[Amps] RF voltage on VHF strip line

[John Lyles]

 >Message: 5
Date: Fri, 28 Oct 2016 15:37:58 -0400
From: N1BUG <paul at n1bug.com>
To: 'Amps' <Amps at contesting.com>
Subject: [Amps] RF voltage on VHF strip line
....
 >Suppose you have a 4CX1500B sitting in the middle of a half wave plate 
strip line for 144 MHz. Some may prefer to think of it has two quarter 
wave lines with the tube between them. This is a "teflon sandwich" line 
where the upper brass plate is connected to the tube anode, the lower is 
not. The lower brass plates are grounded at the far end from the tube. 
Between the upper and lower plates we have a thin sheet of teflon that 
forms the dielectric for the DC blocking capacitor.

 >Can someone help me understand whether (and if so why) there would be 
significant RF voltage across the blocking cap? Obviously RF voltage is 
going to vary along the *length* of the line, being high at the tube and 
near zero as we get toward the grounded end. But what about voltage 
across the teflon sheet between the upper and lower plate *at any given 
point along the line*?

 >I am trying to understand WHY this thing would want to arc or break 
down only when RF is present. My mind wants to believe that the RF 
voltage across the capacitor at the point where the teflon shoulder 
washers insulate the screws holding it together would not be 
significant, and that the voltage would be just B+ regardless of whether 
RF is present or not. However I see evidence this is not the case and 
want to know what it is that I don't understand.

---------------
Reply:
With respect to earth, the RF voltage swing is largest at the center of 
the tube. But across the dielectric of the blocking cap, the RF voltage 
should be low. The capacitive reactance is Xc=1/(6.28xFxC) and if the 
blocker is built right, the C is large enough that Xc is very low, few 
ohms or less of -j. RF current through the capacitor should also be 
reasonable there, being a current node (min) near the anode. Reactive 
voltage across the cap would then be I x Xc.

A good design might have a few hundred volts of peak RF voltage 
superimposed across the blocker with a significant DC voltage present. 
If the blocker has some sharp edges then the RF voltage can easily 
initiate partial breakdown, i.e., corona effects around the edges. The 
most intense electric fields are near the point where the dielectric and 
the electrode (metal plate) are exposed with air at the edges, the 
so-called triple point.  This is where capacitors will usually fail with 
RF voltage first. And the shoulder washers of a Teflon blocker are 
places for E field enhancement due to geometry.

Another consideration is the harmonic voltages, if they are getting 
enhanced by some 'feature' of the layout. There could be higher voltage 
at one of the harmonics due to the standing wave structure across the 
resonator. At higher frequencies, the blocker may not exhibit a purely 
capacitive reactance, and could even have some parasitic resonance 
associated. At this point, it could have an RF voltage superimposed on 
the DC. I haven't measured such a capacitor myself with a network 
analyzer or UHF Z meter so am just postulating.

73
John
K5PRO