Measuring range of tank and input circuits with AIM4170 antenna analyser?

[BASIL]

I am trying to convert a 2000D Henry RF generator built for 27.1 MHz to run on
the 15 meters band. Initially, leaving the 3CX3000A7 tube in it, I have started by
seeing what range the standard tank circuit and input circuit has. The tank coil
is tuned by a brass slug entering into it on a threaded rod. The cap is a "flapper" cap,
two sheets of brass separated by a Teflon sheet. The moving sheet is curved and
as it presses against the Teflon separator, under the influence of another threaded
rod, it flattens, and the capacitance rises.
I did a custom cal to a 5 foot length of RG-58 BNC terminated coax on
my on my AIM4170C analyser and squirted a frequency range from
20 to 30 MHZ into the RF output socket of the deck, with tank coil and flapper
cap at maximum, and then again with them at minimum. I then put the signal into into the
exciter (input) socket. The tube was still in the deck. I have put the results at:

http://www.gatesgarth.com/henry/scans.zip

I can't understand why a generator set to run at a fixed 27.1 MHZ gives the
input scan that I have produced with a resonant frequency of 25.065
MHz? The input coil has a ferrite trimmer, but is untouched from when
the generator came out of working service. Should it not be at very
near 27.1 MHZ? Is my test procedure flawed?
The range of the tank circuit seems sensible.
Thanks!

[K8AXW]

I'm not that well versed on amplifier design and operation but I have read where, in order to tune the input circuit the amplifer needs to be operating.  In other words, HV applied and loaded into a dummy load.

No doubt this question will be answered shortly by some very knowledgeable people!

[KG6YV]

The initial comments are correct.  Without the tube powered up, you are probably not seeing the load impedance for the input network that is reflected when the tube is pulling grid current.  You might check W8JI's website to see if he has posted any technical papers on the subject.

Greg

[K6AER]

What bands are you wanting to operate?

[BASIL]

I am wanting to see if I can get it on 15 meters, but even 27 to 21 MHZ is a circa 25% change.

I have done some more reading and my belief that just having the valve present in its socket would reproduce the same impedances as when it's powered up were totally wrong. I have added a 1700 ohm resistor (carbon film, don't have any old carbon composites suitable for RF yet) across the anode to ground, and a 50 ohm to mimic the valves input impedance, cathode to ground. I am not sure I have the input impedance match correctly done yet, although I am using Eimac's specs for this valve, at

http://www.gatesgarth.com/henry/impedance.jpg

the input graphs still seem "wrong", but the tank stuff looks better again. By squashing the tank coil up tighter (circa 10mm OD copper tube) I can get the resonance down into the 21 to 22 MHz range, with the brass slug right out.

http://www.gatesgarth.com/henry/output_ ... sistor.jpg

Maybe a new coil with say 2 more turns, then trim to suit? Will have to do some more research on imitating the valves input impedance, I can't believe it's miles off 21.7 MHz.

http://www.gatesgarth.com/henry/input_w ... sistor.jpg

Thanks for the replies.

[K1ZJH]

I used my CIA-HF analyzer to optimize my SB-220 conversion. (Full WARC coverage and 160 meters.) 

You can do the testing with power removed. You need to add a non inductive resistor at the cathode to approximate the input Z for the cathodes. Same at the plate choke, add a resistor to the chassis that is equal to the plate impedance.  This will get you very close to the optimum values.

Pete