[Amps] Alpha 87a

[Tom W8JI]

> Pappenfus et al write about the length of cable adding to 
> the input tuning such as to make a complete number of 
> electrical half waves back to the plate of the driver 
> tube: the tube acting as a constant current source. under 
> those conditions, the input SWR isn't so important, except 
> when using a solid state driver. I would not be surpised 
> if  the length of coax was chosen with this in mind.

That theory absolutely does not work.

That theory would only work if the tube acted as a constant 
current source and if the impedance matching was a broadband 
transformer. Unfortunately the tube  does not look like a 
current source, and the tank is frequency selective. The 
tube is actually a non-linear time-varying resistance. The 
tank a low pass filter.

Tank circuit Q smoothes the tube resistance into a mean 
resistance value at the operating frequency. At harmonics 
the output reactance in the tank is the primary termination.

If the PA uses a pi-L, any harmonics are terminated into a 
moderate to high inductive reactance. If the network is a 
pi, the load is capacitive and very low reactance.

The harmonics simply terminate in the coax and whatever the 
tank output reactance is at the harmonic frequency. This 
reflects back on the tube cathode (in a GG amp) as an 
impedance in series with the cathode, and so cable length 
can greatly affect IM performance and efficiency of the PA 
if the PA has inadequate input filtering. An SWR meter will 
show the harmonic energy as imperfect SWR, even though the 
actual SWR is not changing!

The optimum termination at the cathode depends on several 
factors but as a general rule we want an input circuit that 
presents a LOW impedance to the cathode for all harmonics of 
the drive frequency. This means a moderate Q low pass C-L-C 
pi-network or parallel resonant tank at the cathode.

All that stuff about certain length lines and planned phase 
shift is nonsense since the impedance and phase varies with 
the type of equipment driving the PA and how it is 
adjusted....and it is never the same on harmonics as it is 
on the fundamental. The real trick is to stop the harmonics 
at the tube, then the cable length doesn't matter a bit and 
the PA works the same with ANY driver.

73 Tom