All Things Wireless & Letterpress

All Things Wireless & Letterpress

Sunday, February 22, 2015

A Homebrew MOPA for NRR

The above title translated: "A Home Made Master Oscillator Power Amplifier Transmitter for Novice Rig Round-up."

Novice Rig Round-up is the brainchild of  Bry, AF4K and WD4NKA (me).  Both of us are CW addicts, both of us love simple, close to the basics radios.  And such were the rigs of my Novice Days of 1977 - 1979. My Novice days were a two year romp in the world of Wireless Telegraphy of 1929 through 1960.  This was largely owing to:

A.  I was 22 years old, my profession, that of a Letterpress Printer and Offset Platemaker and Halftone Cameraman had bit the dust thanks to the Oil Embargo a couple years earlier.  I was apprenticing as an Optician, a change of careers, and my bankbook balance roughly equaled my shoe size.  I was perpetually strapped for cash.

B.  The Old Timers that I had for Elmers would give me their old QSTs from the 1930s to read.  My first radio magazines, until I became a charter subscriber to Ham Radio Horizons, were the red jacketed QSTs of 1938.

C.  It just so happened that our local Orlando Military Surplus depot, "Skycraft", then down on Orlando's  Gore Avenue, carried all the parts necessary to build the projects in those old QSTs!  In 1977, you could still get World War 2 surplus for pennies on the dollar.  807s, 6L6Gs, Transformers, Chokes, Isolantite Sockets, DCC wire, it was all there, for very little cost.  If Homebrewing was your thing, Skycraft was your place . . . as long as it was Valve/ Tube oriented, that is.

You have maybe seen the black and whites on my Wireless Page.  Those were all built back then from "Milsurp".  The only thing I did not show was my BC-453 "Q5er".  What a combo that was.  And it was this very rig that I thought I might reprise for Novice Rig Round-up.  What follows is a photo Show and Tell of piecing together my version of the Modified Pierce MOPA, which was featured in the ARRL Handbooks of 1954 and 55.

Starting off with three chassis which I obtained from eBay, I began by getting a feel for the components on the top deck.  These chassis are 5 x 7, not really all that much bigger than the QSL-Forty was.  Here is a shot of the transformer and 5Y3GT tube socket, ceramic.

The transformer is a drop-through chassis mount type that requires a large square hole to be cut.  Fortunately, I saved my good old Novice tools: the tin nibbler, and my Greenlee Socket Punches.  Some of you Orlando History folks might be interested to know that I purchased those punches from the old Harry P. Leu place just off Mills and Montana.

I start by drilling holes in the corners that I ruled out.  To the right is the pilot hole for the octal socket.

The nibbling commences.  the square nibbling head makes it pretty easy to stay aligned with the pencil mark.  After about ten minutes of cutting, I am done.  The cuts don't have to be perfect, because the transformer will cover the hole itself.  But you don't want it sloppy, either.  Just know that if you wander outside the lines a little, there's a lot of forgiveness.

The transformer drop and the octal socket hole are complete.  Not too awfully bad, if I may say so myself.  The Greenlees never punch a bad hole.  You align these punches by scoring or drawing lines that intersect at the center of  your desired punch area, then line up the four alignment marks on the punch itself.  This centers your punch exactly where you want it.  Then,  after drilling the pilot hole for the punches central nut to drop through, place the top part of the punch above deck, and thread the lower cutting portion of the punch below the surface.  Turning the nut pulls the cutter below deck through the metal, cutting as it goes, into the receiving "cup" at the top surface.  When screwed all the way through, the cutter punches a very clean hole through the metal.  You then unscrew the punch nut to take out the metal that was cut away.

Once the main openings are cut and punched, mounting screw holes must be located and drilled.  Above, I am inserting the transformer into the square hole I just cut, and observing where the mounting screws make contact with the metal surface.  I will then mark the place with a pencil.  I will also mark the valve socket holes for the 5Y3GT.

Here we have the completed punches and cuts.  All there is left to do is to drill out the front and rear apron holes for the power switch, the fuse holder and the power line and A.C. lines. 

Next comes the transmitter chassis itself.  Here, I do the same thing by marking where the valve sockets go, and in this photo, where the tuning capacitor will mount.  The capacitor mounting will not be simple because four holes must be marked with precision.  Now, many times the tightness and exactness can be somewhat mitigated by drilling holes wider than the screw and shaft, but you must also be careful not to drill so large that the screw heads pull through the apron itself.  So in this case, I chose to drill one "drill-bit size" larger than the 6-32 screws that mount the capacitor, and that of the shaft diameter.  

In the print shop I use transparencies to help register my  colours.  I simply took a transparency and cut a two inch by two inch square and used it to mark my holes.  I first punched a hole for the capacitor shaft, and slid it over the shaft.  Then, with the plastic squarely on the face of the capacitor itself, I took a Sharpie Marker and spotted the mounting screw holes.  This was then taped to the exact spot where I wanted to mount the capacitor, providing me with the drill marks needed.  I had to make sure that I had room below deck for the full rotation of the rotor plates below deck.

I use a hand held dremmel drill bit to make my initial marks on the chassis.  You can see one of these marks in the center of the shaft hole.  These are sufficiently deep enough to position the actual drill head that does the final drilling.

The capacitor is trial mounted: success.  Exactly as I wanted, and there is clearance for the rotor plates.  They do not make contact with the the metal at full open mesh.  This variable capacitor came from out of a Heathkit Apache, by the way.  It seems never to have actually been mounted.  I estimate the value to be approximately 250 pF.

Having drilled out the mounting holes for the variable tuning cap, I turn my attention to the valve and xtal sockets., all of which will be ceramic octals.  The coil will not be plug in, because I have a grand total of four to work with.  The power will be brought in from the power supply using Cinch Jones plugs, which I ordered on-line.  Here are the positioned sockets, for the 6AG7 oscillator, the 6V6G final, or alternately, the 6L6G, and the xtal socket which roughly positions between them, and the final tank coil with positions above the tuning capacitor.  A 6-32 screw and nut will be used to mount the coil form.  This was why I wanted to be sure I had a lot of overhead clearance for the rotor plates.

Hear is an overhead "above-deck" view of the mounted components and hardware.  On the side you can see the taper reamer that widens out the pilot holes just enough for the socket punch screws.  Once positioned and the intersect lines drawn, the pilot holes are drilled and reamed to size.

The pilot holes are marked, and the 6-32 screw and nut is already drilled and mounted.  The tuning capacitor is also already mounted.

And now, here we have it.  The transformer, the octal sockets, and the coil form is drilled, punched, and mounted.  All that is left is the SO-239 and Cinch Jones openings on the rear apron, the power switch on the front apron of the power supply, the key jacks, and the below-deck wiring.

The power switch is mounted, and those two screws you see next to it hold the bleeder resistor on the inside.  The transmitter itself (right) uses two banana jacks for my Russian Military Key.  Note the coil form is different.  I opted to use the black 35mm film cansister form rather than the white.  The screw holds the cap of the canister, and the coil form, inverted, snaps right into the cap and is held secure.  Since this is a link coupled antenna, the secondary of the tuning coil feeds the antenna directly via the SO239 jack on the rear apron.  There is a #46 pilot bulb in series with the antenna output.

Rear apron shot shows the SO239 jack, the male Cinch Jones jack, and the rear of the Power Supply, now sporting two mini-valve holes.  These were made to support regulator valves should they be necessary.  As it turned out, the transformer supplies much lower current than I thought, and regulation was not necessary.  A larger transformer will substitue this one soon.  As it is now, this MOPA does an honest five watts output, with 380vdc @ 20mA on the plate of the final.  Note also that the final is a 6L6GC.  This rig is capable of close to 40 watts with a decent transformer.  But as it is, it makes contacts on all five watts.

A below-deck view of the power supply.  That switch is a 20-amp heavy duty SPST switch.  The supply is lightly fused.  The filter choke is actually the primary of an audio plate to voice coil choke.  It does very nicely!  Four power leads coming from the Cinch Jones plugs permits separate B+ to the Oscillator and to the Final should it become necessary to regulate the oscillator.  It can also separately supply the screen if needs be.

And the below-deck view of the transmitter.  Not a whole lot there, isn't it?  This circuit uses a 100uH choke as the RFC to the 6AG7, which brings the oscillator output at about 5 mHz (heretofore known  as "Mc"!), sufficient to drive the 6L6GC, on both 80m and 40m.  Thus, no drive tuning is required.  This really keeps the parts count down.

And there she is!  All we need to do is plug in the antenna, key, tune her up, and fly.  She pairs well with the homebrew receiver, which is a Regenerodyne Regenerative Superhet.  On the top is a spotting oscillator, which helps the receiver tune the transmit frequency of the transmitter.

What self respecting Letterpress - in - a - Ham - Shack doesn't design it's own QSL cards?  I was hoping to have this ready by NRR, but it just didn't happen.  It is still slated to have a plate burned and I'll run it from the Chandler and Price there in the shop.  The silhouette of the key is actually taken from a silhougram of one of my own keys!
I would be remiss if I did not mention that I can do this same card for others!  If you, dear reader, might wish to have a quantity of these printed for your NRR station, let me know at!
That's it for today's installment.  More to come!
de wd4nka a r

Saturday, February 21, 2015

The Legacy of the "Simple" 6L6 Transmitter.

This installment will largely concern itself with what we now call the 6L6 power oscillator, or "PO".  These days, the 6L6 and it's descendants finds a ready audience with the electronic music crowd, but from time to time, the single valve (or, tube, but I prefer the more accurately label "valve") 6L6 or 6V6 transmitter turns up either in discussion, or as a project transmitter.

Very recently, we had an event on the air called "Novice Rig Round-up", in which about 300 hams trotted out their beginner rigs, whether stored in the cellar or purchased and restored for the event - and exposed the world to some very hard core nostalgia.  Tones and signal dynamics that I haven't heard in years.  And among the Home Brew offerings was the single valve 6L6'er, and I believe more than one.  All of these were, of course, crystal controlled, as per the Novice protocol.

It is remarkable that a valve on the brink of turning eighty years old is still stirring up the Ether!  Shades of Grammar and Goodman!

The 6L6 is as ubiquitous as the day is long and icicles form on a frosty winter's day.  As the photo above hints, a lot of this use was portable, as per the Depression Era concept of "portable".  The valve began life as an intended Audio Beam pentode, but from the very first announcement of it's introduction in May of 1936, this tube found itself the frequent visitor of the Ham Shack Work Bench.  I thought it might be fun to catch a few early episodes of the Saga of the 6L6 Power Oscillator, beginning with the announcement itself, as it appeared on the pages of QST, fetching the attention of the 10,000 hams in North America:

The original 6L6 was, of course, a metal valve.  The idea was shielding, and in truth, it was an effective shield.  Eventually the concept of the metal valve went by the wayside simply because just as effective shielding can be obtained using metal slip cans, which became the norm with the advent of the 9 and 7 pin "mini-tubes".  The 6L6 gave a respectable amount of output at Class A, to the tune of about 8 watts, realistically.  But as we shall see, a lot more was realised as a Class C cw transmitter.

The very next month's issue of QST saw the first mention of the 6L6 as a transmit tube.  We see the first, and probably the most enduring morph of the 6L6 in cw service in Frank Edmonds' June '36 article "The 6L6 Beam Power Tube as a High-Output Crystal Oscillator":

And here was shown the most enduring schematic.  We will find this oscillator circuit very nearly duplicated in a later and more famous article from February 1938.  The opening line of this article pretty well sums up what the legions of Ham Radio Homebrewers held as principle for the past Century: "The advent of a new tube always kindles the fire of conjecture as to it's adaptability to transmitter oscillator design, even though it may have been intended for other uses." (italics mine)

And adapt we did.  Now figure, this tube sold for about two or three 1936 dollars.  Depression dollars!  This tube could have purchased dinner, a cigar, and an overnight stay at a not-to-disrespectful boarding house.  It wasn't cheap . . . about 40 dollars today.  Thus, the 6L6 was not an economy move.  It was, rather, an efficiency move.  You got more bang for the E x I buck, input to output!  The concentrated electron beam flow was a thing of genius!

Check out that Crystal Oscillator Performance Data Table.  This was the article that most inspired the later creator of the famous "QSL-40", Fred Sutter.  Check out that last line: 425vdc plate, 285vdc screen, 70.1 watts input!!  Yikes!!  At 51.5% efficiency at Class C, it realised 36.2 watts output.  Oh, but hold on to your electrodes, this puppy could be driven higher than that!  

What about the xtal current?  Maybe I should have shown a close-up of the xtal, but it is the size of a tumbler combination pad-lock!  Bliley made them, we call them "Door Knob" crystals today.  They handled enough current to light a #47 pilot lamp.  No kidding!  There were a few for sale at the recent Orlando Hamcation, I should have purchased one just for display purposes.  They really are a marvel to behold, the ancient Bliley.

Later, in November 1936, there appeared what to this author was the real impetus that put the 6L6 on the radar for the more commonly resource endowed ham of the middle Depression Era.  Byron Goodman published an article in that month's QST called "A Simple Two-Band 6L6 Tri-Tet Transmitter".  The TTT.  Ok, so that monicker never caught on, but boy, the portable 6L6 PO transmitter sure did!

"In these days of inexpensive crystals and tubes, there is really no reason for the beginner or amateur of limited means to deprive himself of the advantages of a crystal controlled transmitter."  Thank you, Mr. Goodman!  This article, in my opinion, is what put the 6L6 in the Shack.  Here it is, in all it's Masonite and Plywood Glory.  And yes, fans, she is still built very much like this today.  Only, uh, with a little less grid current.  Oh, but Byron was not one to hyper-charge this valve, much more sensible voltages are utilised.  While he holds out that this rig can be powered up to 400v, he speaks largely about using the 250v "power pack".

I, personally, am a big fan of Masonite, if it is double tempered.  This would be the type used for clip-boards.  Masonite has an oil impregnation that permits longevity that is quite impressive.  As a Letterpress Printer that deals with very old wooden type cases, I find my Masonite cases have outlived their wooden counterparts in un-air conditioned Florida climes 2 and 3 to 1.  Wood rot sets in down here fairly quickly, Masonite is very nearly immune to this.  Polished, it is a dead ringer for Bakelite.  In the 1930s, often they were backed with aluminium below deck, for shielding and to create a ground planing effect.  It was also easier to ground, but cheaper than an all metal chassis.  This construction mode lasted all the way up into the 1960s.  Not a bad run, boys!

The back-page finish of the Goodman article.  A foundational piece of literature from a radio pioneer in his own right, Goodman published great construction articles, along with the likes of Millen, Lamb, and Grammar.  And a lot of their work is informational to this day.  Why?  Because the Laws of Physics have not yet changed, ya young Squirt!

The last entry closes out the 1930s to a large degree, and is probably the most famous of the 6L6 articles.  Entitled "The QSL-Forty", by Fred Sutter, this article uses the second morph of the 6L6, the glass "S" shaped envelope 6L6G.

The QSL-Forty was so named because the chassis dimensions were QSL-card sized, 3-1/2 by 5-1/2 inches, and stood 3 inches tall.  Oh, and beware that schematic.  It was corrected the next month: the cathode needs to have a key inserted between it and ground, and not shunted across cathode and ground.  Yeah, guess what?  As a Novice I built this rig right out of this magazine, and I didn't catch that.  So I can hardly blame the QST layout guys!  

Fred Sutter cites as his source the Edmonds article, but I think he also read Goodman's article, too.  Just a hunch.  To me, of equal value in this article is Sutter's description of how he made that tank/antenna coil!  I followed his instructions, and you need to know, this was a very nice coil !

This is my own version of the QSL-Forty, using what I call a "Modified Slat Board" chassis, consisting of angle aluminium with hardwood end-blocks.  Very, very solid chassis construction.  The transmitter was a bucking bronco, though.  To keep a good note, you cannot simply tune to a dip in current like a MOPA.  I found I had to detune slightly to the high capacity side, or "Hi-C" of rez tuning.  This brings the efficiency of the PO down considerably, but it keeps the FCC off your back.  This rig here could do 20 watts, but if I wanted to keep the "youp" down to a dull roar, it was detuned to about 5 watts output.  And it had to be touched up constantly, regardless of the xtal.  But it made contacts, and was very much a 1930s romp on 40 meters, no doubt!  But, back to the Sutter article. . . 

Feel like making one?  Here's the chassis fabrication and punch-out plans.  Check out the space you have to work in, with that huge Triplett meter!  Yow!! 

The power supply, with the tapped bleeder to supply the screen.  I might imagine these days you'd want to stack a couple OD3s or V150s for some regulated 300vdc on that screen.  Notes on cw signals in 1936 were more tightly controlled than they were "Pre Cairo", in '27, but a far cry from what is required today.  This was why these ops simply tuned to 'dip', and let 'er rip!

Heh, I have to point out that pilot lamp in the grid circuit.  It served both as regulator and as a warning  indicator to unkey!  Ahh, the Thirties . . . 

And this about wraps things up.  The final page and finishing comments by Fred Sutter.  He was a very candid writer, and some of the comments are as interesting as the tone they take!  But then, all these old radio mags have an aire of authenticity, a very "face to face" nature to them.  Very plain language was used: the publisher understood that the average Ham was the average "Joe", educated maybe through 6th grade, many completed high school, some collegians in the mix.  Mostly tradesmen. Very easy to understand, and the humour is priceless.
I hope you liked our little fifty-cent tour of the Legacy of the 6L6 PO.  Perhaps you might like to give one a try for NRR?  Just remember: Plate and screen voltages bite, use common sense, and don't tune for power, tune for a good note.  Keep yer dits and dahs short, and yer QSOs long.  And above all, have fun!
de wd4nka
Paper Wren Press.

Late Edit!  One of my favourite radio persons, and possibly the only other person on the planet who even knows who the Dixie Squinch Owl is, is not only a commercial broadcast engineer, but Bobbi is also an authority on the subject of 1930s Radio in general.  Here is a link to her blog, and the page that contains her own construction of the QSL-Forty.