1kHZBdZbeBs9ruRuTX8hgUWUE/Ricardo/ChinaMod%2BU87hybrid/ChinaMod%2BU87hybrid.doc
 
Well, you have to be a member of the Yahoo group :) And the file's too big to attach here (1.8MB)

Then again... I guess i can shove it up onto my Dropbox ;D

https://dl.dropboxusercontent.com/u/20127094/Mics/ChinaMod%2BU87hybrid.doc
 
OK, I'm back again after a couple of weeks abroad...
Many thanks to Ricardo, for his link. I knew this modification already, since I am also a member of the MicrophoneBuilders group.
The capacitor (+resistor) modification in parallel with the drain resistor of the FET works.
But the thing that I don't like, is that is causes an unbalance in the output.
Some people say that this doesn't matter, but I would prefer the correction to work on both output legs, so that symmetry is maintained.
But the fact that surprised me most, was that I need way bigger capacitors than Jim Williams suggested.
The suggested value of 3300 (or 5600) pF across the inputs (B-C, C=Ground for AC) of the PNP transistors in the Schoeps circuit, only gave a negligable effect on 10 KHz.
A citation from Jim Williams found on the Internet:

"[Rode NT2]
On one I kept the K67 capsule. There are a pair of Wima FKP-2 470 pf green box roll-off caps fitted. I increased those to tame the tops of that capsule. I ended up with Wima MKP-2 3300 pf, it sounds very much more like a U-87 curve, the screaming 12k hz peaks are gone. On another I used a C-12 capsule. I left the 470 pf caps in place. It sounds very good, but the headbasket design adds some low mid resonance that I don't like. That quality is gone in other mics I've installed those in.


and:

"For the stock K-67 capsule I replaced the 470 pf caps across the 100 k resistors with 5600 pf. That tunes the response down closer to flat and removes that large 12k hz peak."

Well, I tried the same, but there is almost no measurable (let alone: audioble) effect with 3300 pF added.
To get a measurable effect, I have to increase the values of the capacitors to 22 nF or even higher.
As far as I know, Mr. Williams is a well respected microphone 'modder', so I am a bit puzzled...
 
RuudNL said:
Many thanks to Ricardo, for his link. I knew this modification already, since I am also a member of the MicrophoneBuilders group.
The capacitor (+resistor) modification in parallel with the drain resistor of the FET works.
But the thing that I don't like, is that is causes an unbalance in the output.
Some people say that this doesn't matter, but I would prefer the correction to work on both output legs, so that symmetry is maintained.
...
As far as I know, Mr. Williams is a well respected microphone 'modder', so I am a bit puzzled...
You may like to listen instead to well respected microphone designers like Dip Ing Jo Wuttke, Neumann & Calrec instead of 'modders'.

Try Douglas Self, Rod Elliot, Jensen & others about the benefits of impedance matching as opposed to this misguided idea that 'signals need to be balanced'.
But the fact that surprised me most, was that I need way bigger capacitors than Jim Williams suggested.
The suggested value of 3300 (or 5600) pF across the inputs (B-E) of the PNP transistors in the Schoeps circuit, only gave a negligable effect on 10 KHz.
....
Well, I tried the same, but there is almost no measurable (let alone: audioble) effect with 3300 pF added.
To get a measurable effect, I have to increase the values of the capacitors to 22 nF or even higher.
There is an IMMEDIATE detrimental effect on distortion and overload.  By the time you get audible response changes, this is quite nasty.  One of William's more yucky mods.
 
Yes, you are right. The distiortion increases a bit, although not unacceptable.
I think the best solution to 'tame' the HF boost would be (frequency dependent) negative feedback.
 
RuudNL said:
I think the best solution to 'tame' the HF boost would be (frequency dependent) negative feedback.
That's exactly what 'ChinaMod+U87hybrid' is.
 
OK, to sum things up:
An often suggested method to introduce de-emphasis, is putting capacitors across the B-C junctions of the PNP transistors in the 'Schoeps' circuit. This is the same as connecting the capacitors across the source and the drain resistors of the FET.
To simplify this even further, one could use a single capacitor between the drain and the source of the FET:

Cbb.PNG


As Ricardo pointed out, this increases distortion. Measurements I did confirmed this.
As an alternative, it is suggested to place a capacitor across the drain of the FET only.
For low frequencies, the FET would then act as a phase splitter, but for high frequencies the FET would act as a source follower:

Cd.PNG


This works. Disadvantage is that the attenuation of high frequencies only works on one 'leg' and that the maximum attenuation can be no more than 6 dB.

In the circuit I suggest here below, the attenuation will be (more or less) equal on both 'legs'.
Also the attenuation can theoretically be infinite.
The source and drain resistor are usually 2.2 K.ohms.
The input impedance of the PNP transistors is roughly Hfe * Re = ~400 * 6800 = ca. 2.7 M.
This impedance is in parallel with the 150 K. resistor, say 140 K.
So a series resistor of 10 K (R7 and R8) between drain (or source) and the transistor would lower the output less than 1 dB.
The two 10 K resistors would even be at high frequencies a modest load for the FET.

Rbb.PNG


Would this be a useable configuration, or do I overlook something?
 
RuudNL said:
As an alternative, it is suggested to place a capacitor across the drain of the FET only.
For low frequencies, the FET would then act as a phase splitter, but for high frequencies the FET would act as a source follower:

This works. Disadvantage is that the attenuation of high frequencies only works on one 'leg' ....
Would you like to point out a 'real life' case where this is a disadvantage?  What 'real life' problems does it cause?  (I know some but only in such yucky situations where it would hardly matter.)
In the circuit I suggest here below, the attenuation will be (more or less) equal on both 'legs'.
Also the attenuation can theoretically be infinite.
The source and drain resistor are usually 2.2 K.ohms.
The input impedance of the PNP transistors is roughly Hfe * Re = ~400 * 6800 = ca. 2.7 M.
This impedance is in parallel with the 150 K. resistor, say 140 K.
So a series resistor of 10 K (R7 and R8) between drain (or source) and the transistor would lower the output less than 1 dB.
The two 10 K resistors would even be at high frequencies a modest load for the FET.

Would this be a useable configuration, or do I overlook something?
Yes.  The 10k resistors introduce a HUGE amount of noise.

Why don't you make something with my method first?  It's easy.  Then you have a baseline to see if you can do better.
 
Back
Top