Help designing PP tank circuit

[W8ACR]

I am in the process of designing an AM transmitter with push pull T-20's in the final. I plan to have about 600V @ 150mA on the finals. I understand that a P/P tank may have a Q as low as 6 or so. I will probably design mine to have a somewhat higher Q. One detail that I haven't been able to quite figure out is how to calculate the inductance of my coil so that I get the proper Q value. This will be a single band 160 meter rig, and I have a B&W 160BVL coil. The BVL coil should be adequate for this power level.

Here is my question: do I calculate the total inductance by figuring out the inductance of each half and adding them together, or do I calculate the inductance by considering the coil as one long coil and calculating the inductance of the entire length as if there were no gap for the link. By method one, I get 53uH and by method two I get 72uH. This coil is 2.5 inch dia, 19 turns each side, and 1.0 inch length each side. Appears to be about 20 or 22ga wire.

Perhaps ther is an even easier way..... maybe there is an old B&W catalog that lists the inductance of their various coils.

[k4kyv]

Calculate the inductance of the entire coil as if there were no gap, since there will be mutual coupling between the two halves of the coil.  If the two halves were widely separated or oriented 90° from each other, so there were no mutual coupling, then you would calculate each one separately and add the inductances together.

In all the years of building transmitters I have rarely, if ever, bothered to use the formulae calculate inductance and then the physical size and number of turns for  the coil.  I figure out from the charts and tables what the optimum resonating capacitance is, then by trial and error, find how many turns of coil it takes to hit resonance with a coil the appropriate physical size and wire gauge, using that given amount of capacitance.

I always would shoot for a Q of 12, whether pushpull or single ended.  Better to be a little on the high side than too low.  Class-C finals work best when there is plenty of flywheel effect, and the extra Q gives better harmonic suppression.

I notice the older handbooks recommend a minimum Q of 12, whereas later ones recommend 10.

The reason you can get away with lower Q with pushpull is that the tank circuit gets kicked with a pulse every half rf cycle or twice per cycle, whereas with single ended, it gets a kick only once per rf cycle.

I usually go by the charts in the late 30's through early 50's handbooks, which recommend the  resonating capacitance based on plate voltage/plate current.

The total capacitance for a balanced tank circuit, whether pushpull or single ended, is one quarter that of an equivalent single ended circuit, since the capacitance in the balanced tank is split in two, and the two sections are in series.  If a single section cap is used in a balanced circuit, think of it as equivalent to taking the single ended capacitance, splitting it in two, and wiring those two sections in series.  It will need twice the voltage rating, or slightly more than twice the plate spacing of that of a single ended circuit.  If a split stator cap is used, the spacing of each section may be the same.

In fact, the same split stator capacitor may be used in either balanced or unbalanced configuration.  For unbalanced, tie the two sections together in parallel, to act as a single capacitor.  For balanced, connect to the two stator sections, which connects the two sections in series, with symmetry to ground.

For the same voltage/current ratio, the coil should have twice the number of turns for balanced as for single ended.  Twice the number of turns gives 4 times the inductance to resonate with a quarter of the capacitance.

The Q is the same in each case because in the pushpull class C circuit, only one tube functions at a time.  Effectively, each tube is tapped down across half the tank coil while the other half remains idle, which means that the entire coil is acting like a step-up transformer, and the tubes each take turns seeing a load impedance that is one quarter of the impedance of the whole coil.

[W8ACR]

Don, Thanks for the comments. As it turns out, the 160 BVL coil actually calculates to 51uH total inductance - I had miscounted the number of turns earlier. This should resonate at about 1955KHz with a tank capacitance of 130pf. That amount of capacitance should yield a Q of 11-12 with Ep=600 and Ip=.150. In other words, the coil value is just about right. Do these numbers sound right to you?

Also, my tuning capacitor is a B&W JCX100 - 100pf per side, 50pf total if connected in series. I will have to use a padder to get the additional capacitance. Is it OK to use a single HV transmitting capacitor (say 100pf) in parallel across the plate leads or should I use two 200pf capacitors in series, paralled across the plate leads, to maintain balance. Have you ever heard of someone using two air variables in parallel?

I plan to use crystal control with a 6AG7 osc and 2E26 or 6L6 buffer. I need 8 or 9 watts to drive the T-20's with a little to spare. Do you think the buffer tube is necessary? I could probably drive the finals directly from a 6L6 oscillator.

I am hoping this project will be a warmup for a larger rig - 8000's modulated by 810's. I've got most of the parts already, just need some mod iron. However I wanted to try a smaller and simpler rig first.