When a dual section split stator tuning cap is used with a pair of tubes in push pull, what value do you use to determine how many tank coil turns you'll need? the entire value of both sections combined or only one section?
I have always found it easier and less trouble to find the number of turns by trial and error, than by calculating mathematically the inductance required, and then calculating the number of turns for the size coil I wanted to use, to get that inductance. I use the tank
capacitance as the starting point.
The capacitance needed to resonate the coil is based on the total capacitance across the coil. It can be done with a single section, but it is better to use a split stator. Each section of the split stator needs to have twice the resonating capacitance, but the plate spacing can be reduced, since only half the rf voltage appears across each section.
The amount of capacitance is based on the plate voltage/plate current ratio. You want a Q of approximately 12 for plate modulated AM. You could probably get by with less Q running push-pull than single ended, since the resonant circuit is getting two kicks in the ass per rf cycle from the tubes, instead of only one.
The formula for calculating the amount of capacitance can be found in the older Radio or ARRL handbooks. I would look in the ones published in the era before the pi-network became the standard for ham transmitters. You basically come up with a formula for picofarads per metre.
You want the capacitor to max out at maybe 25% above the calculated value, for the lowest frequency band. You may need the extra margin just in case the load is reactive enough to de-tune the tank circuit, so you can always hit resonance and dip the final as you tune up, plus you want to CYA regarding component value tolerances.
My computer is in the house and all my radio books are in the shack, so I can't look up the formula right now and don't remember it, but to give you a rough idea, my HF-300 final was designed to 2000 volts @ 500 MA. That is the equivalent of 4000 ohms, and that is the same value as the modulating impedance. My tank capacitor is 240 pf per section, and it runs about 70% meshed on 160, and only about 25% meshed on 75. If I had planned to run it on 75 but not 160, I could have had the same result using 120 pf per section. My first QRO rig, a single 304-TL, ran 2500 volts @ 400 MA, for 6250 ohms. The capacitance needed to maintain a given Q is
inversely proportional to the modulating impedance, so that rig needed 4000/6250, or 64% of what the present rig uses. That would calculate out to approximately 150 pf/section. If the lowest frequency band was to be 75m, 75 pf/section would have been sufficient. 6250 ohms modulating impedance calculates out to about 1 pf/metre for each section of the split stator cap. Extrapolate to other plate voltage/plate current ratios by calculating the inverse proportion.
Another standard for comparison is the BC-610. It was designed to run 2000 volts @ 250 MA, or 8000 ohms modulating capacitance. The final tank capacitor is 150 pf/section. It will tune 160 without the extra fixed vacuum padders if the 1.5-2.0 mHz coil is used. With the 80m coil installed, it will hit 75 with about 30% mesh, as I recall.
To figure out the number of turns needed, I have always wound a JS prototype coil for each band, the same diameter and spacing as the finished tank coil, but with more turns than I thought I would need. Then I would put a link over the coil and load a low power transmitter into it. A couple of watts would be enough. Then put the capacitor across it, and tune for resonance as indicated by a neon bulb or a kick in the transmitter plate current. A good grid dip meter could be used just as well. Then I would trim the coil until I got resonance at the desired capacitor setting. For 160-80-40 mine run about 70%, 25%, 15% respectively. When I used the HF-300 rig on 20, it hit resonance with the capacitor just barely meshed, probably 5%. If you don't design it for 160, make it hit resonance at about 80% full mesh at 75m, etc. Once you figure out the number of turns and diameter for the prototype, then you can build the final version of the coil.
Over the years of building rigs, I have become pretty good at guesstimating the size of the coil pretty close, first try.
I have a dual section 12 to 74 pf per section @ 4.5 kV thats very compact yet can take big voltages. Would make a nice cap for one of my H&K decks, as they are only 10" long. but I dont think it's got enough pf to use it on 75. I think it would probably be ok on 40 tho, my gut feeling is.
As explained above, that should be sufficient for 75m if the plate voltage/plate current calculates to somewhere about 6000 or higher. Another thing, the tube capacitances and neutralising capacitors will add to the tank capacitance, so if the prototype hits resonance @ 95% meshed, it will probably be closer to 75% meshed in the final circuit with tubes installed and neutralising caps adjusted. So that gives you a little fudge factor.
H&K says to only use a higher than 100 pf tuning cap if the tx will be used on 160, if so double it.
I assume they are talking about per section. That sounds about right. If you design it for 160, for most final tubes and plate voltages you will need more than 100pf/section, but if for 75m, 100 pf will probably be enough.
If you make it here next week, we can look in some of the old handbooks for the exact formula. I also have a box full of scrap coil stock, so we might be able to find something to build the coil with. If you need the info right away, let me know and I'll look it up this evening.
Don
