Towertalk
[Top] [All Lists]

Re: [TowerTalk] tuners and power rating

To: <towertalk@contesting.com>
Subject: Re: [TowerTalk] tuners and power rating
From: "Paul Christensen" <w9ac@arrl.net>
Date: Wed, 1 Dec 2010 10:30:07 -0500
List-post: <towertalk@contesting.com">mailto:towertalk@contesting.com>
Steve,

Perhaps it would be more accurate to state "I-squared-R loss attributed to 
SWR," rather than "Additional loss due to SWR?"  The caveat that addresses the 
effect seen on short lines is still needed.  

Not sure if this link was provided earlier, but VK1OD presents an excellent 
analysis of a relatively recent QST article to illustrate his point.  

http://vk1od.net/transmissionline/VSWR/aldv.htm

I think the QST article still offers an excellent explanation for beginners 
even when he refers to the "additional loss due to SWR" graph without further 
clarification.  By the way, there's a gross error on one of the graphs!  Looks 
like the artist didn't do a good job of lining up the X and Y axis and it 
wasn't caught during proofs.  

Still, anyone who can even get to that level of understanding without the 
caveat is more knowledge on the subject than the vast majority ops.  For 
example, eHam is still publishing articles from authors who proclaim that only 
single-band, resonant antennas can achieve any semblance of high efficiency and 
that ATUs do nothing more than "make our transmitters happy."

Paul, W9AC
  ----- Original Message ----- 
  From: Steve Hunt 
  To: Paul Christensen 
  Cc: towertalk@contesting.com 
  Sent: Wednesday, December 01, 2010 9:52 AM
  Subject: Re: [TowerTalk] tuners and power rating


  Paul,

  I'm not sure it helps to distinguish between "SWR losses" and "I-squared-R 
losses" - at HF, all the losses are predominantly "I-squared-R losses".

  It may help to picture qualitatively the current profile over a short length 
of feedline at the load end - in all cases delivering the same power to the 
load:

  * If we have a matched load, the current is constant along the line and the 
loss-per-unit-length will therefore also constant along the line. Cumulative 
losses increase linearly with length.

  * If we have a load with a moderately high resistive component the current at 
the load will be lower, and therefore loss-per-unit-length will be lower. But 
slightly back from the load the current will have increased due to the standing 
wave pattern, and therefore the loss-per-unit-length will be higher; eventually 
it exceeds the matched case loss-per-unit-length, and even further back the 
cumulative losses exceed those of the matched case.

  * If we now have a load with a *very* high resistive component, the current 
at the load will be very low and the loss-per-unit-length will be even lower 
than in the previous case. However, moving back from the load, the rate of 
change of current with distance is higher because of the increased ISWR, and it 
may be that we reach the "break even" point sooner, despite the 
loss-per-unit-length adjacent to the load being lower.

  Incidentally, there will be a load value which maximises the distance from 
the load of the "break even" point.

  So, all the losses are "I-squared-R losses" - it's just that the current 
profile (and therefore the cumulative loss profile) changes if the ISWR is not 
unity.

  73,
  Steve G3TXQ





  On 01/12/2010 13:44, Paul Christensen wrote: 
    The additional loss attributed to a mismatch is still relevant once SWR 

becomes part of the loss.  In the 10 ft. examples we've been using, that 
occurs when the line get to roughly 40 degrees in length.  When we approach 
1/4 wave, loss due to SWR becomes equal in loss to I-squared-R loss.  I 
think that's was one of Steve's points in that the additional loss 
attributed to SWR needs some clarification in the footnotes o be completely 
accurate.

Paul, W9AC


_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk


_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

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