Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

Question about prevalent 6L6GC Screen Grid Voltages

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

  • Question about prevalent 6L6GC Screen Grid Voltages

    Hello Again,

    I have recently become the owner of a Mesa Boogie Bass 400+ amplifier. Planning on adding an adjustable bias circuit for the power stage, I took a look at the only free schematic I could find for it online, Power Supply.

    After reading about notoriously cold biasing of power tubes in Mesa amps, I decided to modify the circuit, rather than just sticking in a pot of the same value as the original bias resistor. This led me to take a look at the datasheet for the 6L6GC power tubes in the amp, to get an idea of the typical operating circuit for these tubes, as well as their max plate dissapation. Comparing the operating parameters listed on the datasheet with the circuit of the Power Amp in my amp, I was very surprised. The schematic is of poor quality, but I'm fairly certain it lists the plate voltage as 480VDC and the screen grid as 465VDC. This screen voltage is off the charts on the datasheet, with 400VDC being the upper threshold they bother to graph. The bias voltage is an icy -54VDC, also on the extreme low end/unplotted on the datasheet charts. Looking at a the schematics of a few of the other non-adjustable bias 6L6GC amps I owned, the same scenario presented itself:

    Fender Silverface Quad Reverb

    Peavey Butcher

    I've read that independently, excessive control grid bias causes nonlinearites, same for high screen grid voltages in relation to plate voltage. When the two conditions are combined, would this distortion be increased? Also, does the high bias voltage affect the slew rate of the tube, reducing the "tightness" of the amp? Would decreasing the screen voltage while moving the bias voltage closer to zero improve the sound of these amps? Would this be a "healthier" operating condition for the tube/amp (assuming proper dissipation parameters were observed)?

    - Jack

  • #2
    ya, I know.

    First of all, vacuum tubes tend to be more tolerant of voltage, verses currents. The values listed on those data sheets are for reference only. I've run the 6L6 at 600 volts with no problems, but I run the screen at 300 volts to keep the plate current in check, and thus the total plate dissipation. Hope this helps.


    -g
    ______________________________________
    Gary Moore
    Moore Amplifiication
    mooreamps@hotmail.com

    Comment


    • #3
      Even though these amp designs are within tolerance as far as max plate watts per tube are concerned, are they optimal, or rather designed to allow different brands/ages/plate hours of tubes to be swaped without adjusting bias levels?

      Comment


      • #4
        I'd guess that the design criteria for Leo was to get the absolute max out of the tube / parts compliment, without stressing the tubes available at the time so much that they failed in warranty, or so soon that it got them a bad name.
        A few % distortion was probably seen as immaterial for guitar amps when the speakers are adding even more. Peter.
        My band:- http://www.youtube.com/user/RedwingBand

        Comment


        • #5
          So, would it be beneficial to knock the screen volts to about 300VDC, while increasing bias to about -20VDC? Or would the difference be negligible? Even for the sake of making the tube easier to bias, because I'd be somewhere around the middle of the tube charts, rather than at the extreme ends.

          Comment


          • #6
            ...FWIW, Leo Fender seemed to idle his amps at about 55% of tube plate dissipation rating for 6L6 amps and at about 60% for 6V6 amps.

            ...CBS/Fender changed from those values when they introduced the UL-output models.
            ...and the Devil said: "...yes, but it's a DRY heat!"

            Comment


            • #7
              Originally posted by apehead View Post
              The schematic is of poor quality, but I'm fairly certain it lists the plate voltage as 480VDC and the screen grid as 465VDC. This screen voltage is off the charts on the datasheet, with 400VDC being the upper threshold they bother to graph.
              The max spec for the valve is 450V on the screen, but you can push it, if you use big screen-grid stoppers.



              I've read that independently, excessive control grid bias causes nonlinearites, same for high screen grid voltages in relation to plate voltage. When the two conditions are combined, would this distortion be increased?
              Hard to see what sort of distortion you mean. Are you referring to crossover distortion? in which case biasing colder will worsen it, although the detrimental effects have been exaggerated by 'amp techs'.

              Also, does the high bias voltage affect the slew rate of the tube, reducing the "tightness" of the amp?
              No, that's determined by the valves input capacitance, which doesn't change with voltage. In any case, slew rate limiting is of no consequence in a guitar amp.

              Would decreasing the screen voltage while moving the bias voltage closer to zero improve the sound of these amps? Would this be a "healthier" operating condition for the tube/amp (assuming proper dissipation parameters were observed)?

              - Jack
              Healthier, yes, but not necessarily better sounding. There comes a point where it is so low that you hit triode-like operation, which can be a bit bland. By all means try it though. You should be fine down to 400V or so.
              Last edited by Merlinb; 05-12-2009, 01:24 PM.

              Comment


              • #8
                Every musical instrument amp I've ever seen runs the screens over the recommended voltage. It seems you can get away with it fine, if you use properly sized screen resistors.

                One thing to think about is that unbypassed screen resistors reduce the tube's gm. The bigger the resistors, the more gm you lose.

                Another thing is that screen voltage directly translates to output power. The more voltage you apply (or the smaller screen resistor) the more power the tubes will deliver. So you probably don't want to reduce the screen voltage in a Bass 400+, since they never made anywhere near 400W to start with. Or maybe you do, since the tubes will last longer.

                If I were designing a high-powered amp from scratch, I'd use something like 400V for the screens and 800 for the plates. Eight 6550s running off these voltages is about what it takes to make 400W.
                "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

                Comment


                • #9
                  Thanks for all the replies. This is really helping me to understand the nuances of the power sections in these amps.

                  Another thing is that screen voltage directly translates to output power. The more voltage you apply (or the smaller screen resistor) the more power the tubes will deliver
                  Reducing screen voltage would reduce plate current if control grid bias remained the same, but if you increased bias voltages closer to zero simultaneously, wouldn't the plate current, and therefore the power of the amp, remain constant?

                  FWIW, Leo Fender seemed to idle his amps at about 55% of tube plate dissipation rating for 6L6 amps and at about 60% for 6V6 amps
                  Anyone know how hot Marshall or Hiwatts were run at? Will running @ higher/lower % of max plate watts affect the sound of the power stage?

                  Comment


                  • #10
                    if you increased bias voltages closer to zero simultaneously, wouldn't the plate current, and therefore the power of the amp, remain constant?
                    No. The output power depends not on the idling plate current, but on how much instantaneous plate current you can achieve at signal peaks. The control grid is driven as far positive as it can go at this point, usually about 0V, so the idling bias voltage has no effect.

                    Class-AB2 amps like the Fender 400PS use a beefy driver to force the control grids more positive than 0V, and this gives more power for a given screen voltage.
                    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

                    Comment


                    • #11
                      Ah, I understand. So before I go messing with voltage parameters, I have to know what the input signal amplitude (RMS?) from the phase inverter will be. Is there some way of determining that, without an o-scope and a func. gen.?

                      Comment


                      • #12
                        No, you don't. Your screens will take 500v+ (as they have to in many, many amps that run reliably, usually at just a volt or two lower than the plates). Bear in mind that you don't actually appear to have measured anything, you are quoting the schematic which may be different to what your amp runs at (though still within normal operating perameters).

                        The perameter you need to check is plate current, in mA, not negative voltage (-54 is not "icy", it's pretty well ball-park for the voltages that you are expecting, some tubes might require -60v+, varies with brand). Start at 30-35mA per tube.

                        The reality is that guys who build desirable tube amps are not the same guys that write the data sheets, most successful amp designs run plates & screens well over "suggested" voltages on data sheets.

                        On the other hand, if your amp is functioning correctly and you still don't like the sound, then the world is your oyster, but really try to have a think about the tonal qualities that you do & don't like & come back and ask how to resolve specific tonal issues, rather than analysing this voltage vs another.

                        I'm not trying to disuade you from learning about how tube amps function, but many guys make the classic mistake of getting a reasonably working amp, then read a tube data sheet & panic - thinking the sky will fall in because their 5881 runs at over 400v on the plate, or their 6V6 runs its screen 100v over the RCA data sheet. It won't, the voltages in your amp are pretty typical of 6L6 equipped guitar amps...amps like this largely run very reliably and sound good to most ears.

                        Comment


                        • #13
                          Originally posted by MWJB View Post

                          The perameter you need to check is plate current, in mA,
                          Yes, this is absolutely correct, and also allows you to calculate the idle plate dissipation. Just as a side note, however, this is one point where I tend to drift away from the herd. Most ; if not all ; gurus on these forums teach to measure the plate current off the cathode circuit ; whereas I will always measure the plate current ; off the plate.








                          -g
                          ______________________________________
                          Gary Moore
                          Moore Amplifiication
                          mooreamps@hotmail.com

                          Comment


                          • #14
                            So, the general consensus is to install test points (I usually use the 1 Ohm resistor on cathode method), replace the bias resistor with a pot of generally the same value, turn it down to a cutoff (or at least -54VDC) condition, and gradually turn up to taste (staying below 30 plate Watts, or 62.5mA in this instance). Is this the basic idea, or should I reveive another typed smackdown

                            Oh, and by the by, I guess my beef with this amp is its thundering low end, with little top end clarity/definition. I would be willing to trade some of the rumble for a little more punch. After checking the tone stack values and comparing them to my silverface via the Duncan TSL program, the response curves are quite similar, but they sound nothing alike, and my butcher is all grind with no bass, even though I've "silverfaced" the tone stack on that amp as well. So I was trying to find an explanation for the drastic tonal differences of three amps that use almost the exact same compliment of tubes/tone circuits. What do you think? Coupling caps, transformers, overall circuit design, year/zodiac sign of manufacture?
                            Last edited by apehead; 05-12-2009, 09:45 PM. Reason: confusion and general malaise

                            Comment


                            • #15
                              Bias resistor - which one are you thinking of?
                              I'd use a 47k pot with a 33k resistor in series to replace the 47k resistor (R103?).
                              30 watts - that's the maximum disspation,which is too high for it to be sitting at all the time. The received wisdom is to set bias so that plate dissipation is between 50-70% of max.
                              Remember that if you're measuring it at the cathode you will also be including the screen current. Also the plate voltage will drop as the current goes up. Peter.
                              My band:- http://www.youtube.com/user/RedwingBand

                              Comment

                              Working...
                              X