All Things Wireless & Letterpress

All Things Wireless & Letterpress

Wednesday, December 31, 2014

Revisiting the Regenerodyne.

With the Novice Rig Round-up coming up this February, I thought I might review a circuit I developed in the late 1980s and early 1990s.  What became known as the Regenerodyne began as an attempt on my part to come up with a small travel receiver to take with me on the road.  I traveled for a living, and lugging the Drake or the Q5'er was just not practical.  I needed something I could toss on a night stand at a hotel, clip the antenna to a bed rail or window screen, and yield enough sensitivity and selectivity to be worth using.  It needed to be stable.

I originally started building a Direct Conversion receiver for 80m, and two converters.  The audio amp was about 100dB in the gain department, and the plan was to use a chain of 88mH toroids with 1uF caps shunted and bridged around these surplus toroids.  I built these before, they were ok.  I really hated the massive audio gain necessary for these rigs, and I hated the lack of selectivity, the native widebanded nature these receivers embody.  All these deficits can be made up for, but in the end what we have is a not-so-simple "simple receiver".  And I still had what I considered an annoying high-gain amp.

But I did like the idea of converting high frequencies down to a lower frequency, where greater gain and selectivity is natural.  Radio circuits simply behave better at lower frequencies, the lower the better.  It's an old concept going back to the dawn of the Superheterodyne receiver.

Having already built a regen for 80m cw - which was built like a VFO - I simply took my 3n211 / 2n2222a xtal downconverters and coupled them to the grid of the regen.  

Wow.  I had 40 and 20 meters, tuning and behaving like 80 - 160 meters.  That means better stability, selectivity, and overall gain than ever I had using a regen as a TRF on those bands.  Since the output of those mixers were untuned, I could tune pretty much the entirety of 40 and 20 meters with equal performance from end to end.  The selectivity and gain of the regenerative tuner was sufficient to select the intermediate frequency from the mixer output.  This made for a very simple IF transformer.

By 1990, I had substituted the solid state modular converters with a standard cathode injected tube-type mixer incorporating a 6J7 pentode and a 6C5 triode xtal oscillator.  The regenerative detector itself was made to tune 3.0 down to 2.5 mc.  This enabled me to use a 10mc xtal to produce a tuneable range of 7.0 to 7.5mc.  The converter and tuner was built on the same chassis as an integrated unit.  

This had several advantages, not the least of which was instant band changing. 10.0, 13.0 and 17.0 xtals gave me 40, 30 and 20m respectively, and the dial never had to be recalibrated for each band.  I had a converter standing between the regen detector and the antenna, so virtually NO RF leaked, and images were not even a hint of an issue.  From the front panel, this receiver looked like any other regen.  But in operation, these higher HF bands tuned like 140 meters, which meant lots of selectivity, lots of gain, and the regen control barely had to be touched every 100kc or so, if operated at critical threshold.

It became the "tamed regenerative receiver", although technically it's a superhet.

So, now you have the basics of the Regenerodyne, and a history of it's design concept.  What surprised me was that when I did research to see what other folks did in the past along these lines, I discovered that in over 50 years of published articles, I could not find anyone doing the exact same thing.  One letter to an editor described using an outboard converter for 10 meters, feeding a regen at 160 meters in 1938.  That was about it.  It seemed that nothing was written about building both the detector and xtal converters on one chassis!  Probably because by 1938, the virtues of a fixed tuned detector enabled it to enjoy the new technologies, such as single signal filtering.  Anything regenerative became a technological cul - de - sac.

Frank Jones developed what he called the "Super Gainer", which was a variably tuned converter feeding a fixed tuned regenerative detector at 100kc.  Between 1936 and up to WW2, there were several variants of this set-up.  I needed a name to avoid confusing my circuit with the Super Gainer, because my circuit was the reverse of the 'Gainer.  Mine was a fixed tuned converter feeding a variably tuned regenerative detector.  I came up with the name Regenerodyne.  

I shared the Regenerodyne with Dave Ingram at the Atlanta Hamfest, where he encouraged me to send an article to CQ magazine.  While that never happened, I did open up a web page in the early 1990s featuring this circuit.  

I met Arnaldo Corro of Radio Habana's "DX International" on the Glowbugs Group reflector about this time, and we compared notes.  Turns out, he was working on the very same idea, and the most amazing thing was the name he chose for this lash-up:  Regenerodino!!

I built three prototypes of the Regenerodyne, the first is shown in the opening photo.  The last one I built is a miniature version intended for dedicated use on 20m.  I still have that one, the others were sold over time.
Let's take a look at the Regenerodyne circuit itself:

You should be able to click on the schematic to enlarge it.  This is the Classic Regenerodyne.  The only variants I ever made were a double tuned front end and maybe an audio power amp.  But this is the heart and soul of the "R-dyne".

From the antenna, the desired signal is tuned and fed directly to the grid of the 6J7 pentode, while the 6C5 xtal oscillator injects the cathode.  The output of the mixer is untuned, therefore four signals are present at L3: the oscillator frequency, the input frequency from the antenna, the difference frequency, and the sum frequency.

The regenerative detector grid coil L4, which is also the tuner, has an extremely high native Q, which is more than selective enough to pick out the desired frequency, which is, of course, the difference frequency.  Not only does it pick out the difference, but it maintains this difference as the detector tunes.  As such, the IF, as regards the Regenerodyne, is not one fixed frequency, but rather a "window", or continuum of frequencies.  I call this the "IF Window".  In the circuit under discussion, that window is about 300 - 400 kc wide.

The coil form that supports L3, L4, and the feedback coil L5 is a pill vial or a plastic 35mm film canister.  You can pretty much use any form you want.  The RFCs are primaries from old audio plate to voice coil transformers.  Note the capacitor across the audio output choke: it provides enough filtering, in combination with the chokes value . . . whatever that is . . .  to knock out the sonics, or high audio frequencies.

Notice that all the parts sort of date to the late 1930s?  I purposed to use antique radio parts as much as I could.  Mission accomplished.  I also wanted not to depend on manufactured IF transformers.  In the Regenerodyne, you wind that transformer. In fact, the Regenerodyne compares to a standard 1-v-1 regen receiver in overall parts count, circuit complexity, and building cost.

Operating the R-dyne is fun.  Now, personally, I like the idea of regenerative control.  I feel I have more control over the received signal than with any other receiver.  At 2.5 mc, the tuning rate almost matches a far more complex superhet, and while you do need to touch up the regen injection as you tune, at these lower frequencies, the amount of adjustment is far less than operating a regen on 40 or 80 meters.

Here is a sample of a Regenerodyne in action.  Notice that I am still using a window frame and wire clip for an antenna.

The next video is the same Regenerodyne in play, a few weeks later.  The band was, unfortunately, pretty dead.  But it was the only time I had available to set up to shoot.  The audio on each of these videos was fed into an outboard audio amp so the cheap microphone on the camera could pick it up.  All in all, it wasn't a bad couple videos.

Now, lets have a little photo "Show and Tell".  My first Regenerodyne was built in a large BUD cabinet, using National vernier hardware, the famous HRO dial and PW gearing, which I think is a must for any receiver if you can get your hands on one.  Whenever I find them, I buy them if I can.  If you are into building receivers, it's an investment in the enjoyment of your receiver, believe me!  The little "Eaves Dropper" shown in the video uses the National Velvet from an SW-3, and is built into a National SRR cabinet.

This is a photo of the first Regenerodyne, removed from the cabinet.  Yeah, black wrinkle finish.  Is there any other, really?

Top view showing the PW gearbox and the 75pF tuning cap and calibration cap.  To the left is the 6J7, and behind it the 6C5 LO.  Top center is the 6SN7.  Next is the audio choke.

From the rear apron, you find the power connects and the xtal that determines the band.  I did not wire a band switch although I could easily have.  I had no problems just plugging in the xtal I needed as I needed it.  This particular radio covered 40, 30 and 20m.  The IF window was 3.0 to 2.5kc.
 Below deck shows the twin ganged input tuner.  This radio had a double tuned RF input to the mixer, and both coils are wound on the same form with no attempt to shield one from the other.  Worked out pretty nicely, really.  It was sort of a tuneable Butterworth filter.  The coil center right is the detector coil.  The Mixer coil is wound directly over the detector tuning coil, matching the turns count.  I have about 1/16" of insulation between the two coils, which took the form of a thick, foamy double sided tape.  I think it was called "Tiger Tape".  A few layers of electricians tape or even masking tape will do.
 Here is the rig in operation in the field.  Note the power supply is seperate, using two back to back filament transformers to develop the 100vdc B+ operating current.  The stand-by switch is one the power supply.  Headsets plugged in the back, on the rear apron.

This is my current Regenerodyne, the only one I kept.  As mentioned earlier, it is nick-named "Eaves Dropper" because of it's diminutive size.  It is a dedicated 20m receiver, covering 14.000 - 14.250.  When I take this guy to the Hamfests and operate it on my swap table, I always draw a crowd with it.  The OTers all think its a regular regenerative TRF receiver, and when I tell them it's a 20m receiver, they all want to hear it. When they handle the tuning and the regeneration control, they are amazed. I tell them it's a superhet in disguise, they look at the three tubes and next to nothing parts count, the hand wound coils, and they always walk away amazed.  I have turned down a lot of offers for this little rx, and yes, it's still not for sale.  It's the last of my prototypes.

Here is the view through the flip top lid.  Lower left is the 6SN7 detector/audio stage, above it is the detector coil.  Above it is the band tuning coil, and to it's right, the 6J7 pentode mixer, and center right is the metal 6C5 LO.  In the upper right corner you can see the xtal.  Audio choke is on the lower right.

Another view of the upper deck, and the 35pF tuning cap.  There is no calibration cap on this rig, I just wound the detector coil just right . . . luck.  See the masking tape over the detector coil?  That's yer IF transformer, gang!  The detector coil is wound like any other tickler coil arrangement for any other typical "Novice" regen receiver, but then a couple layers of tape is put over that coil.  Then the mixer coil, untuned, is wound directly over it, matching turns count and direction.  I borrowed this technique from the 1929 Bearcat Model 3B.  This actually helps the coupling, because the relative impedance is matched.  More or less.  Nothing is precision here.

Below deck.  Here, I did use 2.5mH RFCs instead of audio choke primaries, largely owing to space allowance.  I can't really tell the difference.  You can see the band tuning/ peaking cap.  Regeneration is controlled by a 50k Cardwell milsurp pot.  Boy, just looking at it, there's almost nothing there.  Yeah, it's a superhet.  Would you believe it just looking at it??

No "Show & Tell" with the Regenerodyne is ever complete without showing Ken Lotts's version. I love his construction techniques!  He sent me these shortly before he parted with it.  Yeah, it sold.  Ken used quality parts and did a quality job.

Just a few more shots before I close this segment down.  I wonder who has this receiver now? Also, I wonder if the person who bought my original regenerodyne, the one shown at the opening of this installment still has it?  And if they sold it, does the new owner know he or she has the very first Regenerodyne by that name?  Hmm . . . . 

Well that's about it for tonight.  Hope it was a fun, maybe educational little excursion.  Try building a Regenerodyne!  Bang for the buck, they are hard to beat, and are a genuine communications device!

For those wishing to view the original article and photos, click here, or copy and paste:

-gary // wd4nka

ADDENDUM  (31 Mar 2015):

At the end of the second week of February, Novice Rig Round-Up came and went.  It was an overwhelming success with lots of participation.  I thought I might share the rig I used, largely because the Regenerodyne played a major role at WD4NKA.

The Regenerodyne featured in the above videos was the primary receiver used at my station for NRR.  I padded the front end tuner for 40m, and replaced the 13mc xtal in the LO with a 10mc xtal.  The xtal was about 8kc off, which made my tuning lag by that frequency, but it was easy to compensate for on the dial.  (Apparently my 13mc xtal was more like 12.920 mc.)

Now, a couple interesting exponents developed as I paired the R-dyne with the transmitter I would use with it, a two valve MOPA consisting of a 6AG7 xtal oscillator, and a 6V6G final, running at around 10w. input, to a 33 foot vertical mounted on my roof, using a four-wire ground system.  One exponent was side-tone.  For a while, my Hammarlund HQ-170 provided that, but I had to tune to each frequency.  I discovered that after a while, I got used to hearing the receiver "block" during transmit, and in that blocked state, I could hear a bit of ripple from the power supply.  Not enough to actually modulate the cw carrier - it sends a pure DC note.  But when the transmitter is parked right next to a regenerative detector, you can hear the slightest bit of hum.  Sometimes more than slight, but in my case, it was just enough to use as a side tone.  So I stopped using the '170 for that purpose.

Another exponent was the use of my transmit xtal to spot my transmit frequency.  Obviously my regen detector was completely blocked for several kHz.  So, I built a small 2n2222a xtal oscillator into a small mini-box, and set it by the receiver. That worked great!  But I noticed something that I will try to describe:

I could control the amount of carrier from that little oscillator box by it's proximity from the receiver.  The closer the stronger, &c.  At some point, the carrier "injection" was about equal to the receiver's own autodyne regenerative feedback.  Tuning the feedback below threshold, I found that I could plainly hear signals being demodulated just as easily as the receiver in it's own regenerative detection mode.

What I had was, in effect, a direct conversion detection action going on.  But more than that!  Further adjustment of the feedback increased the Q beyond what it would have been without the oscillator on and the detector depending on it's own autodyne detection action.  The feedback was actually below just normal threshold, the most sensitive point of any regen detector.  But since an external carrier was producing the audio product, cw signals could still be heard, and quite loudly.  But then, I would tune the detector to make sure I was tuned to the center of the "injected carrier".  I noticed that certain signals audible within that carrier width would attenuate or accentuate depending on how close I tuned to either "fringe edge" of the carrier.

What I was observing was something approaching single signal action.  I would call it "reduced sideband filtering", not complete suppression as you would find in a regular single signal filtered receiver.  But there was a noticeable reduction of signal on the other "side" of the injected signal.

So . . .

My Regenerodyne operated in two modes for NRR.  Regen mode, and enhanced Direct Conversion mode (EDC, to coin an acronym.).  One nice thing about EDC, was that like a regular DC receiver, there are no antenna affects.  Now, the Regenerodyne does not suffer from antenna affects, but a standard twinplex does.  EDC might be an idea to stabilise a TRF regen detector, although the most practical way to inject an external signal to the regen detector is by use of a xtal oscillator, which would require the receiver to be retuned to the carrier of each channel. "Channelized operation".  But then, if you operate with a xtal controlled transmitter, it could be practical.  A regenerative Direct Conversion (RDC - there we go again!) would be more selective than it's standard solid state counterpart, AND require much, much less audio gain!

Food for thought.

Below is the home-brew station used for NRR:

The transmitter is built into the cabinet of a defunct VTVM.  Keys used was the J36 (shown) and an 8-amp straight key from the Royal Air Force.  Not shown is the power supply . . .you can just see the front of it on the extreme right.

I also used my Gonset GSB-100 and HQ-170 combo during NRR, too.  I'd say about half the contacts were made via the rig shown above. The Gonset runs about sixty watts output.  The MOPA runs right around six watts.  My aggregate total between my MOPA at 5-6 watts, and the Gonset at 60 watts was 350 points.  Not a bad showing for a three-valve regen based lash-up!

This is a shot of the actual operating position during NRR.  Left to right: Power Supply, ACR (switch, actually), Spot Oscillator minibox on top, MOPA, and Regenerodyne.  There's my 8-amp RAF key with all my xtals mounted on the same board as the key.  Head-sets are low-Z military from around 1952.

This is the log I kept for NRR.  I am seriously considering designing and printing "NRR Log Books" specifically designed to be used for next year's NRR.  Catagories include power used, type of rig contacted, whether it be VFO or Xtal controlled, QRP or QRO.  NRR QSLs may also be offered.  Letterpress printed, of course!

And look what came in the mail on Monday!  My NRR Certificate!  There ya go!  So, never let it be said that a Regenerative Receiver is unsuited for serious communications!  Nothing could be further from the truth.  It's just . . . not a "plug-and-play" affair. 

So, that's it for now.  Stay tuned for further installments, where in one way or another, Wireless meets Letterpress (in this case, the Certificate!)

-gary // wd4nka.

No comments:

Post a Comment