Gennies - Regenerative Receivers.

[KX4QC]

Ha! the first post here!
Regenerative receivers outperform almost any design for their low component count, and ease of build.
Their principle is so easy: take a little from the output of the RF amplifier and re-introduce it to the tuned circuit. so, the RF goes round and round, being more sharply tune and amplified with the same components. The result can be spectacular performance at the cost of somewhat inconvenient operating.
So what's not to like? While the tuned circuit, often just a simple LC tank, has its Q multiplied by the feedback, some RF has to be sent to the detector, or the wonderfully tuned signal would never be heard. This removed RF damps oscillations in the tuned circuit, degrading the tuning and RF amplification. The solution is often some sort of "detector drive" control to find the right compromise between adequate signal for the detector while preserving the whole reason for regeneration. AF amplification is no different in a regenerating receiver.
Charles Kitchen, the late N1TEV wrote extensively about regeneration for the amateur. A fine example is attached.

9811qex026.pdf

[KX4QC]

WBR with Corrections (3).pdf

The last post described Gennies with a "tickler" coil. Easy to see how it works, but as the construction details show, a different number of turns of the tickler is needed at different frequencies, needing complex switching or interchangeable coil assemblies.
Now if only the regeneration control could be made independent of frequency, bringing RF from after amplification back to the tuner, the radio shoild be much easier to operate and construct.
Here's a curious example of this approach, and also introducing a "bridge" type VCC tuner with a return to Major Armstrong's original oscillating detector.
But is it still truly a "Gennie"?

[KX4QC]

A friend complimented me as the kind of guy who could identify a regenerative circuit at a glance.
Hmm... I guess he thought it was easy!

So, here's one to try:
Ok it's the classic tube variety, but if you can put up with a hissing single gate FET instead, the principle still works.
The key points of this one are the RF from the plate is coupled L3 to the tuned circuit coil L2, so each time some RF is put back into the simple tuned circuit, making the selectivity sharper and sharper. The grid (gate) is kept at groud potential, so the trieode conducts on just half cycles, making a fine detector. Residual RF is kept out of the phones with the bypass C4, though it is likely that the high impedance phone windings offer a huge inductive block to RF anyway.
The volume is adjusted by changing the proximity of L2 - L3, and tuning by C2. Kinda clumsy, but doable.
It's a lot of fun to throw one of these simple circuits together for AM reception, and then tinker to make it better. Try it before the "improvements", and you'll be astonished how much performance comes out of so little. Ok, it's fiddly to adjust and so it lost out to the superhet in the 1930s.

gennie1.jpg

So what's wrong? The coils L1-3 are specific for each band, so there are at least six connections to plug/unplug or switch. The simple to understand but inconvenient component is the feedback "tickler" coil L3. Much better would be to arrange RF feedback to the tuned circuit some other way, not a frequency-dependent coil.

Ok, so we want to do it with bipolar, rather than a FET. The impedance now has to be a little different for the tank and tickler to keep the current-driven Tr1 happy. The principle is the same...

1965NovEISchematic-2017266332.jpg

Feeding some of the RF back for re-tuning and re-amplification make for high selectivity and high gain with minimal components. If the feedback is increases too far, the circuit oscillates and spews RF from the "receiving" antenna. Not good for the neighbors! If one adjusts the feedback so carefully that it is only just at the point of oscillation, CW can be received with a note not a hiss. One step more in fiddly adjustment and a sweet spot can be found for SSB reception, where the circuit oscillates just sufficiently to replace a missing carrier and allow detection of the resulting "almost AM" signal. Can it also do FM? Yes, but badly if detuned to one side of the median frequency.

Some gennie circuits use other means of feedback, rather than tickler coils, usually meaning that tuning ranges can be switched more easily. Think instead of negative feedback used to improve stability, positive feedback used at RF to include a tuned circuit.
There are better "gennies' in other posts here. Ok, superhets work as well or better and are easy to adjust, but I still find it astonishing what can be done with so little in the way of a gennie!

[KX4QC]

Here's an improvement on the gennie theme, with many of the shortcomings addressed, at the price of somewhat increased complexity.

Regen_4_beginners_RF.png

The original article is available to members:

95hb1768.pdf

So let's take a look. Q1 isolates the antenna from the rest, if it oscillates. Neighbors like this. There's also an optional AM trap to combat powerful AM stations of yesteryear when HF DX is being hunted. RF gain is controlled by Q2 bias. Q3 acts as more RF amp, with feedback to the tuned circuit by L3, detector with the gate kept at ground by the "grid leak" R5. Rectified AF is recovered downstream of the RF choke RFC2.
So what's C8 doing? A perfect tuned RF would have a very high Q, and in a gennie, the RF would be retuned to perfect selectivity. Good, as long as you don't want to listen to it; in practise, we take some RF out to the detector, killing some of the Q. The ideal gennie detector then, take only just enough RF to give adequate AF without degrading RF selectivity. The control that sets this compromise is often called detector loading or detector drive. Yet another knob to twiddle on an already fiddly receiver to operate!

[KX4QC]

Here's the venerable and popular Scout. Nice simple design, but maybe could do with adding a "detector-drive" control?

scout.jpg

The instructions for building are here:

Scout_Regen_v1.2.1.pdf

The original is described with an old commercial PCB. but this is so simple that Veroboard or an etch-resist pen PCB might do as well? Everything except this original board seems still currently available.

[KC4NYK]

Hi John,

Wishing everyone a Happy Father's Day and basking in the afterglow of getting my one transistor transmitter on the air and putting out a mind bending 300 mW! Add to that picked up on RBN 233 miles away in PA and you can feel the oscillatory love!!

Now for more "gnarly news", my latest foray down the dark rabbit hole of Regens and I am 0 for2 put in MLB lingo. So what I'm thinking is to have a "Regen Renaissance" of sorts like a Saturday morning workshop where interested parties, having been supplied with the proper BOM for a simple regen, and the wherewithal to creat a "Manhattan style" receiver but with added "Elmering" from - say - KX4QC!!

Does that "resonate"? We could do it at HackRVA where we have a classroom adjacent to a fully stocked electronics bench, most bench equipment and other accoutrements necessary for joy of regen.

Let me know and we can chat!

cheers,

Rob, KC4NYK

[KX4QC]

Here is the account of a first building project, and it works!




The circuit is available here: https://noeyedeer.org/viewtopic.php?p=12#p12 with original article describing design of Gennies in General.