[Boatanchors] My Homebrewed Regen Receiver project, link to pictures

[Phil]

On my Beacon mail list it was requested that I post (actually re-post) 
the information on the tube type regen receiver I designed/built this 
past summer.  I thought it might be of interest to some folks here. . .
**************************************
Good Mornin' (afternoon for you) Václav

Thank you for asking about my receiver.  No manual, no schematic. When I 
designed/built it, I first sat down and figured out exactly what I 
wanted it to do, specifically be an above average regen designed for NDB 
DXing.  A real tool, not a 1 or two tube 'novelty' set that would sit on 
the shelf and not be used.  The ability to work other bands if desired 
was also on the list, which dictated the use of plug-in coils.  More CW 
selectivity than a regen usually has, and the ability to drive either a 
speaker or modern low impedance headphones were also desired.  BUT MOST 
IMPORTANT TO ME, I wanted the set to be based ENTIRELY on technology 
available to the experimenter in the 1930's, NO solid state active 
stages.  I DID use silicon rectifiers in the PS simply because the 
transformer I used wouldn't support rectifier tube filaments etc.

Then I drew up a basic block diagram, knowing that the exact circuit 
would depend heavily on what electronic components I could obtain. So at 
this stage I didn't draw any schematics, but came up with a basic list 
of parts that I would need.

As project cost was going to be expensive IF I had to buy everything 
new, I then went on a scavenger hunt to see what I could scrounge up.  I 
put a want list in our local ham radio newsletter, as well as some of 
the vintage radio and boatanchor mail lists I belong to, wanting to 
purchase old parts.  Things like tube sockets, power transformer, audio 
output transformer, tuning capacitor and so forth.

This paid off nicely in a quite unexpected way.  A man in Washington 
State offered me some adjustable 10H audio chokes, suitable for a 
'passive' audio CW filter.  Sent me three of them and they immediately 
became the components around which the entire receiver was designed.  
With capacitors they would form a parallel resonant circuit.  They were 
tapped allowing  the filter to have a low input impedance (plate) and 
high output impedance (grid of next stage). He also sent me an Emmco 
vernier tuning dial from the 1920's.

An almost local ham friend told me had a lot of stuff on my list, tubes, 
tube sockets, power and output transformers, old military style black 
Bakelite knobs, potentiometers, etc, free for pick-up. Needless to say I 
drove the ~30 miles to get them.

Now things were starting to come together (this was a long summer 
project, NOT one to quickly complete).   I took the stuff that I had 
gathered and laid them out on the workbench to get an idea of what sized 
chassis I would need.  For rigidity/stability a good aluminum chassis 
was important.  I ended up buying one that was 12x8 inches along with 
two aluminum bottom covers (one to use as a front panel).  I thought it 
might be a bit large for the project, but in the end turned out to be 
the perfect size.  At about $50 it was also the single most expensive 
item in the project, but I felt a necessity .

Once I had everything on hand I hand drew up the basic schematic of each 
stage, knowing that some of the circuits would be tweaked once I got the 
set working.

Next step was to set  all the major components on the chassis to figure 
out the best layout.  Important to keep the RF amp and detector stages 
as far as possible from the power supply (regen detectors are prone to 
hum pickup).

NOT a single hole was cut until I felt I had a good layout.  I then 
drilled all the holes for the tube sockets, controls, transformers etc.  
NOTE:  Not having any Greenlee chassis punches like I had as a kid in 
the '50s (they're VERY expensive now) I was told that a step drill like 
Harbor Freight sells is a VERY good, and possibly superior alternative.  
Reasonable in cost, effective, I now have five of them in different 
sizes, from a small 1/8" to a large 1-3/8".  I even sent a couple to 
another member of this list so he could punch some tube socket holes.

I even drilled the holes to mount the Emmco tuning dial.  NOTE: After I 
built the set  I found that even with a lot of tweaking, lubrication 
etc, with it's relatively small tuning knob the dial was a bit too stiff 
to allow my arthritic fingers to tune it for long periods of time 
(remember, I said this radio was being built as a serious DXing tool).  
So I ordered a couple used National Vernier Reduction Drives from 
surplus WWII equipment and installed one. MUCH better and smoother, 
meant I had to enlarge the mounting hole and relocate the tuning 
capacitor slightly. . .

OK, I won't go into detail about how each stage was tweaked as I went 
along.  But here's a description of each stage and what it does, 
following signal flow through the set.

1.  6AK5 tube, untuned Grounded Grid RF amplifier.  Not there for gain, 
but to isolate the antenna from the following detector stage. A regen 
detector has high gain and easily overloads.  The RF gain control is a 
Potentiometer in the cathode input circuit.  There is also a 580KHz 
wavetrap at the cathode as 580 KMJ in Fresno (a 50kW station about 60 
miles away) tended to bleed through on the high end of the NDB band.

2.  1/2 12AU7 dual triode.  Armstrong Regen detector stage.  The classic 
regen detector with tuned grid coil and a tickler coil in the plate 
circuit.  Some other circuits, such as a Colpitts or Hartley were 
suggested to me on the Regen mail list, but I settled on the classic 
Armstrong circuit.  Some folks have had problems with the circuit 
smoothly going into oscillation, but mine works VERY well (think it 
depends on the exact circuit parameters as to how smooth it is).  This 
one is also EXTREMELY stable, after a short warmup of a few minutes, in 
oscillating mode it NEVER drifts more than a hundred Hz or so.  NOT at 
all bad for a tube type free running oscillator!!!  Of course this is at 
LF, it would probably be a different story at HF and above frequencies.

3.  1/2 12AU7 dual triode.  Cathode follower loosely coupled to the 
plate of the Detector tube.  It's output feeds a BNC connector on the 
back panel so that an external frequency counter can measure the actual 
receiver frequency when the detector is oscillating.  NOT really out of 
the spirit of designing the set with '30s technology as it doesn't have 
any effect on how the set itself performs.  It's there to facilitate 
dial calibration or to quickly tune the set to a desired frequency.  I 
rarely use it in actual listening as it only works when the set is 
strongly oscillating, otherwise the counter just bounces around.

4.  1/2 12AT7 dual triode.  First audio stage, audio filter driver. Not 
a lot to say about this one, it does what all audio amplifiers do, and 
it isolates the detector stage from the audio filter stage.

5.  Passive audio stage.  This is comprised of a 6 position switch, 
three parallel tuned circuits resonant at 400, 600 & 1020 Hz, and two RC 
Low-Pass-Filters (treble cut-off).  First position bypasses all audio 
filtering, but I never listen in that "wide" position.  Most of my 
listening is in the narrowest LPF position (treble-cut), switching to 
one of the bandpass filters when needed.   The audio filtering is VERY 
effective in digging out weak beacons.

6.  1/2 12AT7 dual triode.  2nd audio amplifier stage brings the audio 
level back up after the losses introduced by the audio filters.  The 
volume control is located between the audio filter and this stage.  The 
set was intentionally designed with more AF gain than needed.  At first 
the volume control was between the detector and first audio stages, but 
hum from the audio chokes in the filter was objectionable, especially in 
the headphones.  Even though the chokes are in a shielded steel box, 
they still get 'some' hum from the nearby power transformer.  Changing 
the gain distribution in the set by relocating the volume control 
completely removed it.

7.  6AQ5 audio output stage.  Again, this section does just what it 
should, provide sufficient power to drive a speaker or modern headphones.

8.  0B2 Voltage regulator tube.  Provides a stable 105 Volts to both the 
RF amplifier and detector stages.  This is probably one of the 
contributors to the excellent stability of the receiver!

9.  Rectifier stage.  Uses 4 1N4007 diodes in a bridge configuration.  
My one concession to modern electronics in the set. The 'donated' 
transformer did NOT allow for a full wave tube rectifier, but I wanted 
full wave rectification as it is much easier to filter 120Hz than it is 
60Hz.  There is also a large DC filter choke in the B+ filtering circuit 
and VERY heavy filtering via the over-sized filter capacitors.  That 
along with the 0B2 regulator means that the DC supplied to the radio is 
essentially "battery pure".

10.  Filament circuit.  As I mentioned, regen receivers are prone to hum 
and noise, often coupled from the filaments of the tubes.  Part of the 
problem is that one side of the filament circuit often uses the chassis 
as a conductor, a bad design.

In this set I floated both sides of the 6.3 Volt filaments above ground, 
wiring with twisted wires.  At the 12AU7 tube, both sides of the 
filament line is bypassed to ground with capacitors.  In addition, there 
is a 100 Ohm pot wired across the filament lines, and the wiper is 
connected to the chassis ground.  This is a "hum balance" control and is 
common in many older audio amplifier designs.    The result is that 
between the well filtered B+ and the design of the filament circuit, 
there is NO audible hum in the receiver at all (outside of the 
aforementioned bit from the audio filter chokes, now almost completely 
gone).

FINAL NOTE:  I've been asked a number of times if I'd draw up a 
schematic, and somebody suggested using "tubepad" software to do it.  
I've wanted to, need to locate the software, THEN FIND THE TIME TO DO IT

Pictures of the radio may be found here. . .  You do NOT need a 
Photobucket membership to view/download the files.
http://s38.photobucket.com/user/ko6bb/library/Tube%20Type%20LF%20Regenerative%20Receiver 


-- 
73 From "The Beaconeer's Lair"
Phil, KO6BB
http://www.qsl.net/ko6bb/  (Web Page)

PRESENT RADIOS:
Grundigs: S-350 (~2006), G6 (2011) & S450DLX (2014).
HOMEBREW: 7 Tube+Rect 1v3 Regenerative RX for LF (built 2015)
Icom:    IC-735 Transceiver (~1990).
Icom:    R-75, Cascaded 250/125Hz CW-Filt, Panadapter. (~2005)
Icom:    IC-7200 Transceiver (~2015).
R-Shack: DX-380 digital portable (~1990).
SDR:     Softrock Ensemble II LF (built from a kit 2015).
Zenith:  Royal-7000 Transoceanic Portable (~1968).

ACCESSORIES: HOMEBREW LF-MF Pre-Amp, MFJ-993B HF Auto-Tuner.
              HOMEBREW 8 Hz Audio Filter.

ANTENNAS:  88 foot Long Ladder-line fed dipole, 35 feet AGL for MW/SW.
            Active Mini-Whip, 36 Feet AGL for LF/MW/SW.
            37 foot "Low Noise Vertical", 11 feet AGL for LF/MW/SW.

Merced, Central California, 37, 18, 37N   120, 30, 6W CM97rh