[Amps] Odd efficiency shift on 80m band

[Jeff Blaine]

Gentlemen,

2xGI7B HB amp.  Shows a funny efficiency shift on the 80m band and I was 
hoping that some of the more experienced hands on the board may have 
seen something like this before....

On the top of the band (only), the calculated efficiency 
[(Po-Pdrive)/Pin] varies quite a lot on the 80m band..

4.00 mhz      61%
3.75 mhz      57%
3.50 mhz      55%

On all other bands, the efficiency is essentially flat over the range.  
I can understand a significant variation on say 10m, but on the low 
band, the efficiency of the amp should be more linear across the band as 
stray L/C are a much smaller % of the lumped values.

Things tried:

1. Replaced filament choke (much more L)

2. Increased coupling capacitance (2x C)

3. Varied the Q of the input matching network (Q=1.5-3.0)

4. Varied the L of the tank (+/- 15%)

Things not tried:

1. Increase plate choke L (it's a RF parts/MFJ stock unit of 225uH with 
no resonances in the ham bands).  This is the stock choke for all 
Ameritron amps.  Approximate Xl=5K ohms @ 3.5Mhz vs an estimated 1800 
ohms plate load Z.  Seems like it should be plenty...

None of these things made any difference.  Sure would appreciate any 
suggestions on where to look next...

73/jeff/ac0c