To see if I'm on the air right now click here

Wednesday, July 22, 2020

A 1929 Transmitter on 14MHz

This year the Bruce Kelley Memorial 1929 CW QSO Party will include 20 meters. I decided to see how my Push-Pull Colpitts does on 14MHz. It has a fairly good note and is very usable in the BK.

I wanted to stay with a high capacity tank circuit. This minimizes RF current through the tubes and should improve tone and performance. In addition, I wanted a number of turns in the tank coil so that I would have reasonable transformer coupling between the tank coil and the antenna link. I ended up with two 4 turn, 3" long coils wound on a 1.25" form. This 1.25" form resulted in more tank coil turns on 20 compared to the 2.5" form used for the tank coils on 40 and 80 and the antenna link coils. Tank capacitance is about 100 pF on 14 MHz. My Colpitts now runs at about 20% efficiency giving 2-3 watts output on 20 meters.


My 14 MHz CW Signal as recorded on July 22, 2020 using the Northern Utah WebSDR:



For comparison listen to this same transmitter on 80 meters:




Sunday, May 17, 2020

What Do I Sound Like?

Have you ever gotten a 597 signal report after putting a lot of time and effort into building that new rig copied from an old (or really old) issue of QST? What is the other station hearing? More than likely they aren't familiar with the definition of a T7 signal (which to many of us sounds fine). Maybe it just sounds different or maybe it has chirp or maybe it sounds like a buzz saw the morning after a big party.

The first thing I do is check my signal in the shack receiver, but lots of times I'm overloading my receiver and I sound like that buzz saw. Is this really what others are hearing? Maybe being so close makes a minor problem sound worse than it is.

The next step is to have a friend a few miles away listen to you and, possibly, hold a phone up to the speaker so that you can hear your self. This can be real helpful but lots of times I need a more complete listen.

As a final check I recommend taking advantage of one of the many web based radios available all over the world. Link into one of these, turn on your own computer speaker, set the frequency/mode you wish to test and then transmit. Given needed propagation you can listen to yourself and the waterfall display helps you see what your signal is actually doing.

Friday, May 1, 2020

Push-Pull Colpitts Tips

(As questions/answers/tips concerning my Push-Pull Colpitts come up I'll post my thoughts here.)

Where's the key? Many early transmitter construction articles left the keying method up to the builder. Everyone had their favorite. In this case I used cathode keying by placing the key between the filament center tap and ground. Lacking a filament transformer with a center tap I could have simulated the center tap by using two 40-75 ohm resistors. Keep in mind that cathode keying results in some amount of voltage on the ungrounded key terminal. See my blog entry at  https://w0vlz.blogspot.com/2018/04/cathode-keying-safety.html  for how to deal with this. Also, be sure to provide a separate path for B- to ground. Otherwise you will be doing B- keying and have the entire power supply voltage across your key terminals.

Why are your coils different from what the article says? Studying the original QST article I decided that the 80 mtr coils were just too tightly wound for my comfort. The author allowed only 1/6" between each turn. Shorts were bound to happen. I added 1/2" to each tank coil, spacing the tank coil supports 3" apart rather than 2 1/2" apart. I had enough spare capacity in my tank capacitor to compensate for the decreased inductance. For the same reason I allowed 2 1/4" between the inner to tank coil supports rather than the specified 2". My antenna link coil has only 5 turns rather than 7 like the article recommended. I found that 5 turns gave me more output at 12-15 watts input.


What sort of tank tuning cap is that? This Push-Pull Colpitts calls for 250pf across the tank coil with the capacitor rotor grounded and the tank coil connected to the two stators. This requires each section to be 500pf. What I found is a dual 550pf per section Cardwell capacitor. It is physically laid out with the two sections, both rotor and stator, turned opposite each other. It looks like a differential capacitor but it is really just an mechanically unusual two gang variable capacitor. Any 500-700pf per section two gang variable capacitor would work here.

While we are on the topic of the tank tuning capacitor...notice that many variable capacitors have stator connections on both sides. This allows a convenient path for routing from the plate of one tube through a feedback capacitor to the grid of the other.


What is the order of the connections along the back? See the photo below. Facing the transmitter from the outside/back, left to right: B+, B-/Gnd, Key+, Key-/Gnd, Fil CT, Fil, Fil.

Will this design work with other tubes like 45s or 27s? I've never tried a 27 or 45 in this circuit but others have run Push-Pull Hartley and TNT transmitters with 27s and 45s so they will probably work here. R, the bias resistor, will need to change if a different tube type is used with different characteristics. The Push-Pull TNT article in the November 1930 QST recommends a 10K bias resistor for type 10s and a 50K bias resistor for type 45s. 50K is a good starting point if planning to use 45s.

What's that insulator in the center of the tank coils? This insulator is there as a safety precaution. B+ is on the copper strip just below the antenna swinging link. I worried that the antenna link, if not tightly secured, might sag down and short B+ to my antenna system. This insulator limits the movement of the antenna link.


If I'm using 2.5V tubes like 45s or 27s do I need a 2.5 VAC filament transformer? The 27 has an indirectly heated cathode. The two 27 filaments can be connected in series and lit from the 5V winding common on many tube era transformers. 45 tube filaments, while 2.5VAV, are directly heated. I don't know if they can be operated the same way. Try it and let me know.

What sort of B+ supply do I use? I use a regulated 200-325VDC supply. 10s can handle higher B+ but why stress them and risk an expensive tube failure? 15 watts input is fine with me. The regulated supply also results in a more stable signal.


The bare wiring underneath looks like it might short together. Shouldn't that all be insulated? There is more room underneath the "deck" then you might guess looking at the photos. That space is 1.5" high by almost 12" x 12". There is plenty of room for three layers of wiring given stiff wire. In a few places where I needed extra stability I used brass screws as standoffs/supports.

This looks pretty dangerous. Shouldn't you have a cover so you don't get electrocuted? There are lots of places where B+ is exposed on this transmitter. This is not unusual for these early designs. Always shut down the power, checking it twice before doing anything behind the non-existent front panel or underneath the chassis. I also recommend leaving a metal screw driver (with insulated handle) levered between the B+ and B- power terminals while changing coils.








Saturday, February 29, 2020

A Dingey size Boat Anchor

I first started on this project over 15 years ago. I wanted a 6L6 based transmitter with a spotting function. Over the years it has gone through several iterations to the "Dingey Size Boat Anchor" I use today. Along the way I added a loading control, regulated the oscillator plate voltage to 108 volts, shifted the output circuit to a pi network and repackaged it in a metal cabinet to better match my HRO Sr. Like many projects it has been a journey. Here are a few more details about this transmitter as it stands today.

The circuit itself is fairly standard using a separate 300 B+ power supply. About the only thing I don't see in most designs is SW, the spotting switch. It allows me to check my crystal frequency against any station I might want to work or against QRM. As an important serendipity it also allows me to run the oscillator continuously while keying only the 6L6 output stage. This cuts down on chirp caused by the crystal restarting at the beginning of each code element. C10, the loading capacitor, is really two capacitors in parallel. One is a typical 300pf variable. The second is a 300pf high voltage "door knob" capacitor that can be switched it/out with a toggle switch.
Coils are what I could find. The 80 and 40 meter coils are commercial Bud coils that I had. I'm using only the tank portion, not the link. For 30 I found an old plug-in coil that looked about right and tried it. It works. In general look for coils that  dip and load to around 60% efficiency (ie. 6 to 7 watts output for 10 to 12 watts input).

Inside I arranged everything so that it made sense.  Across the back, left to right, is the 6J5 oscillator, the 6L6 final and then the plug-in output coil. Just to the right of the coil is the toggle switch that switches in/out the additional 300pf  loading capacity. The crystal socket is near the 6J5 and the 0C3 108 V voltage regulator is just to the right of that.

When changing crystals or, especially, coils it is important to recognize where high voltage is exposed. In this case the unprotected meter  terminals have 300 VDC, a lethal voltage, on them. Power this transmitter down when under the covers. This is even more important when working under the chassis. Don't take chances. Power down before working on it. As a safety modification consider placing the crystal socket on the front panel.

Are you looking for something more modern looking? Try  W1TS's 6C4-5763 MOPA in the October 1968 issue QST. Just add a resistor and 0B2 to regulate the oscillator B+ to 108V and include a switch to run the oscillator stage while keying the final.

Thursday, February 27, 2020

A Dingey size Boat Anchor on 30 mtrs

After putting my Drake 2B on 30 mtrs I immediately started looking for a matching transmitter.  The trouble is, the WARC bands like 30 mtrs weren't created until about 1980, long after the novice requirement for crystal control had gone away. By then solid state VFO controlled transceivers were becoming the typical rig. I'd need to homebrew a transmitter if I wanted vacuum tubes and crystal control on 30 mtrs.

A search of the QST archives brought up three articles that helped me decide what to do. The first was WD8DQT's article in January 2012 showing how to use today's HC49 crystals in simple MOPA (Master Oscillator - Power Amplifier) transmitters. The other two, W1TS's in October 1968 and WD8DAS's in January 2003 each show 6C4 to 5763 MOPA designs that can be modified per WD8DQT's article to use HC49 crystals.

What next?

I already have a crystal controlled MOPA transmitter on my desk. See https://w0vlz.blogspot.com/2008/02/further-6j56l6-transmitter-developments_5846.html . I did a quick check and found that the 6J5 I used is a close match to the 6C4 WD8DQT based his article on. I could start with the 6J5-6L6 transmitter I already had.  First I added a VR-105 and 12K 5 watt dropping resistor to power the 6J5 oscillator at 105 VDC. Next I converted the PA to a pi network instead of link coupled output. The tank capacitor is now 140pF and  the output/antenna capacitor is now 300pf plus, optionally, an additional 300pF. The coils on 80 and 40 are the same B&W output coils I've been using but without the output link connected. For 30 mtrs I searched through my junque box and found a coil that loads up fine on 30. Finally, I found HC-49 crystals, including 30 mtrs, five for $7.50 on ebay. See https://www.ebay.com/sch/possumlodge/m.html?_nkw=&_armrs=1&_ipg=&_from=

My "new" 6J5-6L6 transmitter uses FT-243 and HC- 49 fundamental frequency crystals. It runs almost 12 watts input and about 6 watts out on 80, 40 and 30 ... outstanding! 

See ya on 30.

Sunday, February 23, 2020

Putting the Drake 2B on 30 and 17 meters

Recently I noticed the optional crystal frequency chart in my Drake 2B manual. It shows that a 14.0 MHz crystal allows you to listen to WWV on 10 MHz and 30 mtrs on 10.1-10.150. Nice, but wait, buy one get a second one free! The chart in the manual also shows that by retuning the preselector that same 14.0 MHz crystal also covers 17.5-18.1 MHz. This includes the lower half of the 17 mtr band. It looked like shifting the crystal up 100 KHz would shift coverage to include the entire 17 mtr band and still keep WWV and 30 mtrs on the dial. Outstanding!

A quick check of AF4K's website at http://www.af4k.com shows that he stocks 14.1 MHz crystals in both the HC49 and HC6 style holders. Being cheap I bought the HC49 version for $12 instead of the HC6 plug and play option for $24. I mounted my new crystal on a hacked off HC6 base and plugged it in. It works fine.

Range C on my 2B now covers WWV, 30 mtrs and 17 mtrs.






Where is Solar Cycle 25?