Amps
[Top] [All Lists]

[Amps] 4CX800/GU74B Socket

To: <amps@contesting.com>
Subject: [Amps] 4CX800/GU74B Socket
From: 2@mail.vcnet.com (Richard)
Date: Fri, 14 Jun 2002 05:59:53 -0700
>Hi Guys...
>
>I'm about to start construction for a 2 meter linear using a GU74B (
>4CX800 ). The socket I have is a military one, with built in chimney and an
>external screen ring. This ring makes contact with the screen area of the
>valve.
>
>At present it has 4  x 1000pf 1kv disc ceramic caps between the screen
>"ring" and ground.
>
>My question....
>
>(a) Is this socket useable at 2 meters

A low-Z screen bypass is essential unless one is building an oscillator. 
Depending on lead lengths, a 1000pF disc-cap exhibits capacitive 
reactance up to around 50MHz - where it is resonant.  Above resonance, 
capacitors look progressively inductive -- i.e., the screen will not be 
at RF zero potential at 144MHz with the original 1000pF capacitors 
because they look inductive.   This is problematic because a substantial 
current flows from the anode to the screen at 144MHz - which creats a RF 
potential on the screen.  Thus, the screen will not be at RF ground.     
My choice for screen bypass caps at 144MHz would be 100pF to 150pF, with 
not all 4 of the same value.  [note - to measure the reonance of a 
disc-cap, clamp it in a vise with Cu foil across the leads and find the 
dip freq with a dipmeter].  
-  To further muddle the picture, one also needs to compensate for the 
inductance in the tube between the screen and the screen ring.  Thus, it 
is better to go with less screen-bypass C rather than more.

>
>(b) I note that about 1100pf  is usually the screen capacitance in eimac
>sockets for the 4CX250...can I just solder an even number of disc ceramics
>to the socket to total about 1100pf, or am I missing something ??
>
The inductance of the Eimac screen bypass C is way less than that of disc 
 ceramic caps. 

cheers, Glenn

Murphy was right:  Things are often more complicated than they look.  


-  R. L. Measures, a.k.a. Rich..., 805.386.3734,AG6K, 
www.vcnet.com/measures.  
end


<Prev in Thread] Current Thread [Next in Thread>