simple tube VFO with mixer

[PA0NVD]

I did like to make a stable VFO with tubes that has less problems with feed back and extensive screening as many VFO's have that oscillate at the basic frequency.  This mainly because I don't have here the possibility to do extensive mechanical work and things like aluminium plate is not available here. So I tried a mixer based VFO
For the variable oscillator I selected a Franklin oscillator at a low frequency like  1 - 1,5 MHz. The Franklin oscillator isolates the tube quite effectively from the resonance circuit, so tube load and capacitance will not influence the output frequency that much as other oscillator do. That gave me the possibility to use almost any kind of tube. The stability depends mainly at the resonance circuit and not at the tube . For the crystal oscillator no special frequency is needed because one can adjust the range of the VFO to get the desired output frequency. So no hunt for crystals. The crystal oscillator is at a higher frequency and will normally give quite a high output birdy because it is close to the output frequency, so I selected a balanced mixer in order to suppress the crystal signal in the output.
For my test set-up I used a 6 MHz crystal for the 7 MHz band, so the VFO ran from 1 - 1,5 MHz. The tubes are 6J6, but many other double triodes can be used as well. The design can be adapted to all frequencies required. I used it for 7 Mhz.
The VFO gives approx `13 dBm output and all undesired signals are approx 40 - 50 dB down, quite nice. The output can be a coupling loop to L3 of approx 5 turns to get 50 Ohms. L2 and L3 have 40 turns and are 18 uH. But the value is not critical at all.
The oscillator coil L1 has 45 turns, the coupling is a 4 turn CT winding at the cold side.  The mixer is a self oscillating mixer. The crystal signal is nulled using a small 30 pF trimmer at one of the anodes while the other anode is loaded with 15 pF to ground.
The output filters L2 and L3 are high LC ratio (high impedance) filters so the tube anode impedance will reduce the Q to a low value where the bandwidth of the filter is 400 kHz when K.Q = 1 - 1.5   I used coils of approx 18 uH tuned with a 40 pF trimmer
The output is 10 - 13 dBm. So a simple penthode amplifier can be added to increase the output to 1 Watt or more.
The HV is 150 V and the current consumption is approx. 13 mA . The frequency drift due to HV variation is close to 2 Hz / Volt, so HV stabilisation is hardly required. Even the simple construction at PC Board did result in a frequency drift less than 20 Hz in one hour after 10 minutes warm-up  So, no more rock-bound frequencies required!

[WA4WAX]

Check out the circuit on page 142.  Class AB oscillator.

https://www.americanradiohistory.com/Archive-Radio-News/50s/Radio-News-1957-06-R.pdf

Ideas on the VFO:

1) Put the LC circuit in a polyfoam insulated box. 1/4 to 3/8 inch thick will do nicely.

2) Wind solenoidal filament chokes on 6 or 7 mm forms, with one coil in each lead.  The goal is to get the cathode and filaments at the same RF potential.  This will virtually eliminate frequency shifts caused by variations in cathode to filament capacitance.  Check with RF probe.  Run filaments off a battery while making this measurement.

3) Use nice brass plate, bearing movement variables.  25 to 35 pF  (No more!) if single section, 75 pF per section max if dual section.  Use reduction drive. Make connections with short, heavy pieces of wire. Protect tube and LC from mechanical vibrations.

4) Adjust mesh screw on the back of the capacitor in order to get rotor and stator plates evenly spaced.  Trick: Tune oscillator in on a receiver, CW setting.  Adjust for convenient beat note.  Now adjust mesh screw.  Beat note will be an minimum frequency when mesh is perfect.  This measure will also reduce drift.

Have fun!

[PA0NVD]

But why should I do it that complicated? I can't get any brass plate here.  The Franklin oscillator at 1 - 1.5 MHz is rock stable. Is 20 Hz / hour and 2 Hz per volt pushing not stable enough? And what when there is a VERY SMALL feedback at the VFO frequency from the transmitter? A VFO running at the transmitter output frequency is VERY prone to feedback and lots of shielding effort is required. For that reason I have the oscillator running at a low frequency and I use a Franklin oscillator, to prevent chokes and whatsoever influence of the tube at the stability.
And the cathode is grounded in the Franklin oscillator, so NO rf at the cathode, no chokes required.
I made many direct VFO's, and this approach takes a whole lot of problem away.

[Opcom]

soldering a small vertical strip of pcb from one end to the other of the 'plate' of PCB material will stiffen it, but it does not seem like that's much of an issue.

[w3jn]

A great use for the common and unloved 6J6 - thanks for posting this novel design! 

[W7TFO]

Cool design, I've always loved heterodyne oscillators.

Kudos!

73DG

[WA4WAX]

Do not get me wrong.....I like your design.  It is always nice to throw some of this stuff around.

Here is another interesting tube VFO.  Please see page 41.

http://www.electronicsandbooks.com/eab3/manual/Magazine/Q/QST%20US/195303%20[148].pdf

[W7TFO]

I'm going to work that up with 6A6's for my retro creations rather than the 6J6.  Almost the same Mu, ought to be OK with the proper english applied.

73DG

[PA0NVD]

soldering a small vertical strip of pcb from one end to the other of the 'plate' of PCB material will stiffen it, but it does not seem like that's much of an issue.

That's a good idea, I will do that. PCB material is nice, but not really stiff.

[PA0NVD]

Do not get me wrong.....I like your design.  It is always nice to throw some of this stuff around.

Here is another interesting tube VFO.  Please see page 41.

http://www.electronicsandbooks.com/eab3/manual/Magazine/Q/QST%20US/195303%20[148].pdf

I like to see that design, but I get a 404 error, page can't be found

[PA0NVD]

I'm going to work that up with 6A6's for my retro creations rather than the 6J6.  Almost the same Mu, ought to be OK with the proper english applied.

73DG

That will be interesting. For which frequency will you make the VFO ?The 6A6 is also quite nicely balanced in the connections, so it may work well. The 6J6 was fine for that reason, easy to null the Xtal frequency out of the output due to the balanced structure. It may be required to increase the nulling caps at the anode slightly with the 6A6. Please keep me informed, I am very interested in the results.

[W7TFO]

75m direct.

73DG

[PA0NVD]

The crystal is nulled out of the output, but take care for the third harmonics of the VFO. That may be a strong birdy. Select the frequencies such that the third harmonics is not present in the output, e.g a vfo that runs from 1.5 - 2 and a 2 MHz crystal or a 5.5 MHz crystal.  Or perhaps better for a 6A6 a low VFO from 0,5 - 1 MHz and a 3 MHz crystal or something near. Than you may use a normal broadcast capacitor of 500 pF

[Opcom]

A great use for the common and unloved 6J6 - thanks for posting this novel design! 

Poor 6J6s.. Got a carton of them in military-like boxes stamped for Sperry Gyroscope. Wish they were more linear for audio, nice comon cathode to make a long tail pair.. I am a fan of feedback but nonetheless. They sit on the shelf next to the cases of 12BQ6s and 6GH8, 70L7s, and 3Q5s I think.. I must love the unfortunate.

[Opcom]

Do not get me wrong.....I like your design.  It is always nice to throw some of this stuff around.

Here is another interesting tube VFO.  Please see page 41.

http://www.electronicsandbooks.com/eab3/manual/Magazine/Q/QST%20US/195303%20[148].pdf

I like to see that design, but I get a 404 error, page can't be found

The [148] must not interfere with the url, but is part of it. This confuses some BBS software.
The  [ brackets ] are used on this BBS for psuedo-html, the markup permitted here. BBcode, it is usually called.

One way is to Copy and Paste the entire line into the address bar.
http://www.electronicsandbooks.com/eab3/manual/Magazine/Q/QST%20US/195303%20[148].pdf

Another way is to convert the [ and ] characters to hex values %5b and %5d as can be seen. This normalizes the URL for this BBS software setup which uses BBcode.
http://www.electronicsandbooks.com/eab3/manual/Magazine/Q/QST%20US/195303%20%5b148%5d.pdf

The url was written by the keepers of that site without consideration of message board compatibility. oh well nothing's perfect.

[PA0NVD]

Interesting approach to isolate the resonance circuit from the tube. But I think that the approach of Franklin is preferable and givers you more parameters to design well. Thanks for the link.

[w4bfs]

A great use for the common and unloved 6J6 - thanks for posting this novel design! 

Poor 6J6s.. Got a carton of them in military-like boxes stamped for Sperry Gyroscope. Wish they were more linear for audio, nice comon cathode to make a long tail pair.. I am a fan of feedback but nonetheless. They sit on the shelf next to the cases of 12BQ6s and 6GH8, 70L7s, and 3Q5s I think.. I must love the unfortunate.

the 6gh8 makes a better replacement for a 7199 ... I have one in my Ampeg B25 bass amp and a few trace cuts and resistor changes were all that was necessary .... better sound

[Opcom]

second that. Replaced 7199 with 6GH8s in Stromberg Carlson amps. Two 'audiophiles' said the amps sounded amazing.