Covid Quarantine = clip-lead fun.

[KD1SH]

   My wonderful wife came down with Covid last week. No worries - just head-cold symptoms, and she's pretty much done with it now. It hasn't touched me - triple-vaxed (we both are) and likely immune to it - but that still means a week's quarantine for both of us, since though I'm asymptomatic I might possibly be carrying it around.
   So, great time to grab a slab of pine, reach deep into the junk-box, and see if I can constructively employ every single clip-lead in my inventory.
   The result was an 807 driven by a 6AG7 oscillator on 75 meters. It's currently putting out 30 watts with 500v on the plate, with efficiency better than 70%. Now I've just got to build a modulator for it. My only concern is that the best grid current I can get on the 807 is just over 3 ma, driving the grid with that 6AG7, which might not be enough under modulation. If it modulates, it might make an appearance in the AM Rally, but I probably won't have time to build the modulator before then. In any event, some involuntary home seclusion resulted in some fun.

[Carl WA1KPD]

Speedy recovery.
Why 2 variables in the 6AG7? Is it acting as a VFO?

You're gonna make Steve envious with all those leads.

[KD1SH]

It is indeed a power-generating VFO; the variable toward the back controls the frequency, and the one toward the front is for the oscillator output tank. A lot of this is very preliminary: both of those variables will be changed out for lower values at some point, with appropriate padding, because they've got far more range than I need and are way too twitchy to adjust. If this all works well, it might just find its way onto a real chassis and a case. Right now, it's just fun.

Speedy recovery.
Why 2 variables in the 6AG7? Is it acting as a VFO?

You're gonna make Steve envious with all those leads.

[KD1SH]

The 6AG7 gives 300V peak on the scope with no load, but by the time the load of the 807 stomps on it a bit it's down to 125V; still enough to get over the -90V bias on the 807 grid.

[Carl WA1KPD]

Very cool radio. Your own design or is there a scizmatic kicking around?
iI have the guts of a Johnson Adventurer (6AG7 to 807) that might be interested in morphing itself

[KD1SH]

Well, I'm sort of a "seat of the pants" guy, especially when I'm playing on a pine-board; it's pretty much a "little bit of this and a little bit of that" sort of thing, but the 6AG7 oscillator was inspired by this:
 https://www.frostburg.edu/personal/latta/ee/wing2e26xctr/schematic/oscillator/2e26oscillator.html

And the transmitter inspired by this: https://www.qsl.net/k5dh/6ag7_807.html

Iterations of the two-tube CW transmitter abound on the Web and old "East Coast" handbooks. In my case, of course, the intent is to amplitude modulate it, which implies a bit more grid bias on the 807, around -90v, where you could probably get away with less for CW only. I'm using a separate bias supply, in the form of my Heathkit bench high voltage supply, but think that for AM use we'd probably want to insert a series resistor between the grid RF choke and the bias supply to allow the grid a little freedom, or simply use grid-leak bias. Conventional wisdom has it that a tube will modulate better that way. Haven't worked out the best value for that resistor yet. What really remains to be seen is how much, if any, the 807 will pull the 6AG7 VFO around under modulation. Typically, a two-tube rig like this is either CW only or, if amplitude modulated, a crystal oscillator is used rather than a VFO. And I'm still not sure if that little 6AG7 will swing the 807 grid enough for good AM use. But, my plan is to use this as an exciter class rig - 15 to 20 watts - for driving an amp, so the modest grid drive might not be an issue. Fun stuff to play around with on a pine-board.

Very cool radio. Your own design or is there a scizmatic kicking around?
iI have the guts of a Johnson Adventurer (6AG7 to 807) that might be interested in morphing itself

[w9jsw]

I also plan to play with an 807 exciter, when I get done with the 4-400 RF deck currently under construction. My plan is to drive it with a W9BHI DDS board and attempt to get 20-30W of drive out of it. The DDS can swing 30V P-P.

How much swing are your seeing on the 807 grid? I may need to plan on a buffer stage.

[KD1SH]

I get a solid 125 Volt peak - 250V peak-to-peak - on the grid of the 807. I'm running 165 V on the screen of the 6AG7 oscillator; I might try raising it a bit  to see if I can get a tad more swing out of it, just for headroom. That 125 Volt peak is getting me around 2.5 to 3 ma of grid current on the 807, which might be marginal once the tube is under modulation.
For an 807 in class C, the recommended grid bias is around -90 Volts, so you'll need a driver that can swing the monkey high enough to get it over that fence; you'll probably need a buffer/drive stage with your DDS. 30V peak-to-peak would be only 15V peak; your 807 would probably never even know it was there.

I also plan to play with an 807 exciter, when I get done with the 4-400 RF deck currently under construction. My plan is to drive it with a W9BHI DDS board and attempt to get 20-30W of drive out of it. The DDS can swing 30V P-P.

How much swing are your seeing on the 807 grid? I may need to plan on a buffer stage.

[w9jsw]

I guess I should just build the k5dh schematic as is and feed the DDS into J1?

[KD1SH]

I haven't worked any numbers about what the 6AG7 would like to see on its grid, but I'd imagine that it would work. Very nice little tube.
My transmitter is very similar to the K5DH design, but the oscillator is the Frostburg.

I guess I should just build the k5dh schematic as is and feed the DDS into J1?

[WB6NVH]

I take it you don't have a cat, at least an indoor one...

[KD1SH]

Some progress updates, and notes and observations, as the project continues:

I've now got it putting out a whopping 50 watts, after increasing the screen voltage on both the 807 and the 6AG7, though I should take a closer look at the new value of screen current on the 6AG7. The 807's screen current is well within limits, and pretty much what I expected.

I'm finding that my Heath IP-32 bench high voltage supply is inadequate to the task. For supplying B+ and screen voltage to just the 6AG7 oscillator, it's fine, but supplying the bias to the 807's grid: not so much. Before I started incrementally increasing the grid drive, it was fine, but now the grid current of the 807 just yanks that poor little supply all over heck. Very nice little supply but out of its element when working with transmitting tubes, even small ones like 807's. I'll need to kluge up a stiffer bias supply. Simple enough.

I find that varying the output load on the 807 will pull the oscillator around as far as +- 500hz. Well, my poorly regulated grid voltage on the 807 - see above - might be contributing to that, and sitting on a pine-board as this thing is, isolation between the two stages is dubious at best.

I changed out the overly huge oscillator variable for a smaller one, with a fixed padder, which makes setting the frequency much easier. The old one would cover way above and below the 75 meter band and was incredibly twitchy.

Got to remember to turn on the bias supply before the B+! Look - the 807 is drawing 150ma. Not terribly excessive for an 807, right? Hah! no RF out, which means that every single watt is being dissipated as heat in the 807's plate. All my old beater 807's are test pilots - the Chuck Yeagers of vacuum tubes - if they were smart they'd run off and hide somewhere when they saw my pine-board come out.

When calculating values to build the resonant output tank for the oscillator, you've got to consider the capacitance of the 6AG7's plate upstream and the 807's grid downstream, as well as any coax you might use to connect them. I wound up needing a whole lot less capacitance in that tank than I figured.

All in all this has been a ton of fun, and I'd highly recommend pine-board projects for others like myself who know their way around the circuits but aren't necessarily experienced power-builders like Steve or Tom. Other than the obvious hazard of exposed high voltage, it's an awesome learning experience.

 All the while I'm working with this thing, I'm thinking that, hey, instead of wasting my time with this, I could be watching reality TV or doing FT8. Not!