Discussion in 'Amplitude Modulation' started by KC2ZFA, Aug 29, 2021.
yes, the 807 plate coils are wound on forms that place them perpendicular to the TVL coils above (see pic below)...the pic above showed a manufactured coil used for testing.
The point is well and wholly taken...the final home for the rig is a Bud cabinet and by the time I have it refinished I will also have a totally shielded mopa exciter in place of what I have now (which on it's own, at 40 watts out, is audible on sdr's across the country).
I like a lot of grid bias in class C.
On the stuff I have built, I make the grid leak a big wire wound pot plus add fixed bias with a variac through a diode.
That way I can run any class, different tubes, and experiment with the settings and values to get the best results.
More complex and takes up space though.
I had a nice push pull rig in the past and plugged in 812a's, 811a's, and 572b's with about the same results.
I think the 812 and 811 tubes were interchangeable and seemed to give the same results!
I never had a better working/tuning rig than that push pull setup.
For best modulation linearity, as much of the bias as possible should be grid leak. In my homebrew rigs, I use a grid leak to generate the full bias, and a fixed protective bias supply is connected in parallel through a diode, so that as soon as the grid leak bias exceeds the fixed bias, the diode effectively disconnects the fixed bias from the circuit.
I tend to run more bias than recommended in the tube charts, driving the stage deeper into class-C, which improves modulation linearity. The grid leak is set with a higher resistance value to generate the increased bias at the normally recommended grid current.
Of course, this requires additional driving power to be delivered by the previous stage.
getting closer…260 watts into a matchbox and then 40 feet up to an 84-foot doublet on 3.880. No reflected power. Decent signal, a few seconds of carrier, onto the wa1sth websdr receiver.
I’m seeing very high efficiency ~80-82%, is it explainable ?
and on a websdr on Northern Utah on 7290. 260 watts out.
The nice thing about zero bias tubes is that a protective bias supply isn't needed.
The bad thing about zero bias tubes is that unlike the older triodes, you can't really put them deeper into class C beyond the data sheet bias v. (which is usually 5 or 6 times cutoff anyway) because you'll exceed the rated grid dissipation power. Whatever triode you are running, find out how much power the grid can dissipate and take that into consideration when calculating your needed grid leak value and grid current drive.
On a related topic, you mentioned that you will leave the HV on during RX. I elected to turn off my HV on RX. That proved of benefit when I went to using a grid leak resistor. I now only generate GRL heat when transmitting. On RX, no heat. You may want to reconsider your thoughts on HV control.
What has the grid leak got to do with the HV?
Grid leak is driven by the drive power, not the HV on my transmitters.
Bleeder resistors ARE in use, maybe that is what you meant?
Some just set the bleeder value high and let the tubes conduct a bit in RX acting as bleeders.
I like a robust choke input supply and just key it.
I used to do step start but have bypassed that as I found it not needed, even with a lot of filter capacitance.
In operation, I could detect no difference at all between 812a's and 811a's in my push pull rig.
A pair of 811a's modulated by a pair at 1200 up 1750 volts makes a simple setup!
A string of diodes to add some bias to the modulators is easy and cheap, add a switch to select the bias amount.
I have adjusted the bias all over and could hear no difference really...