[Elecraft] K3 - SWR Indication

Tue May 5 22:08:17 EDT 2009

Several interesting topics have arisen in this discussion. Here are a couple
comments...

1) Most wattmeters use a diode peak detector for each of the coupler ports
(FWD, REF). The forward voltage drop of the diodes becomes an increasing
source of error as power is lowered. The means than for fairly low SWRs,
there is more error in the REF voltage sample, which is lower to begin with.
Squaring the voltage to calculate power exaggerates the error. Very few
commercial meters try to compensate for the diode drop. The LP-100A does,
using a circuit similar to the one in the Tandem Match design, as well as
using a dual Schottky diode on a common substrate to minimize temp
variations between the detector diode and the correction diode. It is
possible to get a fairly good correction for static voltage drop errors, but
errors due to the dynamic inverse impedance of the diode, which is frequency
dependent, can't be easily compensated for. Still, it's much better than
nothing.

2) Most wattmeters use a simple coupler where samples of line current and
voltage across the load are combined to obtain FWD and REF voltage samples.
There are some inherent phase errors which are relative to frequency in this
simple design, and magnitude errors related to parasitic coupling which
limit the directivity of the coupler. Directivity determines the ultimate
limit of SWR measurement error. As mentioned by K8ZOA, 30dB is an excellent
broadband directivity number, but rarely achieved with high power couplers
from 2-54 MHz. The LP-100A feeds the raw current and voltage samples to a
gain/phase detector chip for determination of SWR, instead of combining them
in the coupler. No diodes are involved, and no FWD or REF voltage samples.
The meter has a frequency counter, and any phase or gain errors in the
coupler samples can be calibrated out, indexed to frequency. This allows the
meter to achieve >40dB directivity from 2-54 MHz.

3) Power circuitry and calculations are completely independent of SWR
circuitry and calculations in the LP-100A. As long as there is about 1-2W of
power, the gain/phase detector will provide the same result at any power
level, with slightly reduced accuracy down to 50mW. The LP-100A uses the
differential magnitude and phase samples to calculate the complex reflection
coefficient, from which all other impedance and SWR numbers are derived. The
meter can display REF power, but it is calculated from FWD power and
reflection coefficient, not measured directly. 

73,
Larry N8LP

Jack Smith-6 wrote:
> 
> A wattmeter built around a directional coupler always has to deal with 
> finite coupler directivity. Making the problem more difficult is that we 
> expect a wattmeter to be accurate over a rather wide frequency range, 
> 1.8  to 30 or even 50 MHz. This places an even greater burden upon the 
> directional coupler.
> 
> If the directional coupler has 30 dB directivity--a very good number to 
> be maintained over a wide frequency range--then 1 KW forward power into 
> a perfect load will show 1 watt reflected power, corresponding to an SWR 
> of 1.065:1 instead of the expected 1.0000... for the theoretically 
> perfect load.
> 
> It is possible to measure the phase and amplitude of  the coupled signal 
> to "calibrate out" coupler imperfections. This is what is done with a 
> vector network analyzer when the standard "open/short/load" calibration 
> is applied. The VNA measures the phase and amplitude of the coupled 
> signal when the through port is operated into an open circuit, a short 
> circuit and a known value (resistance and stray L & C known) termination 
> for each test frequency. The VNA then computes and applies an 
> appropriate correction factor to correct for coupler errors. O/S/L 
> calibration has been supplemented by more advanced techniques in newer 
> VNAs. (There's a very good Application Note AN 1287-3 from Agilent on 
> this subject available at 
> http://cp.literature.agilent.com/litweb/pdf/5965-7709E.pdf. Bird 
> Electric has a less technical Application Note on the effect of coupler 
> directivity on SWR at 
> http://www.bird-technologies.com/techapps/app_notes/StraightTalkAboutDirectivity.pdf)
> 
> A wattmeter using diodes to measure RF voltage used with a directional 
> coupler cannot apply sophisticated error correction to compensate for 
> finite coupler directivity. At most, one can tweak a balance pot or 
> trimmer cap to null the reflected signal at a single frequency and power 
> level. Further complications result from the forward and reverse diode 
> detectors being operated at different points on their sensitivity curve, 
> etc.
> 
> Hence, it is far from surprising that different wattmeters will show 
> different SWR under ostensibly identical test conditions.
> 
> Larry's LP-100 wattmeter operates with a different methodology and I'll 
> leave it to him to explain the differences and how coupler directivity 
> is considered.
> 
> 
> Jack K8ZOA
> www.cliftonlaboratories.com
> 
> 
> 
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
> 
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> 
> 

-- 
View this message in context: http://n2.nabble.com/K3---SWR-Indication-tp2784793p2808130.html
Sent from the Elecraft mailing list archive at Nabble.com.