Can you put KT66s' in a Blackface Fender Bassman?

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by Erik J. Rodriguez, May 7, 2020.

  1. gunslinger

    gunslinger Member

    Messages:
    2,499
    Joined:
    Mar 28, 2010
    Location:
    Louis XVI ville, KY.
    Why not just put the best 6L6s you can find in there? Maybe something like the new production Tung-Sols or JJs.
     
    Tone_Terrific likes this.
  2. HotBluePlates

    HotBluePlates Member

    Messages:
    7,982
    Joined:
    Mar 27, 2017
    It was a bit of a trick question, but your original answer implied the "ideal impedance" is different for KT66 and 5881. Yet your answer only gave a single range of impedance, not differentiated by tube type. Where I'm trying to lead you is the realization that if you're shooting for the same power output, and the tube types used have the same plate current capability, then the same OT primary impedance is "ideal" for both types.

    - The common knee-jerk is to assume that an OT primary impedance listed on a data sheet is "what a tube wants to see" or to observe the primary impedance used in a popular amp as an "ideal primary impedance." Really, each are only "ideal for the particular application observed" especially in the case of a data sheet that often has multiple conditions listed, with multiple OT impedances, and multiple power outputs. So the right answer is "there's no one right answer." ​

    Kuehnel's Bassman Book says the OT primary impedance is 4050Ω, but we could call that 4kΩ.

    You gave the right answer saying the ideal impedance is based on supply voltage, and the rest should be refined as, "tube plate current capability."

    - Power is Voltage * Current

    - The output tube plate swings from its idle value down towards 0v, but the tube needs some voltage on its plate to keep attracting electrons. The minimum plate voltage is where the "knee" of the plate curve happens, and plate current rapidly dives towards 0mA. Kuehnel's book implies we need to keep ~100v on the plate at maximum output.

    - Plate voltage swing from about 400v at idle down to 100v at maximum output, or 300v peak across the OT primary impedance.

    - How much plate current can 5881s or KT66s pull with ~400v on their screen (as found in the Bassman)? Neither data sheet gives us a curve for "G2 = 400v" but we can extrapolate. The 5881 sheet (bottom graph of page 4) shows when the plate is held at 100v and G2 moves from 250v to 300v that plate current increases from ~170mA to ~220mA (a 50mA increase). We might expect plate current to rise to 270mA when G2 is at 350v, and to 320mA when G2 is at 400v. The KT66 data sheet (top graph on page 7) shows a higher plate current capability of 250mA at 100v plate and 300v screen, but a similar ~50mA increase for 50v more on the screen. We can infer both tubes are capable of 320mA peak plate current (and perhaps more for the KT66).

    - Tubes receive a voltage input and have a resulting plate-current change; there is no plate-voltage change except that the tube's plate current is pulled through a resistance/impedance, which creates a voltage drop due to the current (according to Ohm's Law), and leaves the remainder of the voltage at the tube's plate.

    - Altering Ohm's Law, Impedance = Voltage / Current. We have 300v peak plate voltage and at least 300mA of peak plate current. 300v / 0.3A = 1000Ω. For a Class AB OT, the impedance seen by one side at peak output is 1/4 the total primary impedance, so 1kΩ * 4 = 4kΩ, or the usually-stated OT primary impedance for a Bassman amp.

    - RMS Power is 1/2 the Peak Power, and we've been dealing with peak figures for voltage & current. 300v peak * 0.3A peak = 90 watts peak, and halved that's 45 watts RMS. Sounds about right, no? Kuenhel goes on to show the effects of power supply sag and screen resistors, and finds after all that the amp delivers 39 watts RMS.
    You could "gild the lily" by trying to use a somewhat lower OT impedance to extract every last watt from the KT66s. Supposedly, they could deliver 350mA peak plate current, so what primary impedance might we have used for the 300v peak plate swing? 300v / 0.35A = ~857Ω -> ~3.4kΩ plate-to-plate. 300v peak * 0.35A peak = 105 watts peak, so 52.5 watts RMS assuming no sag.

    - 10 log (52.5 watts / 45 watts) = ~0.7dB ---> an almost imperceptible volume increase.
    Bottom-line On Bottom: There are lots of methods/rules-of-thumb to calculate the ideal load impedance for a tube. For pentodes/beam power tubes, the method I use most is envisioning the OT primary as a resistor. The voltage applied to the tube's screen sets the peak plate current capability of the tube, and the minimum voltage that must remain across the tube sets the peak plate voltage swing obtainable. From there, calculate the OT primary impedance (seen by one side of the push-pull stage) as Voltage / Current, and multiply by 4 for a Class AB power section (or multiply by 2 for a Class A power section). Power "dissipated by the primary resistor" is the power transferred from OT primary to secondary (minus losses and effect of sag) to drive the speaker. If we used the same tubes at some other supply voltage (and/or had a different power output target), we would arrive at a different answer for "ideal" OT primary impedance.
     
    Tony Bones, CoyotesGator and pdf64 like this.
  3. Tony Bones

    Tony Bones Member

    Messages:
    1,215
    Joined:
    Nov 1, 2014
    Location:
    beneath the moon
    And, on top of all of that, the advertised primary impedance of any particular OT assumes that the load on the secondary is exactly 4, 8, or 16 ohms, or whatever. We know that the load a speaker presents is pretty much never exactly as advertised so the load seen by the tubes is pretty much never exactly what you calculate.

    And then there's the inductance and capacitance of the primary winding which the tubes see in parallel with the load. And so on... It's an inexact set of conditions.

    In the end, all else being equal, 3.2k, 4k, and 5k transformers might sound a little different, but from a functional point of view any will work about as well as any other.
     
    HotBluePlates and pdf64 like this.
  4. HotBluePlates

    HotBluePlates Member

    Messages:
    7,982
    Joined:
    Mar 27, 2017
    And the "advertised primary impedance" of a nice round figure is almost never what you get when you do a voltage-ratio measurement on the actual part...
     
    Tony Bones and pdf64 like this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice