4x 6AG7 Amplifier

Discussion in '"Boat Anchor" & Classic Equipment' started by NW2K, Feb 11, 2019.

ad: L-HROutlet
ad: l-rl
ad: Left-3
ad: L-Geochron
ad: Left-2
ad: L-MFJ
ad: abrind-2
ad: l-BCInc
  1. K9MOV

    K9MOV Ham Member QRZ Page

    Those were the days. We tried the modified 1625's but the C.E. 10a was just not enough power to drive them in GG.
    73, Glen
  2. NW2K

    NW2K Ham Member QRZ Page

    The power supply is done. Variac at the front, cap input or choke input can be selected, up to 1300vdc/300ma.
  3. NW2K

    NW2K Ham Member QRZ Page

    Here is a picture of the grounded grid parallel 6AG7 amp, before connections for input, power and the output link. I allowed for six 6AG7 tubes. I paid $1 for each tube and the tubes were not new.

    6x 6AG7.jpg

    All of the cathodes are tied together. All of the suppressor grids, screen grids and grids are tied together and grounded.

    Drive is supplied by a Heathkit AT-1 and is connected across a 2.5mH choke, which goes from the cathodes to ground. I used an 80m crystal, and the output of the AT-1 was about 8w. HV is supplied through a 2.5mH choke and I used a single parasitic suppressor in the plate circuit.

    Tank uses 13T of THHN #14 close wound around a small piece of 1.3" OD PVC pipe. It measures a bit less than 4uH. The link is 2.75T close wound around the cold end of the tank coil. I used a series variable cap with the link.

    A guess is that the plate impedance is somewhere around 1500 ohms.

    The amp draws about 15ma at idle. It's not fully cutoff, but nearly so.

    The power supply that I used is inadequate for the task; it's only good for 850vdc/250ma.

    I measured input VSWR as a function of the number of tubes, at operating power. Oddly enough....best match was for four tubes. About 1.4:1. I'm using only four tubes right now.

    More work needs to be done to better understand it. However, first measurements with 8w of drive gives 70w output into 50 ohms with 850v on the four plates and 250ma of cathode current. Efficiency is about 1/3. A quick look at the output signal on the scope indicates that it is "pretty clean." It sounds as good as the AT-1 on the receiver. But...I need to look closer to gain a better understanding of what it is doing.
    Last edited: Mar 20, 2019
    KE5OFJ, K9MOV and KA0HCP like this.
  4. W9BRD

    W9BRD Ham Member QRZ Page

    That reminds me of a line in the first article by Larsen E. Rapp, WIOU ("Putting Dynamic Prognositication to Work", April 1941 QST): "With 900 V on the plate [of the 6L6], plate current will run around 240 mA, for several seconds."

    This design is idiotic. The phrase "careful testing" used in conjunction with a circuit that abuses its components in this way is an oxymoron.

    Just a few years later, TV sweep tubes and regulator tubes more capable of standing up to such abuse were available in forms that included "separate beam-forming plates" designs. (The authors of the CQ article explain why this is important for grounded-grid operation.) To anyone considering building the fundamental circuit on which this publishing mistake is based, I recommend replacing its 6AG7s or 6CL6s with xJN6s or xKV6As, where x is the heater voltage. Especially the variants with heater voltages other than 6.3 or 12.6 can be pretty cheap on Ebay (17JN6, 17KV6A, 22KV6A). The xKV6A is especially cool because it was intended for use as a voltage regulator tube rather than a sweep tube; it works just fine as an RF power amplifier. The 25 W output homemade CW transmitter I use for everyday operation uses a neutralizated 17KV6A in grounded-cathode driven by the pentode of an 8CX8. (I recommend Sylvania xKV6As over RCA equivalents; the Sylvanias have physically larger plates that can, I suspect, take more temporary overload than the RCAs. The tubes I've named are single-ended; you can locate double-ended [plate cap on top] that also have separate beam-forming plates.)

    Fun point for discussion: For a given output power, why does a tube operated grounded-cathode require much less driving power than the same tube operated grounded-grid?
    N2EY likes this.
  5. W9BRD

    W9BRD Ham Member QRZ Page

    Second point for discussion: So why don't we see grounded-grid RF power amps that operate in class C?
    N2EY likes this.
  6. W9BRD

    W9BRD Ham Member QRZ Page

    If you get down to the detail of checking the grid currents of the tubes, you may find them significantly overloaded even in this respect. A tube's grid has far more control over its plate current in grounded-grid than its screen, is therefore much easier to drive positive than the screen; the screen is intended to operate positive relative to the cathode.
    N2EY likes this.
  7. K9STH

    K9STH Ham Member QRZ Page


    Especially before World War II, more than a few amateur radio operators ran their transmitting tubes way beyond published specifications including the plate voltage. Several of these "olde tymers" told me that they had to send CW at a high wpm rate because a short "dah" would turn the tube plate red hot and an long "dah" would melt the plate! :rolleyes:

    Glen, K9STH
    K9MOV likes this.
  8. AA5CT

    AA5CT Ham Member QRZ Page

    Are you familiar with what is termed "grid-leak" bias?
  9. W9BRD

    W9BRD Ham Member QRZ Page

    Yep, but that's not why I asked a question I already know the answer to. Whether a tube class C amplifier is biased by means of a fixed negative voltage or grid-cathode rectification plays no part in the answer.

    So I'll restate the question with a bit of preamble:

    We don't see grounded-grid amplifiers run in class C. Why?
    N2EY likes this.
  10. W9BRD

    W9BRD Ham Member QRZ Page

    I am aware of the history.

    RCA came up with "Intermittent Commercial and Amateur Service" ratings for transmitting tubes as a result of such practices (and to compete with manufacturers that played faster and looser with specs for the ham market). An even more life-endangering spec, Intermittent Mobile Service, exists; all three specs--Continuous Commercial Service, ICAS, and IMS--are all well-defined, as are the expected tube-life consequences of using a tube within them.

    So, another question for discussion: Let's say that overheating a tube's grid doesn't permanently damage the tube through thermal runaway. (Grid emission can do this.) Let's also say that overheating a tube's plate to the point where it glows when it shouldn't glow somehow doesn't damage the tube permanently. Let's also say that we don't overstress a screen-grid tube's screen. (Which grounding all the grids of a screen-grid tube pretty much assures, since in that case the screen won't be operating at anything near a positive-voltage equivalent that could stress it.) What remaining critical limitation remains?
    N2EY likes this.

Share This Page

ad: SuperQSO-1