[Amps] [rfamplifiers] Unexpected loading capacitance

[Steve Flood]

Looking further into my lower-than-expected power output on my 3x813 amp....

1.  Checked all tubes in my 813 supply one at a time.  Only 1 out of 12 had slightly lower plate current than the rest.  I selected 3 'good' ones.
2.  Checked calibration on my Ig and Ip meters.  Good there.
3.  Power supply sag on key down is 3100 down to 2850 on the high setting, and 2500 down to 2200 on the low setting.
4.  Checked my input network again for any mismatches.  No problems there.
5.  Checked continuity of all  three plate leads to the plate choke.  Good.

The result is still the same:

- I have to use much higher loading capacitance than what I found tuning the tank with the MFJ259B on the ouput.
- My power output is much lower than I expected.  I should be making over 1 kw, and I can only get 850 to 950w.

I'm using the G3SEK adjustable zener bias circuit for about 3 to 27v of bias.  I brought the 10k adjustment pot out to the front panel so I can have another knob to twiddle.

So, here's my next questions.

1.  Do I figure input power based on the sagged or unsagged plate voltage?  On 160m, on the "high" power supply setting (3100v), I get 2850v and 575mA at full 100w drive and just under 150mA Ig.  That makes about 58% efficiency at 950w output.  I am thinking to raise efficiency I need more bias?  I was hoping to run Class B at 65 to 70%.  Is that reasonable?

2.  Still can't resolve the issue with way-higher-than-expected loading capacitance.  With the MFJ on the ouput and the 2.5 mH output safety choke disconnected, I could get the tank to tune to 1:1 SWR easily on all bands.  Of course I reattached the safety choke, and try it out with real RF, and the load cap settings require much higher capacitance.

What to try next?

Steve
KK7UV