[Elecraft] K3 SWR Accuracy - reprise

[Don Wilhelm]

I believe Steve is onto something critical to this discussion - here is 
my 'take' on the differences --  Remember that we *are* talking about 
the resultant SWR indications on a mis-matched line.

The way most "wattmeters" indicate SWR is to detect the forward power 
and the reflected power - then an SWR is computed from those values of 
associated detector output voltages.  The result is an SWR indication 
based on the absolute values of those magnitudes.  No consideration is 
provided for the phase angles (the actual forward and reflected values 
are complex numbers).  The forward power is indicated accurately (it is 
proportional to the square of the forward voltage minus the square of 
the reflected voltage), but the computation of actual SWR is more involved.

When the impedances are close to the design point (balance point) of the 
meter, the error is small, but the error grows as the actual impedance 
departs from that design impedance.

Those meters that properly detect the phase as well as the magnitude of 
the forward and reflected powers can indicate that the SWR does not 
change as the meter position is moved along the line - but most do not 
have phase detection capability nor complex number computation 
capability, so for those meters, the SWR indicated will change with the 
meter position along the feedline.

In other words, use a good VNA and you should see a constant SWR along 
the line, but common wattmeters are not VNAs, so some error in SWR 
indication is to be expected when the impedance is removed from the 
design point.

Even the well-respected Tandem Match computes the SWR as Vf+Vr/Vf-Vr, 
which is the correct formula, but the detector reports only the 
magnitudes of Vf and Vr and does not consider the phase angle, so it is 
not entirely correct either - it will be entirely correct when the SWR = 
1.0.

73,
Don W3FPR

Steve Ellington wrote:
> Nope:
> The 1/4 wave line transforms the high impedance to a low one and the SWR 
> meter reads low. It's called a transmission line transformer and is very 
> common. It's the reason everyone is having trouble understanding why SWR 
> meters read differently. The ONLY way to compare them is to swap them with 
> each other. Putting them in series fouls up the readings for both meters.
> Steve
> N4LQ
> N4LQ at carolina.rr.com
> ----- Original Message ----- 
> From: "Kok Chen" <chen at mac.com>
> To: "Elecraft Reflector" <elecraft at mailman.qth.net>
> Cc: "Steve Ellington" <n4lq at carolina.rr.com>
> Sent: Wednesday, November 04, 2009 7:55 PM
> Subject: Re: [Elecraft] K3 SWR Accuracy - reprise
>
>
>   
>> On Nov 4, 2009, at 4:41 PM, Steve Ellington wrote:
>>
>>     
>>> 4. Example: A full wave dipole center fed with 50 ohm coax. SWR reads
>>> infinite at the antenna but with 1/4 wavelenth of coax, SWR reads low!
>>>       
>> Nope -- the *impedance* at the end of a 1/4 wave transmission line
>> when it is looking at a very large impedance, is close to zero,
>> therefore the SWR remains close to infinite.  The SWR definitely won't
>> read low unless there is something wrong with the instrument.
>>
>> 73
>> Chen, W7AY
>>
>>     
>