Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

Understanding and tuning the presence control

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Understanding and tuning the presence control

    Hello peoples, it's been a while.

    I'm intersted in possibly modifying the presence control on one of my amplifiers to try to tune it's voicing, so I'm trying to come to an understanding of how it works. It is my hope that by fully understanding the presence control, I can desgin my mod intelligently, instead of just blindly swapping parts and praying that the mojo gods will smile upon my guess.

    To be clear, I do have a basic understanding of how the presence control works: Most often, it is a low-pass filter in the feedback loop tied to a potentiometer. Turning the pot reduces or increases the effect of this filter, in turn increasing or decreasing the negative feedback above the filter's cutoff frequency. This changes the power amp's voicing by altering its gain above said frequency. I do know that some amplifiers are exceptions to this, e.g. several Mesa amps.

    I've done some searches on Google and done some RC filter calculations based on the component values I see in various schematics, and this is where I am running into trouble. Per most of the articles and forum posts I've found with Google, the presence control affects the upper midrange and high frequencies. On many amps I've played with, my ears tend to agree. However, I must be reading these schematics wrong. When I calculate the RC cutoff frequency of the components I see as most likely to form a low-pass filter, I'm getting cutoff frequencies in the range of 10 to 40Hz!

    For example, take the presence control in the Marshall 1959, 1987, 2203 and 2204 circuits. If I'm understanding this correctly, the negative feedback enters into the PI at the bottom of the tail resistor, from the voltage divider created by the 100k resistor attached to the output transformer, and the 4.7k resistor to ground. Parallel to the 4.7k resistor, is a .1µF capacitor in series with a 22k pot wired as a variable resistor. Okay, so it seems that the low pass filter would be formed by the .1µF cap and the 100k resistor. But when I calculate the cutoff frequency, I get 15Hz! That covers the entire bandwidth of a guitar, not just the high frequencies! There must be other factors involved to make the presence control effective only over upper mids and highs, but as much as I look at these schematics, I'm not spotting them.

    Where else should I look, and what other things should I be considering?
    Last edited by Koreth; 01-24-2014, 03:08 AM. Reason: Phrase my question better

  • #2
    Originally posted by Koreth View Post
    ...When I calculate the RC cutoff frequency of the components I see as most likely to form a low-pass filter, I'm getting cutoff frequencies in the range of 10 to 40Hz!

    For example, take the presence control in the Marshall 1959, 1987, 2203 and 2204 circuits. If I'm understanding this correctly, the negative feedback enters into the PI at the bottom of the tail resistor, from the voltage divider created by the 100k resistor attached to the output transformer, and the 4.7k resistor to ground. Parallel to the 4.7k resistor, is a .1µF capacitor in series with a 22k pot wired as a variable resistor. Okay, so it seems that the low pass filter would be formed by the .1µF cap and the 100k resistor. But when I calculate the cutoff frequency, I get 15Hz! That covers the entire bandwidth of a guitar, not just the high frequencies! There must be other factors involved to make the presence control effective only over upper mids and highs, but as much as I look at these schematics, I'm not spotting them.

    Where else should I look, and what other things should I be considering?
    You are using the wrong resistor. Try it with the 4k7 resistor.

    Edit -
    Here's a simulation with the 22k pot shorted. Where it levels out at 2kHz is the 4k7 resistor and where it levels out at 100Hz is the 100k resistor.

    Click image for larger version

Name:	Presence.gif
Views:	2
Size:	21.6 KB
ID:	832244
    Last edited by Dave H; 01-24-2014, 09:53 AM.

    Comment


    • #3
      Well, calculating an RC filter with the 4.7k resistor, I get a different value: 338Hz. While not as uselessly low as 15Hz, that still seems really low for a presence control. What program are you using to simulate that? Attempting to draw a circuit like the one I see on the marshall schematic in LTSPICE, the only way I can get a trace similar to the one you posted, is to check the current through the 100k resistor on the simulation, and the scale I get is much different -- less than 1 dB from the highest and lowest points on the plot.
      Last edited by Koreth; 01-24-2014, 07:48 PM.

      Comment


      • #4
        The simulation was LTSPICE. I got it wrong then. I should have calculated 4k7 and 338Hz. It’s in the feedback loop so perhaps it’s the 4k7, 100n which defines where it flattens out a 100Hz and where it flattens out at 2kHz is due to its running out of feedback i.e it’s approaching open loop gain.

        Comment


        • #5
          Perhaps. Could you post a picture of the circuit you're simulating? Mine is extremely simplified, with just a voltage source, the resistors, a pot and a cap. Perhaps it is necessary to be simulating the PI or output tubes as well.

          Comment


          • #6
            Originally posted by Koreth View Post
            Perhaps. Could you post a picture of the circuit you're simulating? Mine is extremely simplified, with just a voltage source, the resistors, a pot and a cap. Perhaps it is necessary to be simulating the PI or output tubes as well.
            Yes, I'll find somewhere to upload the LTSpice schematic and models I used.

            Comment


            • #7
              The LTSpice file and 'inc' models can be downloaded from here. It shows the contents of the zip file, select 'File' then 'Download' and it will download the zip. Extract the contents of the zip to a directory and it should run from there. The simulation I did was the output to the 'LOAD'.

              Comment

              Working...
              X