There's no connection here, is there?

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by Leonc, Feb 1, 2012.

  1. Leonc

    Leonc Wild Gear Hearder Gold Supporting Member

    Messages:
    17,183
    Joined:
    Jan 4, 2002
    Location:
    Upstairs, L.A.
    See V6, for tremolo in lower left of this schematic. Seems to be somewhat reminiscent of the Vibrolux 6G11 tremolo, though not identical.

    [​IMG]

    Check out the grid on V6 pin 2. There's a .01uF cap that connects to another .01uF cap. That second .01 does NOT connect to the .022uF cap coming off the plate (pin 1) does it? I assume it just connects to the 2.2M dropping resistor coming off the B+ right?

    And I assume the .022 (pin 1) just connects to the 220K resistor that's on the 3rd terminal of Rate pot (and to the the grid on pin 7, etc.). Is that what it looks like to you?
     
  2. diagrammatiks

    diagrammatiks Member

    Messages:
    3,030
    Joined:
    Jan 6, 2010
    that looks like a conncetion dot to me
     
  3. Leonc

    Leonc Wild Gear Hearder Gold Supporting Member

    Messages:
    17,183
    Joined:
    Jan 4, 2002
    Location:
    Upstairs, L.A.
    Yeah, it kind of looks like that but:

    1. it does NOT connect in the 6G11 (which this tremolo circuit seems to be modeled after) and moreover

    2. it would mean that we're bringing B+ to both the grid and plate on that first triode in V6 and that doesn't make sense to me.
     
  4. Jerry Glass

    Jerry Glass Member

    Messages:
    880
    Joined:
    Feb 11, 2010
    Location:
    Fort Wayne, IN
    Hey Leon!

    Actually that has to be connected or you don't have a complete oscillator.

    The two schematics are drawn a little differently; the 2.2M resistor that folds back to the power supply is performing the same function as the 1M that connects to the cathode in the 6G11.

    I see a ground reference but I don't see any grid connection to B+.
     
  5. DT7

    DT7 Member

    Messages:
    2,799
    Joined:
    Apr 28, 2009
    I think he must be referring to the cathode-follower grid connected to the plate of the other half of the triode.
     
  6. Leonc

    Leonc Wild Gear Hearder Gold Supporting Member

    Messages:
    17,183
    Joined:
    Jan 4, 2002
    Location:
    Upstairs, L.A.
    Hey Jerry - happy belated new year!

    Thanks man. Looks like you've saved my bacon from my own bad self again.
     
  7. reaiken

    reaiken Member

    Messages:
    1,798
    Joined:
    Jan 4, 2002
    Location:
    Greenwood, SC
  8. Leonc

    Leonc Wild Gear Hearder Gold Supporting Member

    Messages:
    17,183
    Joined:
    Jan 4, 2002
    Location:
    Upstairs, L.A.
    Hey Randall! How goes it?

    You know, in typical novice-dufus fashion, it didn't occur to me that if I rearrange this tremolo circuit, it's really the same kind of phase shift oscillator you described on your site. I could see similarities in this and the 6G11...but that's just because they're drawn similarly. So if you rearrange this schematic, you can orient like this, right?

    [​IMG]

    Of course, my first problem recognizing this was that it wasn't clear to me whether or not the B+ coming in via the 2.2M resistor was actually connected to what are C1 and R1 in the drawing above.

    But anyway, if I understood that part of your article, you're suggesting that you can pretty much eliminate the slowness in the start up of the tremolo (once the amp's warmed up) by using a large value resistor from R2 to the power supply and connecting the tip of the footswitch jack at the junction of the two resistors, yes? Pardon my ignorance...but would that wind up looking like this?

    [​IMG]

    And please pardon the Neolithic-cave-dwelling nature of my drawings. :dunno
     
  9. reaiken

    reaiken Member

    Messages:
    1,798
    Joined:
    Jan 4, 2002
    Location:
    Greenwood, SC
    Yes, by putting a large value resistor to B+, you are setting that node voltage to B+ (by charging up the caps), then, when you click on the footswitch and pull that end down to ground, you introduce a nice high voltage transient into the circuit which "kick-starts" the oscillator, guaranteeing a fast startup. All oscillators require a certain amount of noise to get them going. If there is a large gain, it only takes a tiny amount of noise to get things going. In the case of a phase-shift oscillator, which has a very low gain of around 29, it takes a much swifter kick in the grid to get them going.

    There's an old engineering saying: "Oscillators don't, amplifiers do."

    Actually, on looking closer at that circuit, the 2.2Meg is not properly configured as a startup resistor. Either it is a mistake in the design, or is just drawn wrong. There is no need to bias that particular node to a positive DC, and that configuration won't help startup at all. I simulated it in PSpice, and sure enough, it won't start up until about 8 seconds after the footswitch is opened as drawn, but if you move the 2.2Meg to the top side of the 1Meg resistor going to the footswitch, it starts right up. The good news is that real-world circuits have a lot more noise in them than simulated circuits, so even if it is connected wrong, it may still start up.

    In your second drawing, move the 2.2Meg connection to between C2 and C3 instead of between C1 and C2. You don't need the 10Meg resistor.
     
    Last edited: Feb 2, 2012
  10. Jerry Glass

    Jerry Glass Member

    Messages:
    880
    Joined:
    Feb 11, 2010
    Location:
    Fort Wayne, IN
    Happy belated New Year to you too! Sorry about the quick and dirty answer but it was bowling night and I could hear the beers calling my name...I think Randall gave you far more detail that I was prepared to do anyway. I think you're in good hands!
     
  11. Leonc

    Leonc Wild Gear Hearder Gold Supporting Member

    Messages:
    17,183
    Joined:
    Jan 4, 2002
    Location:
    Upstairs, L.A.
    Interesting! So you're suggesting that the original schematic has the power supply in the wrong place, correct? (I.e., I think my initial drawing accurately replicated the original drawing, yes?) If I understand you correctly, you're suggesting something like this will work:

    [​IMG]


    No worries mate. Your input is always helpful and most appreciated!
     
  12. Keyser Soze

    Keyser Soze Member

    Messages:
    1,477
    Joined:
    Dec 14, 2008
    Location:
    Johnson city, TN
    That newest drawing looks like it should work (still needs B+ on the plate though.)

    Merlin offers a slightly different approach to supplying a transient - one that doesn't involve having DC in the foot switch.

    http://www.freewebs.com/valvewizard1/trem1.html

    In his version you replace that 2.2M to B+ with the foot switch to ground and connect the bottom of R2 to the top of a cathode bypass cap on the oscillator tube.
     
    Last edited: Feb 4, 2012
  13. Leonc

    Leonc Wild Gear Hearder Gold Supporting Member

    Messages:
    17,183
    Joined:
    Jan 4, 2002
    Location:
    Upstairs, L.A.
    KS - great link!! I wonder why DeArmond didn't use a cathode-bypass cap in this circuit?
     
  14. reaiken

    reaiken Member

    Messages:
    1,798
    Joined:
    Jan 4, 2002
    Location:
    Greenwood, SC
    The cathode-supplied transient Merlin shows (which is the same as used on the Fender 6G11) is obviously a better approach than moving the resistor to the footswitch node, because it doesn't put a large DC on the footswitch. Another method (as mentioned in my paper) is to use a larger resistor to B+ and move the footswitch to the junction of the two resistors instead of the bottom of the second resistor. The voltage divider formed by the two resistors will reduce the DC to a much lower value. The Fender circuit is still better, though, because if the bottom resistor fails, there will still be a large, albeit safely current-limited, voltage at the footswitch terminal.

    In retrospect, I suspect there is enough noise on the B+ to provide the kick-start it needs as originally drawn. There will usually be several volts of 120Hz sawtooth-wave ripple on the B+ at that node, which will couple into the oscillator path to supply the startup transient. It will still start up slightly slower than a good DC kick, but within a second or so (about the same as the lower-voltage Fender circuit), and definitely much faster than no kickstart.

    The capacitor is not needed on the DeArmond because the circuit as shown has enough gain to oscillate without it. Remember, you only need a gain of 29 to sustain oscillations, and any more can contribute to waveform distortion. Plus, it saves the cost of one capacitor, so the circuit is effectively "Muntzed"...you have to marvel at the ingenuity of the old designers in coming up with the simplest, lowest cost circuit to do the job.
     
    Last edited: Feb 4, 2012

Share This Page