How to measure (Z)impedance@X Frequency and reactance?

[rwgast_lowlevellogicdesin]

First of I am working in the L-Band mostly, and soon probably X band too. So right now I can measure return loss using a directional coupler and an RF signal source or noise source, and an SDR or RF power head to get the RL reading. Now I see that a bunch of cheap antenna analyzer can at least display impedance of the antenna at the input frequency. Im not sure if I am right but in order to get a Z reading on top of return loss you would also need phase, so what is the easiest way to measure phase and what is the phase relative too? I would also like to measure reactance that way I can tell when the inductance and capacitance cancel and I am at the resonant point.

I am not very big on transmitting, but I would like to build the most efficient rx antennas and dish feeds possible. Im interested in space stuff... anyways logically to me the best way to make an RX antenna is build it then tune it till it is purely resistive at the frequency of choice and then match it to the feed line using an impeadence transformer/balun/unun whatever you want to call it. If one did it that way that would mean the antenna is almost a perfect match to free space at the desired frequency and then the matching transformer takes care of matching to the 50 ohm line and eliminates any losses from SWR. Or am I totally wrong here? Either way I would like to know how to measure the impeadence of things.

[rwgast_lowlevellogicdesin]

Ok so after reading for hours i think I understand how to make an impedance measurement at rf frequency. I think there are two viable methods, please let me know if I am wrong.

First you can find the magnitude of the signal, and the voltage vs current phase angle, and go from there, Now there is a technique called the three meter method that can be used but im not sure it will work to well with discrete components at frequencies over HF. So maybe someone knows if there is a chip out there that can measure the phase angle of current vd voltage lag? I found the AD8302, but I don't think it will work in this case, it seems like it only measures the phase angle between two different signals? If it is only useful for finding phase between signals, maybe there is a way to sample voltage and current separately and fed each in to one of the AD8302's inputs? This method would require some sort of receiver i.e specAnn, SDR, LogAmp+ADC and a signal generator. I also know you can find the angle with dual directional couplers like a VNA does but that sort of set up would need at the least coherent SDRs to sample  I think.

Secondly and what I think may be the easier path is to just measure R,L, and C then find the inductive and capactive reactances @ desired frequency, using standard equations for doing so and then it would be as simple as Z=R+|Xc-Xl|. Obviously measuring very low inductance and capacitance of antennas and transmission lines is not trivial but these people are doing it, so it's not impossible and you could even just by the pcb only and use its serial interface to talk to another micro that handles everything else.

Any ideas here guys, seems like there a multiple ways to approach making these measurements with out building a full blown VNA.

[xaxaxa]

Ok so after reading for hours i think I understand how to make an impedance measurement at rf frequency. I think there are two viable methods, please let me know if I am wrong.

First you can find the magnitude of the signal, and the voltage vs current phase angle, and go from there, Now there is a technique called the three meter method that can be used but im not sure it will work to well with discrete components at frequencies over HF. So maybe someone knows if there is a chip out there that can measure the phase angle of current vd voltage lag? I found the AD8302, but I don't think it will work in this case, it seems like it only measures the phase angle between two different signals? If it is only useful for finding phase between signals, maybe there is a way to sample voltage and current separately and fed each in to one of the AD8302's inputs? This method would require some sort of receiver i.e specAnn, SDR, LogAmp+ADC and a signal generator. I also know you can find the angle with dual directional couplers like a VNA does but that sort of set up would need at the least coherent SDRs to sample  I think.

Secondly and what I think may be the easier path is to just measure R,L, and C then find the inductive and capactive reactances @ desired frequency, using standard equations for doing so and then it would be as simple as Z=R+|Xc-Xl|. Obviously measuring very low inductance and capacitance of antennas and transmission lines is not trivial but these people are doing it, so it's not impossible and you could even just by the pcb only and use its serial interface to talk to another micro that handles everything else.

Any ideas here guys, seems like there a multiple ways to approach making these measurements with out building a full blown VNA.

At rf it's difficult to measure voltages and currents directly because it's impossible to make very high or very low impedance devices; in other words no rf voltmeters or ammeters. Instead you always deal with waves.

The easiest way to do what you proposed is using a bidirectional coupler (directional coupler with both forward coupled and reverse coupled ports); then you simply measure the magnitude ratio and phase difference between forward and reverse, and this will give you the magnitude/phase of the reflection coefficient. The ad8302 will do that if you feed the coupled signals into INPA and INPB. Then you can use the formulas from https://en.wikipedia.org/wiki/Reflection_coefficient and solve for the impedance.

[rfeecs]

Here is a low frequency (100MHz) VNA made with the AD8302:
http://www.tapr.org/~n5eg/index_files/final_QEX_mcdermott.pdf

It points out some of the problems with the AD8302 for phase measurement.  One main problem is the phase output is symmetrical about 0 degrees, so you need a method to resolve the sign of the phase.  The other problem is the phase output is nonlinear so you need to compensate for that somehow.

[radiogeek381]

Welcome to the ranks of the microwave-philes! 

If you're just getting started, check out some of the ARRL publications on UHF/VHF/Microwave experiments and construction.  RSGB also has some very good material.  Even the older stuff is informative.

For what it is worth, I've build a lot of 10GHz equipment with not much more than a power meter.  But I'll say that a signal generator (not a synthesizer, just a generator) was rather handy.  Even at that, some of the weak signal sources (an oscillator at 1152 MHz run into an amplifier that is well beyond its linear region) work just fine to set up receivers.  Depending on where you live, there may be beacons on 10GHz and lower bands.

For the 10GHz antennas (mostly offset-fed dishes) I've used tried and true feed designs.  Most of the tuning has little to do with the impedance match, and much to do with the position of the feed relative to the dish surface. (Yes, we learned in physics that the feed center has to be at the focus. Alas, they never told you precisely where the feed center would be.)  For alignment, and general RX performance measurement, you need two things:  a big honking source of broadband RF energy like a thermonuclear event, and a sensitive power sensor.  For the former, I use the Sun.  It is available for much of the day, though not every day.  For the latter, I use the panadapter display on my IF receiver (it is a very fancy SDR, but you could use one of the inexpensive dongles.)

For lower band antennas, a power meter and a directional coupler -- two if you're lucky enough to find a good deal -- is really all you need.  Yes, you don't get the phase with a power meter, but you get the magnitude.  When you're standing next to a 15 foot antenna tweaking a loop and finding the right tongue angle, one dimension is about all you can optimize over.

If you really want to measure magnitude and phase, the least expensive technique I know is with a slotted line, a probe, a signal generator that can modulate a 1kHz tone on the carrier, and an ancient HP 415 SWR meter.  There ought to be something on the web about how one of these works.  I've done that once or twice, but found little value in it.

Far and away the biggest bang for the buck in my shop has been my HP432A power meter.  I'm partial to the old fashioned kind with a needle.  When you're tweaking, you look for the needle to move in the good direction.  It is much more natural to react to a needle moving than to recognize that 2.38 is greater than 1.38.  (And the needle puts in perspective just how insignificant 1dB really is, in the grand scheme of things.)  Be aware though that you can get the meter itself for cheap on Ebay or other places.  It is the measurement head, and -- believe it or not -- the cable between the measurement head and the meter that are rare and often expensive.  Don't buy a meter body without the cable and head.  My goto measurement head is an HP 478 that goes from about DC (100 MHz) to just shy of daylight (10GHz). These are fragile -- I bought one of mine from a "trusted source." The second was a wedding present from the same "trusted source" as we had done a fair (ahem) bit of business by then.  My wife was touched. The meter tops out at about 10dBm input, so you need a stack of attenuators and an occasional directional coupler to make it useful. 

By mass, my shop may be dominated by adapters to go from one RF connector family to another.  All the good stuff that I connect up is either SMA or N.  But every once in a while there's some joker with a BNC or even a UHF connector.  (Queue the chorus of replies of the flavor "yeah, adapters, I've got a BillumQuick 5832 Microwave Wobulator with APC-7 connections on one side and a three pronged Russian triax connector on the other.  Finally found the adapter at Pasternack for a little less than the cost of a Ferrari.")

This setup won't rival the average dumpster at NIST.  And, notice, that I haven't mentioned a spectrum analyzer yet. 

I have a spectrum analyzer.  It is amazing how little I use it.  It gets powered up every now and then, but it is creaky, has grown spurs, and does everything but leak oil.  I do, however, have some bodacious broad-band RX, signal generators that, between them, cover 10MHz to 13 GHz, a power meter, and the ability to make a filter or two when required. 

Enjoy building stuff.  Don't let the absence of a piece of fancy test gear get in the way. 

keep your antenna on the ice.