[Amps] Ameritron Al-811H Diode String Grid Bias Question

[bills stuff]

Ameritron tech support suggested that an AL-811H oscillation/RX noise 
problem (arising when using the QSK-5 unit where the input floats during 
QSK Rx times) might be helped by putting in a string of diodes (6 
1N5408 rectifier diodes to give ~4.2 V bias) in forward bias along the 
existing cathode line from the filament center tap.  This will reduce 
the resting plate current by providing a cathode to grid ~4.2V bias. 
Some have written that this is desirable, even vital, for dissipation 
reduction (and some 811A amps do have such bias using a Zener or bias 
supply) independent of oscillation issues.  Some writers have raised 
concerns about the diode bias effects on IMD performance, but quoted no 
data - others say not to worry.  Spec sheets for the 811A do indicate 
-4.5V for grid bias for a class AB2 (e.g., the claimed AL-811H class). 
Anyone have any info on this topic, tried it or looked at the IMD (or 
other) effects of bias use?

Bill, N6MW