Amps
[Top] [All Lists]

[Amps] QRO Balun?

To: <amps@contesting.com>
Subject: [Amps] QRO Balun?
From: RFlabnotes at aol.com (RFlabnotes@aol.com)
Date: Tue Apr 15 14:38:28 2003
Phil:
       I agree with your overall opinion that this was an economic tradeoff. 
However, you are oversimplifying a major point, which was what aroused my 
attention. That is the issue of power loss. The core materials all have a 
very strong dependence on power level, almost like thresholds. They all work 
well at low levels, but increasingly go to hell as power and frequency 
increases. 

       When you say that they are fine when operated within their power 
limits is like saying that 'everything is okay is when it is okay''. In 
reality, the actual flux level in the core (which is the origin of the loss) 
is determined by a number of variables within any one circuit. I have seen 
very few people who are even able to properly determine its value, either 
theoretically or experimentally. 1 db of loss in an RF transformer is 
virtually effortless to produce as power levels get above 100W, in reasonable 
size cores. What I am trying to say is that it is not a trivial job to 
calculate the loss from basic core data (some of which is not even 
available). It IS easy to measure it - just hold the key down for a few 
minutes and feel the core with your hand.

       I would guess that you come from that faction who believe that 
manufacturers are infallible and totally honest. If they tell you that this 
core in this amp has no losses, then you believe them without reservation.

       Well, I am not so believing, based upon my own practical experience in 
building RF transformers. As far as the transformers that antenna tuner 
designers have been using for years, as noted,  I have my suspicions about 
them too. I may attempt to try some measurements on one, and I will let you 
know whether or not it will handle 1.5kW into a 500 or 1000 ohm load as per 
the mfrs. specs.

       Meanwhile, I am still hoping someone will get me a set of dimensions 
for the actual core in question, so I can do my own arithmetic. 
73
Eric K8LV
      

 

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