The AM Forum
March 04, 2024, 09:30:44 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: The Hail Mary Transmitter  (Read 1000 times)
0 Members and 1 Guest are viewing this topic.
KD1SH
Member

Online Online

Posts: 726



« on: February 09, 2024, 02:22:57 PM »

     Two weeks before the AM Rally, after yet another year of procrastinating and unfinished projects, I remembered that I really wanted to have a home-brew rig up and running for the Rally. I've got lots of stuff in various stages of completeness, but nothing close enough for any possibility of completion in time for the Rally. What to do?
     The clock is running; time for a Hail Mary pass! This could be fun, or at the very least, thereís the irony of a guy who knows absolutely nothing about football making football analogies (that is a football analogy, right?)
     Iíve got this little home-brew 75 and 40 meter CW transmitter that I picked up at a hamfest a while back; maybe I can transform this into a workable AM rig in time for the Rally. Iíve already got an external modulator that I built a while back, so part of the project is done. http://amfone.net/Amforum/index.php?topic=47922.0
     This little CW rig was built from plans in the Dec. 1958 issue of QST, according to the information thoughtfully marked on the inside of the chassis bottom plate by its original builder. Itís got two tubes, a 6146 as a keyed crystal oscillator, and a 5U4 rectifier.

      Drilling and blasting went as follows:

   I solid-stated the power supply, and used the 5U4ís octal socket for a 6AG7 driver, the tuned output circuit of which is visible on the top of the chassis.

   I added a second octal socket, for another 6AG7, as a buffer for the external VFO. The output of this buffer is resonated for 40m via a slug-tuned inductor, typical of a DX40, DX60, Eico 720, and a host of others, but where those guys left the buffer un-tuned on 75m (on 75m the inductor appears as an RF choke in those rigs), I installed a disc capacitor switched in parallel with the inductor for 75m. I got a stronger, cleaner output from the buffer that way. I had to install a clumsy toggle switch on the back of the chassis for that, but since I was starting from an existing build, I had little freedom with the layout.

   I changed the B+ supply from a capacitor input to choke input, to get better regulation under load and to lower the B+ to 350v, since all I wanted was around 20-25 watts out, to drive my amp.

   I built a mosfet regulator to provide 300v for the 6AG7ís (the perf-board is visible in the picture), added a little 120-150vac transformer, on top of the chassis, and built another perf-board with a zener regulator to provide negative bias for the 6146, as well as -180v for grid-block keying to stifle everything when not transmitting.

   Iíve got a couple of DX60ís and their HG-10 VFOís, so I added an octal socket at the rear of the chassis, to supply all the needed voltages: 6.3vac for the heaters, 300v for the HV, and negative 180v to grid-block it. Unfortunately, that necessitated moving the RF output connector, leaving the SO-239 to make a rather undignified exit through the top of the chassis, but hey, this was never intended to be pretty.

   I installed a barrier between the final PA section and the buffer-driver section, made from double-sided copper-clad, with acrylic plastic insulated feed-throughs (courtesy of my lathe), with bypass caps.

   I decided to provide grid bias to the 6146 with a combination of grid-leak and fixed bias, at a roughly 50/50 ratio; that way I donít have to mess with clamper tubes or switch the B+. The only downside to that, I found, was that with only the -51v output of my bias regulator, at key-up, the idle current of the 6146 was higher than Iíd have thought, like 60ma, so I added some connections to my PTT relay to apply additional grid-block voltage to the 6146 when not transmitting, which settles the tube down to 30ma at idle.

   All of this is pretty straightforward stuff, but I got a bit more experimental when it came time for applying screen voltage to the 6146, referring to this excellent dissertation by Dean, WA1KNX: http://www.amfone.net/AMPX/71.htm. Following his example (see his link to the schematics and formulas) I used his recommended 60/40 ratio of un-modulated to modulated screen voltages, through the appropriate resistances arrived at through his calculations. It worked out quite well; in my next post Iíll include some pictures of output waveforms under modulation.

   The story, though, didnít have quite the happy ending that Iíd planned. The rig is now working very well, and itís as mechanically ugly as Iíd hoped, but it never got onto the AM Rally. It had some last-minute glitches that drove me crazy, and I put it aside at the start of the Rally, and ran with my old reliable DX60. While it would transmit a fine carrier, it was very unstable, grid-drive jumping around spastically, both driver and final tuned circuits showing multiple nulls and peaks, and some very weird looking stuff on the spectrum analyzer. Come to find out that, in my haste, Iíd entirely forgotten to bypass my heater connections, or ground one side of the heaters. Iíd put bypass caps to ground where the heater voltage entered the final PA compartment, at the feed-through insulators, but nowhere else. Once I grounded one side of the heater terminals at the sockets, and bypassed the other side, it worked perfectly.

   I still want to install a metal barrier, on top, to shield the driver tuning section from the 6146, and also some sort of housing to protect inattentive fingers from the 300v on the driver tuning air-variable. The driverís tank circuit is in series with the plate voltage supply, and isolated from the chassis with a nylon block and fiberglass shaft coupler, so itís hot, as my left pinky-finger will readily attest.

   Check out the cute little horizontal grid current meter; found that in one of my mystery boxes in my shop.

   Some pictures of the butt-ugly but functional result:


* Hail Mary Front.JPG (90.06 KB, 1024x768 - viewed 86 times.)

* Hail Mary Rear.JPG (120.07 KB, 1024x768 - viewed 92 times.)

* Hail Mary Top.JPG (112.48 KB, 1024x768 - viewed 97 times.)
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
KD1SH
Member

Online Online

Posts: 726



« Reply #1 on: February 09, 2024, 02:26:43 PM »

Some more pictures:


* Hail Mary Bottom.JPG (130.7 KB, 1024x768 - viewed 81 times.)

* Insulators Installed.JPG (73.76 KB, 1024x768 - viewed 52 times.)
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
KD1SH
Member

Online Online

Posts: 726



« Reply #2 on: February 09, 2024, 02:31:43 PM »

And some scope-shots. Frequency response is amazingly flat from 100hz all the way to 15khz, with a slight bit of roll-off over 12khz. I didn't include a scope-shot at 100hz because I wanted to use an old analog scope, to avoid the ugly digital "jaggies," but with the sweep so low it was hard to get a picture of it. It even did a respectable triangle at 5khz.


* At 5khz.JPG (54.14 KB, 1024x768 - viewed 32 times.)

* At 10khz.JPG (62.58 KB, 1024x768 - viewed 30 times.)

* 5khz Triangle.JPG (47.76 KB, 1024x768 - viewed 35 times.)
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
KD1SH
Member

Online Online

Posts: 726



« Reply #3 on: February 09, 2024, 06:35:59 PM »

  Another picture, this one from my Steve-O-Meter, showing excellent symmetry: I can hit 100% positive while only hitting approximately 97% negative. I do find, though, that the symmetry becomes slightly negative after about 12khz. I keep my Inovonics 223 set for 10khz bandwidth, like a brick-wall, anyway.


* On REA Mod Monitor.JPG (98.69 KB, 1024x768 - viewed 60 times.)
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
KD1SH
Member

Online Online

Posts: 726



« Reply #4 on: February 12, 2024, 03:26:54 PM »

  A couple of spectrum analyzer shots. Second harmonic down 35 dB on 75 meters, and 31 dB on 40 meters. Shy of the official requirement, but it's not exactly slobbering all over itself, either, and not terribly shabby for a pi, rather than a pi-el.
  Now, to rework the tank into a pi-el will entail more drilling and blasting, and I'm not sure, really if there's room in there. We'll see.
  I run a tuner on both bands, and resonant coax-fed inverted V's, so I doubt much of that 2nd harmonic will get radiated, anyway. Everything else looks clean.


* SA on 75.JPG (117.1 KB, 1024x768 - viewed 45 times.)

* SA on 40.JPG (109.09 KB, 1024x768 - viewed 41 times.)
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
KD1SH
Member

Online Online

Posts: 726



« Reply #5 on: February 13, 2024, 01:31:17 PM »

   Okay, so the 2nd harmonic thing got me thinking: what sort of harmonic attenuation could I expect from a tuner?
   So here are a couple of analyzer shots, one without a tuner, and one with an old (but very nicely built) Yaesu FC-301 tuner.
   A roughly 10dB improvement in harmonic suppression using the tuner.
(I remembered this time to use the "save" function on my analyzer, rather than inflict another crappy picture, with my reflection in it, on everyone)


* 75 meters no tuner.png (62.77 KB, 1236x653 - viewed 32 times.)

* 75 meters with tuner.png (64.81 KB, 1236x653 - viewed 33 times.)
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
W1DAN
Contributing
Member
*
Offline Offline

Posts: 899



« Reply #6 on: February 14, 2024, 01:54:19 PM »

Excellent work!

Dan
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2641


Making AM GREAT Again!


« Reply #7 on: February 14, 2024, 02:14:24 PM »

Harmonic rejection in the tuna will depend on what it is trying to match.  It will change as the impedance ratio does.

--Shane
WP2ASS / ex KD6VXI
Logged
KD1SH
Member

Online Online

Posts: 726



« Reply #8 on: February 14, 2024, 02:24:08 PM »

  Good pointóI will dig out some non-reactive resistors and kludge up some differing loads for testing.

Harmonic rejection in the tuna will depend on what it is trying to match.  It will change as the impedance ratio does.

--Shane
WP2ASS / ex KD6VXI
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
KD1SH
Member

Online Online

Posts: 726



« Reply #9 on: February 14, 2024, 02:29:39 PM »

  Thanks, Dan. It's about as handsome as the south end of a northbound cow, not very professional, and very much a cobbled together contraption, but it was fun and educational. In other words, it's what AM is all about!

Excellent work!

Dan
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
óRobin, in the 1960's Batman TV series.
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.04 seconds with 19 queries.