Announcement

Collapse
No announcement yet.

Triode cascode compressor anyone?

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

  • Triode cascode compressor anyone?

    Hello everybody,

    starting from this article: http://www.valvewizard.co.uk/cascode.html and this quote:
    Grid-leak biasing the Upper Triode: Alternatively we can create a self-biased version by simply using a grid-leak resistor on the upper triode, and a large grid-bypass capacitor. At quiescence, the upper triode is zero-biased of course, but when we input a signal a charging current flows into the capacitor via the grid-leak, which lowers the voltage on the screen grid. The larger the signal, the greater the bias developed, which creates a subtle compression effect on sustained passages, rather like power-supply sag. This method is simple and works well for guitar.
    So when the input signal goes negative the capacitor connected to the grid of the upper triode is charged through the grid leak resistor (that's the first RC constant).

    I figured out that could be interesting to add a grid stopper on the upper triode, so when the input signal goes positive the same capacitor is discharged through grid current (and grid stopper if used) together with the capacitor above.
    Something around 220 nF, with a 47 kOhm and 1 MOhm, would give RCs of 10 and 220 ms respectively.

    DBX 160A values are: attack 5 to 15ms and release 8 to 400ms (in that case are both program dependent).

    Has anyone ever tried something similar, or knows about similar compressors?

    Thank you in advance

    Roberto

  • #2
    No need to specially build one, all overdriven tubes (triodes and pentodes) shift bias when grid starts passing current and charges, the previous bypass capacitor, which changes both tube idle current and most important, transconductnce, which directly affects gain.

    So in any tube: high drive signal > grid rectification > charge storage in the coupling cap > bias variation > transconductance variation > gain variation depending on signal level > compression.

    Thatīs why "tubes compress" , they do so even without asking for it.

    Time constant? .....
    Charge/attack : RC is previous tube generator resistance, so think about 40k for a typical 12AX7 gain stage and C is couplig cap ... do the Math.
    Discharge/release: RC is same coupling cap (.01 to .1uF) and grid reference resistor: 100k to 1M . Again do the Math.
    Many high gain amps add resistors (100k to 470k) in series with grids to slow attack and avoid grid blocking, which is what happens when an excessive peak charges coupling cap "way too much" and tube plain cuts off (instead of just reducing gain) for an appreciable amount of time.
    Remember these are **VERY crude** compressors, and donīt do much individually, justa few dB .
    Thatīs why a single Tube stage (as in Valvestate) only adds a little flavour, best is having many cascaded stages each contributing "something".
    Juan Manuel Fahey

    Comment


    • #3
      Thanks Juan Manuel,

      Thanks for the information, I have a discrete knowledge of standard cascaded common-cathode gain stages, especially on 3-4-5 stages. In this case I was talking about the "pentode" behaviour of the cascode configuration, and the behaviour of the upper grid as a pentode screen. The cap I'm referring to is the one from the upper grid to ground:
      Click image for larger version

Name:	Cascode3.jpg
Views:	1
Size:	36.2 KB
ID:	849906

      The behaviour should be quite different.
      Last edited by Roberto; 06-19-2018, 06:19 AM. Reason: wrong link to the image

      Comment


      • #4
        Originally posted by Roberto View Post
        Thanks Juan Manuel,

        Thanks for the information, I have a discrete knowledge of standard cascaded common-cathode gain stages, especially on 3-4-5 stages. In this case I was talking about the "pentode" behaviour of the cascode configuration, and the behaviour of the upper grid as a pentode screen. The cap I'm referring to is the one from the upper grid to ground:
        [ATTACH=CONFIG]49375[/ATTACH]

        The behaviour should be quite different.
        Roberto - I've used the cascode in several amps. The link below describes a dual cascode-stage preamp as a second channel in a Deluxe Reverb. The compression is very gradual and mild - nothing like a studio compressor - but it is there. With a scope it's easy to see the upper grid voltage swinging in the working circuit as the stage enters/exits overdrive.

        https://sites.google.com/site/string...me/deluxe-plus

        Note that the upper triode grids in my cascode stages are "self-biased" where as the one you show is fixed by a resistor divider. I imagine that will cause some difference.
        “If you have integrity, nothing else matters. If you don't have integrity, nothing else matters.”
        -Alan K. Simpson, U.S. Senator, Wyoming, 1979-97

        Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.

        https://sites.google.com/site/stringsandfrets/

        Comment


        • #5
          Thanks uneumann,

          Both for the link and for pointing me to the fact that I linked the wrong circuit.
          I meant the grid leak biased version, tomorrow I'll amend the link.

          Comment


          • #6
            How about just using a small signal pentode and modulating gain by changing the screen voltage?
            Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

            Oh, wait! That sounds familiar, somehow.

            Comment


            • #7
              That would be for sure the easiest path.
              I also have to say that the compression is part of the role I would like to give to that stage.
              A compressed power amp stage would be a more complete description of its real role.
              Another option would of course do a power amp with triodes and reverb OT, ā la Marshall.

              Comment

              Working...
              X