[TowerTalk] SWR is what SWR meter measures

[Ian White GM3SEK]

Tom Rauch wrote:
>Hi Ian,
>
>Nice to see you again!
>
Hi Tom! Nice to be back to this list.

We're still commuting between two homes, in G and GM, and work levels 
and Internet access are causing some "QSB" on the newsgroups and mailing 
lists...

>> 22. Don't even THINK about "net" power being the
>difference between
>> "forward power" and "reflected power".  You don't need to
>go there... so
>> don't.
>
>It works perfectly, or as perfectly as I can measure, on my
>meters.
>
>> The problem with using those real-life measurements to
>calculate the
>> ratio of "forward power to reflected power" is that it
>suddenly requires
>> us to know things about the transmitter! We're now saying
>that we need
>> to know the type of transmitter, how it is tuned, driven
>and loaded, etc
>> etc. But that flatly contradicts what we said in the
>previous paragraph
>> - namely that we can know everything about the
>transmission line and
>> antenna, without needing to know *anything* about the
>transmitter.
>
>No we don't. It works with any source Ian.
>
>When you use a directional coupler power meter to measure
>forward power and when the load is not matched to the meter
>the forward and reflected readings increase. When reflected
>power is subtracted from forward power we are left with the
>power delivered to the feedline.
>
>It works just fine in every case I have seen. It never is
>affected by the transmitter.
>

I agree with your bottom line, Tom... but not completely with your 
reasoning.

The problem is that the so-called "directional wattmeters" don't 
actually measure forward or reflected power. They sample RF voltage and 
current on the line, run the current through a resistor to get a 
voltage, and then  add or subtract those voltages in vector form. 
Finally they detect the resultant RF signal, and the DC current pushes a 
meter up-scale. Nothing at all in this process is responding to RF power 
as such, only to  RF voltages and currents.

What a "directional ??-meter" actually measures is reflection 
coefficient. Assuming perfect rectifiers and a linear DC voltmeter 
movement, the ratio Vf/Vr *is* the reflection coefficient. If you always 
normalize Vf to full-scale on the meter, then you can switch to reverse 
and measure reflection coefficient on a simple linear scale.

Reflection coefficient is a real measurement. Everything else that you 
see on  the meter scale comes from either math, calibration or both.

For example, the familiar non-linear SWR scale comes from the 
mathematical relationship between reflection coefficient and SWR. But it 
didn't actually measure SWR.

The power calibration on the meter scale comes from terminating the RF 
output port of the instrument in a load equal to the Z0 for which the 
meter was designed, connecting some other instrument that genuinely does 
measure RF power (eg a thermal wattmeter) as a reference, and then 
calibrating the forward reading in terms of watts. So the meter is only 
being *calibrated* in terms of "forward power" - it doesn't actually 
measure power at all.

If you now connect a mismatched load, there will be some indication of 
"reflected power". Sure enough, the difference between  "forward" and 
"reflected" readings will equal the actual power delivered to the load. 
If you choose to think in terms of forward and reflected power, those 
results follow from the math. But I'm just not sure if those results are 
anything anyone needs to know.

The same reasoning applies to the entire range of directional bridges, 
couplers and circulators: under the hood, all their physical internal 
workings involve vector RF voltages and currents (or E and H fields if 
we're in waveguide or YIG).  If anyone actually has a device that really 
does separate a wave of forward power from a wave of reflected power, 
and whose physical principle can be explained with NO reference to 
voltage and/or current; then please show.

But if (as I strongly suspect) no such device exists, then it really 
isn't helpful to talk about "directional wattmeters" - because nobody 
actually has one.

On the other hand, the concept of "forward and reflected power" does 
have a proven down-side: arguing about it has swallowed large chunks of 
the lives of a lot of highly intelligent people, to little or no 
advantage.

Not wanting to add to that problem, I'd better stop right here.


-- 
73 from Ian GM3SEK         'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek