No announcement yet.

Tube Amp myths...

  • Filter
  • Time
  • Show
Clear All
new posts

  • #61
    There is nothing particularly special about the design of the Dual Rec's power section. It is pretty much like 90% of other amps except for the feature of disconnecting the NFB in modern mode ( the amp does have an NFB that is engaged for the other modes ), which does still leave some components in the PI circuit. Not having messed with one enough to know how it acts, I can't or won't say why it does what you say it does. Obviously, not 100% of amps will react the same way, depending mostly on their topology. The Mesa DR does have stuff left in the PI circuit, so that could be a factor as to why it does what it does?

    If you read what I said about how the feedback loop works, you will see I said exactly what you have. I did err in saying low pass filter, where what I meant was Hi-pass filter to allow the highs into the PI. I did iterate the function correctly with the exception of that. That was an oopsie.

    The FM curves can cause the perception of an amps sound to be different at different volumes, but would work in the opposite direction to what most would complain about an amp. I would argue that most people with a 50 to 100-watt amp complain that at bedroom levels, it sounds thin, bright, and anemic. The " loudness " button that comes on 99% of consumer electronics is an FM curve eq implementation that boosts the highs and the lows to create an equal perception of sound at lower volumes. The DJ smile if you will. So if your amps is too bright at bedroom levels, then it is because the amp is that bright, or you could say that it has very little low-frequency energy. Another consideration to make for the FM curves is that the fundamental frequencies for the guitar are just about equal to the higher frequencies of the guitar's harmonics. Meaning that if the amplifier was fairly linear, you would perceive the highs and lows about equal at bedroom levels. The equal loudness contours show us how much actual SPL at X frequency is required to hear them equally at X target SPL. This seems to support my theory that the feedback loop in guitar amps is there because, without it, they would sound less pleasant. Negative feedback loops are used in audio amplifiers for the purpose of making the power amp more linear. It helps even out the frequency response. Since the feedback loop requires forward gain to work effectively, it explains why at bedroom levels ( where you are dissipating 10ths of a watt perhaps ) the feedback loop is not as effective.

    I do agree that the cabinet does have an effect on the whole of things, but that is a completely different subject. The reason I wouldn't clump it into tube amp myths is because it is a fixed variable. If you only have one cabinet and 5 amps, that cabinet is going to react the way it does regardless of which amp you plug into it. I feel that cone breakup occurs later than most people think it does. Of course, only a scientific test would weed that out. Despite the result, you can't really change that outcome. If you operate the amp loud enough to start creating PI distortion, Power tube distortion, and speaker distortion, then all you can truly say is that you have a LOT of distortion. If that is the sound you are going for, then how you got there is irrelevant.


    • #62
      You were right in saying the presence knob is a low pass filter. It lets low frequencies pass while dumping high frequencies to ground, so that those high frequencies don't cancel out frequencies to one side of the phase inverter. Remember, whatever frequencies are present in the NFL are the frequencies that won't be in the final sound, so everything is reversed from what's intuitive. The NFL frequency response shouldn't be too interactive with volume on an amp with a traditional MV knob (which comes before the PI), because as you turn volume down, you cancel a smaller waveform going into the PI out with a smaller waveform from the power amp (the same is NOT true on an amp with a PPIMV, which is one of the major drawbacks of that design; with the PPIMV down, you have a very midrangey sound, precisely because the NFL isn't being effective in linearizing the sound, so the frequencies are overwhelmingly those that a guitar naturally produces, which are all in the midrange).

      The reason a dual recto sounds so different with the MV up or down in modern mode is precisely because of its lack of an NFL. One of the side effects of an NFL is that it makes it harder to overdrive the PI because as you apply more volume, you get more negative feedback. If you've got access to a Marshall 2203 or 2204 and a place to play it loud, you can see this in action easily -- play it with the MV dimed, preamp gain at about noon, and presence knob all the way off, and see how little distortion you get. Then dime the presence knob, and try playing it. You'll find that you get a ton more distortion and saturation. By removing the NFL in modern mode, the recto gets a bunch of overdrive on that stage, giving it a lot more distortion (and volume!).

      There's a lot more going on with cabs and volume than just speaker distortion. As you turn up more, long before the point where the speakers themselves distort, the acoustics of the cab will produce more and more low end, especially if it's a closed back. The lack of low-end at low volumes you describe is largely due to this, not the amp-head itself; you can use an attenuator to control your volume, and you'll still notice the same lacking low-end as if you had turned the amp's volume down (this is why a lot of attenuators have bass-boost switches). Also, speaker cone excursion happens sooner than you might think -- on a 5150/6505 into a cab loaded with V30s, it happens on about 2 on the MV knob (you can easily verify this by shining a light through the grille cloth, a good video of this can be found here: ). Furthermore, cabs aren't really just a passive box -- because their impedance is variable, they're interactive with the output section of the amp itself, meaning that you can't just say that it's a fixed variable. One of the purposes of an NFL is to isolate the amp's power section from the cabinet's impedance curve to a degree; an amp with stronger NFB will react to different cabs differently than an amp with weaker (or no) NFB.

      Re: the importance of power distortion even on "modern high gain amps", here's a useful video: Notice how the SLO100, with its supposedly "clean power section", is already adding a ton of distortion when he turns it up to 3, a.k.a. "roughly band rehearsal volume".

      Finally, the idea that "distortion is distortion" is waaaaaaaay wrong. You'll never get the sound of The Kink's "You Really Got Me" with the distortion from any tube head. Power distortion and pre-amp distortion have a different feel and sound -- again, you can verify this easily on a 2203 or 2204.


      • #63
        The reason I say that the speaker & cab is a fixed variable is that its impedance curve et all is set. The impedance curve doesn't change when you plug a different amp into it. The NFL as you say does in fact help even out that interaction and also does reduce distortion in the PI, again as you say. But if you plug an amp that doesn't have a feedback loop, the speaker/cab will do what the speaker/cab will do; the resultant sound is the some of all parts. In either case, one scenario has what is essentially a late circuit tone control, while the other doesn't. The speaker/cab doesn't suddenly change characteristics when a different amp was plugged into it.

        The SLO MV video doesn't show where that distortion is created. The SLO is still a pre-phase MV, so that extra grit is very possibly coming from the PI. When he went to 7 on the dial there was a shift in tone, which could be the point where the power tubes actually start to distort breaking down the feedback loop. Without scoping the signal coming out of the PI and the signal coming out of the power tubes, we cannot definitively say where that extra grit is coming from.

        As to cone breakup and making the cone move, if you present the speaker with enough low end it will move regardless of how loud it really appears to be. As I mentioned before, most guitar cabs are not made with any real Theil Small parameters being considered. When 4x12 cabs were originally designed, it was sized to what could be gotten out of a single 4x8' sheet of plywood. Don't even get me started with open back cabs. A speaker will break up sooner in an open back design, closed-back designs do produce more bass due to their sealed design and will have less excursion. Most guitar speakers have very little excursion or X-max limits, and most cabs they are placed within are too big to control the X-Max of the speaker fully. The resonance of the cab is mostly part of its design. There are in most guitar cabs 4 surfaces that are adjacent to one another making them standing wave generating machines! Getting the speaker loud enough to where you can hear those standing waves over the speaker is not hard to do. This is not necessarily cone breakup yet. A term you may or may not have heard of is cone cry. It is a VERY evident distortion that is independent of the amp that sounds more like a buzz. The only real way to find out if you truly have cone breakup is to use measurement software and see where the speaker goes non-linear. Although you will know when you have it, it sounds more like a crunch that occurs over the top of the core sound. You have to be surprisingly loud to get any of those artifacts to show up. Which one happens first is anyone's guess. While I have not A/B'd an SLO to a 5150, if the so-called point of perfection is around 2.2 on the MV of a 5150, then the MV level of 3 on the SLO is probably of little use then. I think what these videos show is that guitarists are too loud.

        I think distortion is distortion, objectively speaking anyway. If the signal out is non-linear to what went in, that is the absolute description of distortion. How you get it is what makes a clean amp tuned up to 11 and a gain banger like the SLO with the MV at 1 different. They are both distorting in one way or another. I have a different outlook on sound though. I am a sound engineer by trade and have been playing guitar and performing for the past 27 or so years. I started on guitar and after becoming a sound engineer 17 years ago I learned real quick how problematic musicians were to themselves. My idea of sound is more about linearity, naturalness, and realism. I do what I can to portray a musician's character and sound to the rest of the world by augmenting it with my talent and skills so that the whole band is heard as a single organism.


        • #64
          I'm still going to use the standby switch on my JMP . I just don't like the idea of 550V across a 400V capacitor until the cathodes warm up.
          Last edited by NOGE; 01-24-2021, 09:58 PM.


          • #65
            The PI is a part of the power section; if the PI distorts, you've got power section distortion. Even on an amp like a Plexi, the power section distortion is nearly all from the PI, not the big-bottle tubes (which is why you can get good sounding distortion out of those amps with KT88s; anyone who has ever heard serious clipping of a KT88 knows it's not a sound you want, but most classic era Marshalls that were sold in the U.S. came with KT88s, and they sounded fine).

            3 on an SLO isn't far off from about 2.2 on a 5150. Remember, Soldano knobs go to 11 (and are mounted slightly crooked, so you can't trust "'o clocks" either, because the half-way point is slightly past noon), so it's 3/11 (0.27) vs 2.2/10 (0.22). Very minor difference.

            Saying "distortion is distortion" is a big oversimplification. There's no one on this forum that can't hear the difference between an almost square-wave style hard clipping or soft clipping, or we'd all just be using Boss MT-2s instead of tube amps. Most people here could tell the difference between symmetrical vs. asymmetrical clipping, even if they didn't know which was which. The "purr" of a Recto or SLO is immediately distinguishable from the "angrier" distortion of a boosted Marshall -- the texture is just obviously different because of how the waveform is distorted, in ways that aren't just the relative EQ of the amps. Then there's the spectre of blocking distortion, which *always* sounds bad in every case.