EEVblog Electronics Community Forum
Electronics => RF, Microwave, Ham Radio => Topic started by: fonograph on July 27, 2018, 11:06:53 pm
-
Can VNA measure inductance,capacitance and resistance at given frequency as separate things or does it only measure impedance frequency curve? Do I need high frequency LCR meter if I have VNA?
-
the VNA is not going to be accurate on measuring things that deviate alot from 50 ohms.
https://www.keysight.com/upload/cmc_upload/All/ChallengesandsolutionsforImpedance.pdf (https://www.keysight.com/upload/cmc_upload/All/ChallengesandsolutionsforImpedance.pdf)
I am still learning about VNA but I kind of suspect it should be left for things like measuring filter performance.
-
the VNA is not going to be accurate on measuring things that deviate alot from 50 ohms.
Really? That sucks,I thought VNA doesnt care about source impedance
-
Can VNA measure inductance,capacitance and resistance at given frequency as separate things or does it only measure impedance frequency curve? Do I need high frequency LCR meter if I have VNA?
You can measure components using a VNA, and it will give you a direct read out of resistance and capacitance or inductance.
But its much easier to use an LCR meter.
You need to calibrate the VNA and also account for the effect of the test fixture you are using - or test at a low frequency where this doesn't matter.
Its beyond my expertise to answer this question properly though as I don't fully understand the limitations - the limitations are what I think CopperCone was referring to.
Here is a link I quickly found on google where someone does some component measurements on a VNA.
http://www.w0qe.com/Measuring_High_Z_with_VNA.html (http://www.w0qe.com/Measuring_High_Z_with_VNA.html)
Note that the HP 8753B screenshots don't show the values, but it does show them in the Smith Chart view.
I can't quite remember if the HP shows total impedance in ohms or just the resistive component of the impedance in ohms. It definately shows capacitance/inductance and the reactance though.
-
Can VNA measure inductance,capacitance and resistance at given frequency as separate things or does it only measure impedance frequency curve? Do I need high frequency LCR meter if I have VNA?
Depends. I have and use both. I use the VNA for RF work. Hard to see what's going on above a GHz with an LCR meter.
My VNA is a very old 8754A. I have home made cables. My cal kit is home made. The VNA would have been operated with a grease pen. I pull the data into Labview. So I wrote the software to perform the SOLT and make these measurements.
I have to believe that my setup would be the bottom of the barrel. If you watch the start of this video, you can see the cal kit along with the system in operation measuring a some parts. More of a historical video than anything.
https://youtu.be/GJd8niwbic8?t=51
-
With capacitors a realistic test might be to build a oscillator and look at the spectrum/Q on an oscilloscope/SA/Frequency meter, for practical uses, if you have qualms about VNA accuracy
-
This is one of the things that VNAs can be (and often are) used for. At low frequencies, where the impedance will be far from 50 ohms, a VNA is not really the best tool for the job, and you should look at LCR or CR meters. However, past a gigahertz a VNA is pretty much the only option we have. You can't, as CopperCone suggests, build a oscillator with them because in order to get any accurate measurement of the Q-factor etc, you need to have a /extremly/ well characterized oscillator, otherwise you are measuring the Q of your system, and good luck extracting what part of that Q is your capacitor/inductor/resonator.
A VNA is a extremely precise tool, with the right calibration and conditions they can get to sub-0.005 dB error. I think it is pretty much impossible to beat this with any other instrument at anything past a few GHz. However, once you start pushing things far away from 50 ohm their accuracy starts to suffer.
-
And what about this Omicron Bode 100 VNA,it has 1 Hz to 50 MHz frequency range,does use 50 ohm only too?
-
the document I linked has a graph that shows the 10% accuracy of the 50 ohm VNA vs frequency and impedance btw
-
I've been using decent lab VNAs (with decent cal kits) for over 30 years and they can be very accurate for simple components like a chip resistor anywhere in the 5R to 2000R range up to several hundred MHz. They are also very good at measuring sensible values of inductance and capacitance.
However, it's easy to hit the limitations of a VNA if you expect to make measurements of component Q or if you want to measure the ESR of an inductor or capacitor over a wide frequency range. The results can quickly become very inaccurate compared to a decent impedance analyser.
However, see below for a measurement of a 1000.07R chip resistor (measured at DC with a 4 wire DMM) and a 3.3uH coilcraft inductor. This was measured three times so there are three traces for each measurement. Once with a VNA at work using an 85033E cal kit and the other two measurements are here at home with my VNA using a decent Ecal module and also my own homebrew SMA cal kit. So there is a total of six traces per graph.
You can see that the results agree/overlay very closely but you can also see how the data becomes very noisy and uncertain above about 80MHz for the real part of the 3.3uH inductor. This is because the part is near self resonance and the VNA will struggle here.
You can also give a VNA a hard time with something as simple as a small chip resistor value of a few ohms in series with a small capacitor value of a few pF. It will struggle to measure this down below about 50MHz and will produce big errors here if you try and measure this with a classic SOL (reflection) calibration. So you have to be aware of when a VNA can give very good results and when it can give really poor results.
A VNA is really meant to be used as a device to obtain an n-port model of an n-port device rather than as a replacement for an impedance analyser but there are times when it can do a decent job of measuring impedance as long as the operator knows how to get the best from the VNA and as long as they use a decent cal kit and a properly characterised test fixture.
Note that the resistance measurement of the 1K 0805 chip resistor is plotted as the parallel equivalent circuit Rp and Cp rather than as the classic series Rs and Cs for an impedance measurement.
Looking at Rp and Cp shows this resistor to be a fairly pure and accurate 1K ohm resistance in parallel with about 0.08pF package capacitance although a small amount of that 0.08pF is stray capacitance in my test fixture.
-
See below for the 3.3uH up to 1GHz. This is again for all three cal kits and the 85033E based measurement was done at work with a different VNA but they all agree quite well. This was a fairly casual comparison and I didn't bother with a torque spanner for the Ecal tests here at home as I try to minimise any connector wear on it. So I only torque the Ecal connections correctly for critical (important) stuff. But for this kind of comparison I don't think it matters much.
-
I am super thankfull G0HZU but I have no idea what any of the information you posted means becose I am too inexperienced. :-[
Could you please try to translate your microwave wizard language to n00b langauge? Correct me if I am wrong,what I am getting from this is that that I can measure L C R with VNA but it must be high enough frequency and it cant be too small or too large value,is that right?
I am planning this : use DMM for DC LCR measurements,use Bode 100 VNA for 1 Hz to 50 MHz LCR measurements and 2 port USB VNA like Keysight Streamline or Tektronix TTR500 for the 50 MHz to 6 GHz LCR measurements.
-
What I meant to say was that the accuracy of a VNA when used as an impedance measuring device will vary depending on the range of impedances you want to measure and also the range of frequencies you want to measure them over.
It's best to get some experience playing with a VNA to see what it can and can't do before spending loads of money on a VNA.
I am planning this : use DMM for DC LCR measurements,use Bode 100 VNA for 1 Hz to 50 MHz LCR measurements and 2 port USB VNA like Keysight Streamline or Tektronix TTR500 for the 50 MHz to 6 GHz LCR measurements.
However, if you bought all the new gear above you would probably have spent more money than I have ever spent on all my test gear (at home) across several decades :)
Maybe start by having a look at the Analog Discovery 2 as it has the ability to measure impedance from <<1Hz to several MHz. It won't be as accurate as a professional $$$ impedance meter but it will be a good place to learn about the various ways to set up an impedance measuring jig and I think the AD2 only costs a few hundred dollars. You do have to make a test board with a suite of sense resistors for the impedance measurement but it should only take an hour or so to make it and will cost next to nothing to make the board.
You can also use it as a basic low frequency scope, arbitrary waveform generator and spectrum analyser (to a few MHz) so it is a good learning tool for beginners :)
-
Just a quick note to those using (or contemplating using) the Impedance Analyzer tool of the Analog Discovery device. Digilent recently came out with an inexpensive add-on board (< $25.00) to make its use more convenient.
They state that "With the Impedance Analyzer add-on module, you can simply select the “Adapter” option in the resistor configuration drop down in the Impedance Analyzer Instrument, and WaveForms will automatically select the appropriate reference resistor."
https://blog.digilentinc.com/add-impedance-analysis-to-the-analog-discovery/
This add-on module works with the Analog Discovery 2 and the original Analog Discovery.
-
What I meant to say was that the accuracy of a VNA when used as an impedance measuring device will vary depending on the range of impedances you want to measure and also the range of frequencies you want to measure them over.
Yeah yeah it will vary,but how will it vary thats what I want to know.How does DUT input and output impedance affect it? Why does VNA have trouble at lower frequencies compared to literaly all other test equipment that have easier time working the lower the frequency is?
You didnt answer my question about Bode 100 + USB VNA combo.You wrote "It will struggle to measure this down below about 50MHz".That Bode 100 is made for 1 Hz to 50 MHz range,exactly the frequency span you wrote that normal VNA will suck badly at,so I ask again,if I get the Bode 100 + Tek or Keysight USB VNA with 100/300 KHz - 6 GHz range,can I accurately measure LCR from 1 Hz to 6 GHz?
-
Why does VNA have trouble at lower frequencies compared to literaly all other test equipment that have easier time working the lower the frequency is?
It really depends on the impedance of the device being measured and this applies at any frequency, not just low frequencies. I only gave an example at a low frequency and here it is again in more detail:
If you tried to measure 2.7R in series with 10pF at 20MHz with a VNA in the classic reflection mode the reflection coefficient would be very close to 1. It will be about 0.9996 with an angle of -7.2 degrees. This is (2.7, -j796) so no wonder the VNA struggles here because there isn't much of a vector to measure. So the VNA will struggle to get the 2.7R resistance correct once down at 20MHz or so. A really good lab VNA with a good cal kit would still be able to indicate something vaguely correct but many of the cheaper USB VNAs would completely fail here. So a different impedance measuring technique would be needed instead of a reflection measurement. However, with other impedances the reflection method can be really accurate and this applies from 1MHz up into the GHz region.
so I ask again,if I get the Bode 100 + Tek or Keysight USB VNA with 100/300 KHz - 6 GHz range,can I accurately measure LCR from 1 Hz to 6 GHz?
I don't know anything about the Bode 100 but the Tek VNA doesn't look to be a precise instrument when compared to the specs of a decent lab VNA. The new Keysight USB VNA will probably be better but you still won't always be able to measure impedance accurately if you are expecting to always measure to something like +/- 2%.
Can I ask: why do you (as a noob) have a need to measure impedance accurately across this vast frequency range?
-
@fonograph
You Seem to be a nice person, so i think you genuinely benefit from learning how VNAs work and their applications. The best source would be early HP/Agilent application notes. The problem is too many of them but you can identify the level you can consume by glancing over the content, then give it a deep read if comfortable.
Just one of the problems with using a VNA at high frequencies for measuring component values and whatnot is you have to eliminate (calibrate out) the systematic errors and external connections up to your DUT terminals. So you need a calibration kit to just operate a VNA (they cost big buck) and test fixtures which need to be characterized (measured) and ability to separate ( de-embedd) the fixture from the DUT measurements. This is not a trivial task and not all VNAs have de-embedding capabilities, which still can be conducted externally using custom math but you need to download the measured data to the PC , which may be a hassle in some cases.
All in all, VNAs use dfferent principles for taking measurements than conventional LCR meters. You need to understand the operating principles and limitations for what you want to do.
-
G0HZU I dont need to measure LCR in such vast frequency span right now,but,I plan to do microwave stuff in future so I am preemptively learning about it now.
-
fonograph i wanna work for you, since you don't try to save every every penny on test equipment.
do beware the costs of all the microwave components. If you want to put together systems out of waveguide or coaxial stuff you will get hit hard. I guess your interest is PCB design though?
I kinda find infrasonic/audio and DC to be more interesting the microwave in alot of ways.
I don't think a impedance analyzer is that useful for making microwave systems for a hobbyist unless you are pushing the limits of frequency, otherwise you can get components that are rated for higher frequencies and build a 50 ohm system for a VNA. Money better spent on antennas, materials, machine tools, filters, couplers, attenuators, generators, spectrum analyzers, as someone that only recently got into the microwave/RF range I would be much happier to spend money on equipment other then a impedance analyzer. My interest quickly wanes when it comes down to categorizing a single component rather then building a functional system with an antenna, vwsr measurement circuits, making it nice mechanically etc.
Even with low frequencies I never really needed to use the LCR meter that much.. I always ended up designing my stuff with a DAC or whatever, so the frequency is not super important since it can be modified in case a filter I built has a difference pass band or whatever.. I just test it to make sure digikey sent me parts that are somewhat OK. Air coils actually wind really nicely and seem to agree well with simulations and capacitors are very decent.. I kind of think like it would be more useful if I wanted to categorize sub-par parts to see if they work in a design but its not really that important to me because the hobby/design is so time consuming that the cost of parts is rather small compared to the time cost you put into it. I found that with filters you pretty much get good parts to begin with because they are touchy and for stuff like rough filtering power rails and stuff, you just kinda see if it achieves the desired goal. For some reason I never got too concerned about what the power rail was doing exactly when the op-amp has like 100dB PSRR... usually just use a multimeter or scope with an amplifier to verify that it does a decent job.
Most useful I think is if you wind your own transformers or coils on magnetic material, or reuse magnetic material... I never had good luck making estimates from theory in this department. I would be lost without my DE-5000 here. Especially considering you have all those winding tricks, litz wire types, etc.. enhanced permeability magnetics have been a pain in my ass. They seem downright unruly compared to capacitors, resistors and semiconductors, and datasheets never seem to tell you much about them, but they are fascinating no doubt
-
fonograph i wanna work for you, since you don't try to save every every penny on test equipment.
O RLY? Go ahead,come up with equipment combination that can measure impedance from 1 Hz to 6 GHz for less than 17k which is cost of Bode 100 + TTR 500.
My requirements : 1. It cant be used,new only. 2. It must be from high quality brand,no chinese bug ridden toy.
-
fonograph i wanna work for you, since you don't try to save every every penny on test equipment.
O RLY? Go ahead,come up with equipment combination that can measure impedance from 1 Hz to 6 GHz for less than 17k which is cost of Bode 100 + TTR 500.
My requirements : 1. It cant be used,new only. 2. It must be from high quality brand,no chinese bug ridden toy.
that was a compliment bro :-*
-
fonograph i wanna work for you, since you don't try to save every every penny on test equipment.
O RLY? Go ahead,come up with equipment combination that can measure impedance from 1 Hz to 6 GHz for less than 17k which is cost of Bode 100 + TTR 500.
My requirements : 1. It cant be used,new only. 2. It must be from high quality brand,no chinese bug ridden toy.
that was a compliment bro :-*
Ohh... Ok.I.. I thought.. that you.. umm.. nevermind. ( poker face ( fights back tears( hugs you ( nohomo ) <3
-
What you really seem to need is an impedance analyzer / VNA combo.
Such a machine exists, from 5Hz (no typing error) to 3GHz. Its the Keysight E5061B-3L5.
Using some refined measurement techniques it can measure using low (50Ohm) impedances like a normal VNA,
or high impedances for the inputs (1MOhm) for low frequency or control loop measurements.
Impedances down to milliOhms or several 100kOhms can be accurately measured (by shunt-thru or series measurement adapters)
For device measurements test fixtures with standardizes and calibrated properties are available.
A lot of application notes explain how its done (I learned a lot from those).
There is one minor issue, however: The list price including software is about 35k€.
-
Can you use the Keysight USB VNA as spectrum analyser? Can you generaly use VNA as spectrum analyser? VNA to me seems like fancy spectrum analyser with signal generator.
-
Can you generaly use VNA as spectrum analyser? VNA to me seems like fancy spectrum analyser with signal generator.
Not sure about other instruments but the new Siglent SVA1015X is a dual SA/VNA instrument....just at not very high BW. :(
-
I want to make one thing clear,these microwave frequency range VNAs are useless for measuring anything with impedance other than standard 50 ohm.For example,if I had 1 ohm or 400 ohm DUT operating in the 1 to 6 GHz range,then I couldnt measure it with your typical VNA,right?
-
Hi,
it is indeed true that measuring impedances that are either very small or very large present an accuracy problems for a VNA.
Several strategies exist:
- If you know where to look, you could employ specialized measurement techniques like shunt-thru or series.
With these, you could measure PDN impedances in the milliohm range or the impedance of Quartz crystals.
Keysight has very good app notes on this topic, and a specialized machine like the E5061B-3L5.
- Buy a better VNA. The better ones have a broader range, more accuracy and less tolerances.
In extreme cases, impedances of 0.1Ohm or some kOhm can be measured.
- Do not forget to buy extremely good, phase stable cables and a pro cal kit.
Intelligence can beat blunt money in a lot of measurement tasks.