High-Performance Regenerative Receivers

Discussion in 'Homebrew and Kit Projects' started by G3EDM, Sep 27, 2021.

ad: L-HROutlet
ad: l-rl
ad: Left-3
ad: Left-2
ad: abrind-2
ad: l-BCInc
ad: L-Geochron
ad: L-MFJ
  1. G3EDM

    G3EDM Ham Member QRZ Page

    I'll kick off this new thread with a long description of my 7-tube regenerative receiver that uses subminiature tubes. For those not familiar with these, they are about 1 inch (2.5cm) long and not much thicker than a pencil. This technology had a brief heyday in the late 1940s and early 1950s before the widespread adoption of semiconductors. Because they are tough and also diminutive, they were used for a while in military applications such as missile guidance systems. Raytheon was a principal manufacturer of these. You would also find them, for example, in early hearing aids and small, transistor-sized radios (except they were hollow-state!).

    Here's a schematic of the "guts" of my receiver: RF stage, Detector, and 1st AF.


    The receiver was built in modular fashion on wooden "bricks" cut from American 2x4 timber. It is only in the past few days that I reconfigured it, sawing the original large bottom plank in half and moving the power supply outboard.


    The receiver is designed to be dismantled in a few seconds for servicing and modification. There are no screws to remove and the wiring between the top and bottom deck is long enough that it does not have to be disconnected when the decks are un-stacked.


    The top deck contains the elements shown in the schematic at the top of this post.
    • Top left, RF stage.
    • Center and top right, Regenerative Detector.
    • Bottom right: 1st AF stage.

    The bottom deck is for audio.
    • CW audio filter at bottom, nearest the front panel.
    • Push-pull 2nd AF stage, to drive a speaker or modern low-impedance headphones, at top.

    The receiver is powered by a large bank of rechargeable batteries. To save space on the station desk, this power supply has been moved to a high shelf in the shack closet.


    That's enough for this introductory post. Next: More detailed descriptions of each stage, one post per stage.

    A shout-out to @W9BRD. I believe you may be experimenting with sub-miniature regens too....

    73 de Martin, G3EDM
    Last edited: Sep 27, 2021
    KI4ZUQ, KC2SIZ, W8BE and 1 other person like this.
  2. G3EDM

    G3EDM Ham Member QRZ Page

    RF stage

    This is a tuned RF amplifier. While these are not strictly necessary, or even common, in a regenerative receiver they have several advantages:
    • Ironically, attenuation of strong signals, which is needed because regenerative receivers are prone to overload. By turning the RF Gain way down, or by de-tuning the input tuned circuit, some control of strong signals is achieved.
    • Isolation of the detector from the antenna; regenerative receivers are prone to shifting frequency even for something as simple as a wire antenna being blown around in the wind.
    • That isolation also serves to prevent the receiver's local oscillator from become a "transmitter" and inadvertently radiating a signal.
    For the schematic, see post #1.


    The RF stage is tuned using the small vernier dial shown at the top-left of the receiver in the first photo of post #1.

    73 de Martin, G3EDM
  3. G3EDM

    G3EDM Ham Member QRZ Page

    Regenerative Detector

    For schematic, see post #1. This is a standard design but does include a "regeneration throttle" (the variable capacitor at right of picture) which is a major aid in "taming" a regenerative receiver, in addition to the standard Regeneration potentiomer (knob at the left). The knob at top-middle is for setting the plate voltage.


    When I first built this set, those two large chokes (they look like small transformers), 60 Henrys each, were mounted in a more spacious manner some distance away from the detector. The one shown at the top is the detector plate load. (The one at bottom right is the plate load for the 1st AF.)

    When I reconfigured the set to make it smaller, I initially mounted the chokes bang in the middle of the circuitry, where you see those empty holes in the wood, centre of picture. The result was a horrible regenerative howl, almost supersonic. To get rid of the howl, I have moved the chokes to the edges of the set and downwards. The howl went away. The placement of those chokes turns out to be absolutely critical, and I may have to play around with that aspect of the circuit some more.

    (For the record: When I was working on this set over the past week, I discovered to my dismay that my initial build of the detector six years ago was completely wrong. Without going into details, I had basically hard-wired a nasty feedback loop that was creating "images" every 15 kHz or so. It is only in the past few days that I finally corrected the wiring so that it now matches the schematic! That "images" went away, and receiver gain improved dramatically.)

    73 de Martin, G3EDM
  4. G3EDM

    G3EDM Ham Member QRZ Page

    1st AF stage

    For schematic, see post #1.


    (The PCB board on the left is for the muting relay.)

    This stage is mis-behaving. It is giving out a near-supersonic howl, some of the time. I tried to solve this by displacing the 60H choke load but it made no difference. I now know that this particular howl originates in the AF stage because it is still there even when the muting relay is engaged, i.e., with no input to the stage at all. So more work is needed here.

    73 de Martin, G3EDM
  5. G3EDM

    G3EDM Ham Member QRZ Page

    CW Audio Filter

    This is currently broken. Sometimes, inexplicably, it starts working again but most of the time all it does is completely mute the receiver when switched in. Could be a loose contact or a cold joint somewhere. I may also replace the 1AD4 tubes with 1AG4s to get more gain.

    The filter does work rather well, when it works.... Unfortunately, fixing it is low-priority because of more urgent issues with my transmitter and other gear.



    73 de Martin, G3EDM
  6. K6CLS

    K6CLS Ham Member QRZ Page

    wow! I like this.
  7. G3EDM

    G3EDM Ham Member QRZ Page

    This is a classic push-pull circuit. In the past few days I followed @W9BRD's advice and replaced the original 1AD4 tubes with 1AG4s which are more suitable in a power-amp circuit. Sure enough, I now have a stunning 70mW or so. Or do I? Each tube is rated for 35mW. Anyway, at full gain, the audio is much louder and the speaker is now fully usable, whereas before, it was way too quiet. (There was always enough volume in the 'phones even with the 1AD4s.)



    73 de Martin, G3EDM
  8. G3EDM

    G3EDM Ham Member QRZ Page

    The power supply is as follows, from right to left. All the batteries are rechargeable.

    A reminder: Rechargeable batteries provide a lower voltage than what is says "on the tin". So for instance, a "1.5 volt" cell normally provides more like 1.2 or 1.3V.
    • Two C-cells in parallel providing about 1.2V to the filaments (the "A Battery").
    • Seven "9 volt" batteries in series providing 52V for the B+ (the "B battery").
    • Four AA batteries in series providing 4.8V for the grid bias (the "C battery").

    I have two complete sets of batteries, with one of them kept fully charged, ready to swap into the supply when the batteries get low.

    In the longer run, if this receiver ends up being in use for a long time I will build what in the old days was called an AC-powered "battery eliminator" to save time and space. There is enough room left in the bottom deck of the receiver to build this inboard.

    73 de Martin, G3EDM
    Last edited: Sep 27, 2021
  9. G3EDM

    G3EDM Ham Member QRZ Page

    OK that's it for the initial description of this set. I really just wanted to kick off a general discussion about regenerative sets, anchored around the example of my rather odd contraption.

    For more on modern regenerative receiver design I recommend the articles and building projects published in the 1990s and written by Charles Kitchen, N1TEV. His projects are solid-state but his points apply equally to hollow-state designs. Kitchin is among those who believe that regenerative receivers are not just "toys" or "beginner's receivers".

    Here's a link to one of the articles: http://www.arrl.org/files/file/Technology/tis/info/pdf/9811qex026.pdf.

    As for me, I am keeping an open mind and will decide on the usefulness/performance of regenerative receivers when I have had a better chance to use this one on the air. I have already scored several QSOs using it, but it was only in use for a couple of days before the recent complete revamping. We shall see....

    (A "better" receiver, i.e. a superhet, is in my near-term future but for the moment, this set is what I have! Might as well explore its possibilities....).

    73 de Martin, G3EDM
    Last edited: Sep 27, 2021
  10. W9BRD

    W9BRD Ham Member QRZ Page

    Sound like you had created an Armstrong oscillator, or a tuned plate, tuned grid oscillator, at audio by means of coupling between the chokes.

    I happen to think that a 60-H plate choke is practically unnecessary in the detector plate, and likely also in the "first audio" amplifier if it's being used as a voltage amplifier to drive the push-push 1AG4s. Resistors would suffice as replacements, as in a resistance-capacitance (RC) -coupled amplifier (which the plate circuit of a screen-grid detector essentially is at audio). As a first approximation of what the replacement resistance should be, the formula for the reactance of an inductance --

    reactance (X, ohms) = 2 times pi times frequency (f, Hz) times inductance (L, henrys)

    -- where frequency is your favorite CW copying pitch, will suffice.

    So, for instance, at my fave max CW pitch (500 Hz):

    X = 2 x pi x 500 x 60 = 188495, or (looking to the nearest standard value) 180 k

    For decades, I've commonly used the value 150 k. It works fine. 100 k would work fine, too.

    That much resistance in between the B supply and tube plates might look as if it will shut the stages down by starving them for supply voltage, but the plate resistance of a small-signal pentode is sufficient high, and the plate current drawn by such tubes sufficiently low, that RC coupling can and does work well.

    With at least one of the two chokes replaced with a resistor, plate-to-grid coupling by means of induction between two chokes can't happen. I'd replace the detector-plate choke first. (As for regenerative-receiver purists who insist that screen-grid-detector plate loads must be big audio chokes, go check the schematic of the Pilot A.C. Super Wasp.)

    As side benefit, the impedance of a resistive plate load is -- at the audio frequencies we expect to hear during shortwave reception of CW, speech, and music -- constant v frequency, whereas the impedance of a choke increases with frequency.

    You might also want to try bypassing the screens of your 1AG4 push-pull AF power amplifier to common for audio -- say, 1 to 4.7 uF for starters. The screen of a screen-grid amplifier should be at (or closer to than not) at ac-common for whatever signals are handled by the stage.
    Last edited: Sep 27, 2021
    G3EDM likes this.

Share This Page