|
Just wondering about how to increase the highest value of impedance measurements.
Since it tops out at about 600 ohms in HF how would I measure a half wave endfed of similar ?
I was thinking along the lines of resistive shunt and back calculating.
I wonder how others tackle the problem ?
73 de Andy
|
Just wondering about how to increase the highest value of impedance
measurements. Since it tops out at about 600 ohms in HF
I was thinking along the lines of resistive shunt and back calculating. Without fully working thru the math, I >>guess<< that back-calculating shunt would lack better accuracy (S/N) than straight measurement. I wonder how others tackle the problem? A green (not blue) ~US$10 reflection bridge e.g. from eBay with matched SMA references (want a matched pair to calibrate nanoVNA CH1)
|
you can use impedance transformer. Transformer impedance ratio is a square of turns ratio (or voltage ratio).
For example, RF transformer with 1:2 turns ratio will works as 1:4 impedance ratio.
|
Hi Andy,
if you are able to do a little Math, you can use an S21 through measurement to measure an impedance.
Only issue: there is in the moment no software available for the nanovna to do the calculation. With the DG8SAQ VNWA the software supports this measurement as well as much other analysis modes. For the nanovna we will have to wait, if somebody implements it into the firmware (e.g. the Smith diagram for S21 could do the math) or software.
Some calculations: with 70dB noise flour and usable 60dB dynamic range of S21 (up to 300MHz) using a divider X to 50Ohm you could roughly calculated measure up to 50kOhm (60dB is a 1/1000 voltage divider, 50K Ohm and 50Ohm roughly give 1:1000)
shows that for high impedance S21 serious through and for low impedance S21 shunt through should be used. In between the S11 reflection method delivers good values.
vy73 de Karsten, DD1KT
|
Yes, see message #401 from RF And Microwave Mag on this topic.
Alan
|
I have a VNA that I am working to get sold in kit/assembled form that can
measure 10000+ ohms in the HF in reflection, and can measure 1000+ ohms up
to 600 MHz. I will admit, however, measuring high impedance at VHF/UHF
isnt easy as even tiny amounts of stray capacitance ruin the measurement.
Good fixturing is a must.
In transmission, 10000 ohms is approaching the limit of the S21 measurement
(-50 dB) because of leakage in the SI5351 clock generator. The only way I
know really to fix this would be to use two separate SI5351s to generate
the LO for each port. Not sure if this is worth the effort but certainly
could be done.
It uses the bridge of the EU1KY analyzer which (IHMO) is better designed.
I think the oscillator pins of the SA612s on the NanoVNA are overdriven and
the phase jitter of the IF (which can be considerable when doing a
heterodyne measurement) is not tracked. A Gilbert cell mixer to remain
linear must have its input remain well below 26 mV kT/q (at room
temperature). Also applies common signals to both inputs of the SA612 for
a reason I don't understand, this seems counterproductive.
The design is well shielded and can average over long periods to produce a
relatively high SNR signal from the very low signal input into the SA612.
Hopefully the kit will be available soon....
73,
Dan
KW4TI
|
I wonder how others tackle the problem ?
|
On Sun, Nov 17, 2019 at 02:40 PM, Oristo wrote:
I have resisted posting anything for the last day or two, since I want to re-evaluate some of my own tests and thoughts. But reading the link above, am I to understand that it's worth trying to calibrate with a load other than 50R then doing some simple back tracking calcs ? 73 de Andy
|
Hi Oristo It is a difficult task to measure highZ components based on two restrictions. For the first the DUT used to test the VNA test set up is it really a HighZ with flat response over a large frequency range and for the second part does the test setup destroy the measurement of the well calibrated VNA due to parasitic components. I have done several reports on the topic where I demonstrated the need to characterize the test adaptor and proven it is possible to measure a 10Kohm resistor pretty accurate. One of the obstacles is that it is forgotten that VNA read outs are as series impedance R+jX and it changed to parallel representation it suddenly is showing fat better results I see the article you refer to has fought these aspects and not found the clue as the VNA used are probably able to do better but the testjig is "killing" the results. For a test with the NanoVNA the best way is to use a SMA PCB Edge/bulbhead adaptor and solder the component to centerpin and ground with minimal solder (fringe C) and no leads added. The calibration must be with female kit setting in the NanoVNA-Saver and read out as parallel R and Parallel X. There is one more "snake" as the measurement plane is not the calibration plane for the male SMA to which the female test adaptor is fitted but the QRP firmware has added the feature of positive delay shift so that calibration plane is moved to the rear of the bulkhead/PCB edge adaptor and thus elimination the shunt C and series inductance og said adaptor. Thinking of finding a way to measure in S21 mode with the NanoVNA is "waste of time" as it requires that both the Ch0 and Ch1 output/input impedance is pure 50 ohm. In the appendix2 of one of the report S21 measurement documented. to calibrate and measure a DUT like a toroid in shunt mode_2.pdf Http://www.hamcom.dk/VNWA/How to calibrate and measure a DUT like a toroid correctly the smart way.pdf Copy paste the URL and watch out it is tough reading so just skim it Kind regards Kurt -----Oprindelig meddelelse----- Fra: [email protected] < [email protected]> P? vegne af Oristo Sendt: 17. november 2019 15:40 Til: [email protected]Emne: Re: [nanovna-users] Increasing measurement range (ohms) ? I wonder how others tackle the problem ?
|
Hi Kurt - Thanks as always for your expertise and patience for explaining VNA best practices! This second paper motivates mastering T adapter techniques, particularly for low impedances. Thinking of finding a way to measure in S21 mode with the NanoVNA is "waste of time"
as it requires that both the Ch0 and Ch1 output/input impedance is pure 50 ohm.
In the appendix2 of one of the report S21 measurement documented.
|
Hi Andy - am I to understand that it's worth trying to
calibrate with a load other than 50R
then doing some simple back tracking calcs ? The W0QE paper indicates that, seemingly because of errors introduced by parabolic curve fitting in AIM4170 software, calibrating with higher impedance yielded results that he liked better. Some issues are: * At least for the nanoVNA resistance bridge, sensitivity drops away from 50 Ohms. Sensitivity is related to curve slope steepness here: ...so e.g. errors and noise have more impact. Since a reflection bridge can use a reference load higher than 50 Ohms, it might work better for higher impedances, but requires correcting and converting nanoVNA S21 values. * Where to find / how to verify high frequency impedance calibration references? With good high impedance calibration references, a 9:1 transformer may be characterized and used. As Kurt pointed out, W0QE S21 result depends on ideal (or at least precisely known) CH0 and CH1 impedances.
|
Hi Oristo My comment regarding about "waste of time" regarding S21 measurements was solely based on how to save a S21 sweep and convert it from transmission to reflection. There are other option which I have tested to night where an ordinary SOL calibration is done ether in shunt mode or Serial mode. The shunt mode where a T-Adaptor is inserted between the two SMA male male test cables and a SOL calibration done at the third leg of the T-Adaptor. If the T-Adaptor is FFF then use the male calibration kit and in the NanoVNA-Saver you may use either ideal or the settings I published for male kit. If you use a FMF T-adaptor allowing a female bulkhead/PCB edge adapto to be fitted where the DUT is soldered between center pin and ground, then calibrate with the female calibration kit setting in the NanoVNA-saver. I have attached screen copies of the shunt method testing a 10K resistor and it works actually when seeing the parallel R and X values in the data fields for the three marker. The reading a fluctuation due to noise but by continuous sweeping and using the NanoVNA-saver averaging function it is actually not that bad. However the Shunt method are only optimum for impedances from 0 ohm and up to some 1K ohm For the S21 series SOL calibration I used the three pin pin adaptor seen on the picture constituting Short Open and Load (two 100 ohm 1% SMD resistor soldered to the pins) and the screen copies demonstrates clearly that measuring 10Kohm is piece of cake when studying the marker data as parallel R and X. The window for R+jX is on the other hand "terrible" only because the R+JX read out and due to the 110fF measured as the capacitance of the two pins to which the 10K resistor is soldered. Ig I had used an open standard like the adaptor with 10kOhm fitted the this shunt c would have been smaller and been the shunt C of the SMD resistor and the added fringe C due to the solder added to the end of the SMD resistors. As you see when you have control of you test setup like I have demonstrated. The it is possible up to at least 100MHz to reach accurate measurement. The traces in the Smith chart shows also for the shunt method the added delay due to the Bulkhead adaptor (as the calibration is at the SMA male reference plane). It can be tempting to use the Electrical Delay in the NanoVNA menu but due to the lag of error correction in the NanoVNA it does not work satisfactory. Final remark: Prior to do these two new calibration methods for the NanoVNA in conjunction with the NAnoVNA-Saver do a full SOLT calibration with the two male male test cables fitted from the NaNoVNA menu and save to C0. It is always a good idea to have these two test cables fitted. This calibration is essential for the NanoVNA-saver to build upon despite it does account for the delay of the female female adaptor both for S11 and S21 calibration but no to do anything about and that is why the Electrical delay does not work properly. Have fun Kind regards Kurt -----Oprindelig meddelelse----- Fra: [email protected] < [email protected]> P? vegne af Oristo Sendt: 17. november 2019 21:52 Til: [email protected]Emne: Re: [nanovna-users] Increasing measurement range (ohms) ? Hi Kurt - Thanks as always for your expertise and patience for explaining VNA best practices! This second paper motivates mastering T adapter techniques, particularly for low impedances. Thinking of finding a way to measure in S21 mode with the NanoVNA is "waste of time"
as it requires that both the Ch0 and Ch1 output/input impedance is pure 50 ohm.
In the appendix2 of one of the report S21 measurement documented.
|
There is another way to skin the cat so to speak and measure these non resonate high (or low) Z antenna. That is to say an antenna not near 50 ohms.
Consider going through the effort of constructing a careful and well characterized "L matching" system, one where the L and C components are operable thru a HF range where the self resonate characteristics can be neglected. Then this LC matching system can be taken to the antenna under test and simply use the VNA as an accurate high dynamic range reflectometer. You can readily tweak the match system to obtain a 40 or 50 dB return loss! Then the LC match system now retains the information you desire, namely the Z of the antenna under test. If you take the network back to the shack and terminate the VNA port into 50 ohms, then the antenna port conjugate is the desired antenna Z. Now you could try to measure this Z with your VNA again, or have a calibrated L and C value already recorded on this little LC match system tool. Knowing the L and C values required to achieve the good return loss permits you to calculate the antenna Z.
Of course, you may consider a PI match or a TEE, etc... dependent on the range of expected antenna Z in lieu of the L. However, the L is attractive as it results usually in a minimum Q match solution.
Alan
|
Greetings Alan, You have proposed an idea on this subject that my old brain can almost understand! :) I have a couple questions, if you care to elaborate. I will intersperse them below. On Sun, Nov 17, 2019 at 08:20 PM, alan victor wrote: Consider going through the effort of constructing a careful and well characterized "L matching" system, one where the L and C components are Rather than constructing, would something like an inexpensive "QRP Antenna Tuner" be a useful accessory for the nano? It's just a stepped inductor and air-variable capacitor, with box and connectors, with a moderate impedance matching range. Then this LC matching system can be taken to the antenna under test and simply use the VNA as an accurate high dynamic range reflectometer. You can readily tweak the match system to obtain a 40 or 50 dB return loss! Understood. But I think a more typical use-case would be measuring at the end of the feed-line (typically low-loss open-wire parallel "ladder line"), as for high-Z non-resonant antennas, the feed-point impedance is rarely meaningful, and the transmission line impedance varies with length and frequency. I suppose one might measure it at the feed-point if the desire is to design a tuning stub or other Z-match device at the feed-point. the LC match system now retains the information you desire, namely the Z of the antenna under test. If you take the network back to the shack and terminate the VNA port into 50 ohms, then the antenna port conjugate is the desired antenna Z. Now you could try to measure this Z with your VNA again, or have a calibrated L and C value already recorded on this little LC match system tool. I'm not understanding how to connect and use the nano to determine the L and C values of this little L-network. I assumed one would need an LC meter for that. Knowing the L and C values required to achieve the good return loss permits you to calculate the antenna Z. My understanding of such things is acquired by practical "hands-on" examples (very limited in math skills). A practical example of determining the L-network's L and C, then converting those to Z would be very much appreciated! Thanks! --kv5r
|
Hello KV5R, sure. Good questions and good points. I'll try to go through your note step by step question with a response.
==================
Rather than constructing, would something like an inexpensive "QRP Antenna Tuner" be a useful accessory for the nano? It's just a stepped inductor and air-variable capacitor, with box and connectors, with a moderate impedance matching range.
==================
YES. AGREE. However, the motivation in building your own is to use the VNA in making certain that the self resonate frequency of each of the matching components are WELL above the application or TEST frequency. Otherwise you are dealing with a frequency dependent L or C. Ideally the L or C maintains a C value or an L value that over its entire variable range is spot on the same value, say at 1 MHz or at 30 MHz! Not all L's and C's behave in that manner! Particularly LARGE HV variable caps have a pretty significant parasitic inductance. Hence their C value will shift quite a bit with frequency despite, you never mechanical change their value! Now normally we don't care about this affect, but if were trying to use the matching network as an instrument, then we may need to take care about this subtle aspect.
========================
Understood. But I think a more typical use-case would be measuring at the end of the feed-line (typically low-loss open-wire parallel "ladder line"), as for high-Z non-resonant antennas, the feed-point impedance is rarely meaningful, and the transmission line impedance varies with length and frequency. I suppose one might measure it at the feed-point if the desire is to design a tuning stub or other Z-match device at the feed-point.
==========================
YES! I tend to look right at the antenna feed point. Yes, the series lines and stubs will further modify this Z. But that is part of the MATCHING PROBLEM. Our first challenge is to find the accurate Z at the antenna driving point. So that's where the little high performance matchbox tuner comes in. If we adjust this tuner to obtain a super return loss at its 50 ohm port where our VNA is connected, then the TUNER contains the information we seek, namely the antenna impedance!
===================================
I'm not understanding how to connect and use the nano to determine the L and C values of this little L-network. I assumed one would need an LC meter for that.
============================================
WELL, you are not. However, you are using the VNA to assist in the construction of this well performing tuner whose readout of C and L values (assume a simple L network) are ACCURATE. So you need to carefully characterize the C and L's used in the tuner construction to eliminate as much parasitic from their design. Remember, we are not designing a tuner to handle any power! We are designing a tuner to have stable impedance characteristics. Hence small array of C's or piston caps and an array of tapped air wound L's would be suitable. Recall, the tuner is being driver by the nanovna which has feeble power!
==================================================
Knowing the L and C values required to achieve the good return loss permits you to calculate the antenna Z.
==================================================
WHEN the L and C's of the tuner are set to produce a high performance return loss as measured by the vna, then in essence, if the tuner were terminated (where the vna was positioned) with 50 ohms and we were to look into the TUNER where the antenna was connected, we would see the ANTENNA Z CONJUGATE. Wow, that's a mouth full. The best was to see this is to do an example problem and a simulator like LT Spice is a nice tool to learn. Or there are other SMITH GRAPHIC programs that are quite helpful to assist in this process. Standby and I will see what I can assemble.
Alan
|
Here is a picture story presented as 5 figures which hopefully will clear some questions. In the five figures we start with a circuit representation of a dipole which is a bit unusual as it has a resonant port Z of 500 ohms, pix_1. And it is set to achieve this at 7.1 MHz. If the VNA were connected to such an antenna the super VNA would demonstrate a S11 reflection plot contour vs. frequency on the Smith Chart shown in blue, see pix_2. For now, pay no attention to the plot in RED. We now add our simple L tuner shown in pix_3. And we connect our VNA to the 50 ohm match port and tune to achieve the best return loss possible. See pix_3, the tuner with antenna and pix_4, the return loss. We have achieved a return loss of over 40 dB. If we were to return this tuner network back to calculation what we would find is a conjugate S11 impedance profile that at a single frequency, namely 7.1 MHz that is exactly our antenna Z at this point, portrait of this Z is shown in RED in pix_2. Note it exactly crosses the antenna Z of 500 ohms. If the antenna Z at some other frequency point were complex, the same occurrence would occur. Again, at one specific frequency a real and complex point would occur synonymous with the tuner achieving best return loss at that single frequency. It is worthy to note, that if the tuner were allowed to be more complex, add more elements, that the trajectory of this plot shown in RED would match with higher fidelity over a broader range of frequencies. I hope this clarifies some of the points.
Alan
|
Hello,
There is a trick that is used by people that work on particular frequencies for example in power microwaves 2450MHz.(oven-people)
They use a lambda/4 line at that frequency with a typical 50 Ohm VNA to measure high impedances.
Lambda/4 is a transformer: one side is the low -z end the other the side the high-z end.
Of course the lambda/4 is first characterised ( Q, bandwith, etc.).
I really can't remember nor find the reference to it. It's too long ago, sorry.
Jan ON4MMW
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of alan victor Sent: 18 November 2019 23:17 To: [email protected]Subject: Re: [nanovna-users] Increasing measurement range (ohms) ? Hello KV5R, sure. Good questions and good points. I'll try to go through your note step by step question with a response. ================== Rather than constructing, would something like an inexpensive "QRP Antenna Tuner" be a useful accessory for the nano? It's just a stepped inductor and air-variable capacitor, with box and connectors, with a moderate impedance matching range. ================== YES. AGREE. However, the motivation in building your own is to use the VNA in making certain that the self resonate frequency of each of the matching components are WELL above the application or TEST frequency. Otherwise you are dealing with a frequency dependent L or C. Ideally the L or C maintains a C value or an L value that over its entire variable range is spot on the same value, say at 1 MHz or at 30 MHz! Not all L's and C's behave in that manner! Particularly LARGE HV variable caps have a pretty significant parasitic inductance. Hence their C value will shift quite a bit with frequency despite, you never mechanical change their value! Now normally we don't care about this affect, but if were trying to use the matching network as an instrument, then we may need to take care about this subtle aspect. ======================== Understood. But I think a more typical use-case would be measuring at the end of the feed-line (typically low-loss open-wire parallel "ladder line"), as for high-Z non-resonant antennas, the feed-point impedance is rarely meaningful, and the transmission line impedance varies with length and frequency. I suppose one might measure it at the feed-point if the desire is to design a tuning stub or other Z-match device at the feed-point. ========================== YES! I tend to look right at the antenna feed point. Yes, the series lines and stubs will further modify this Z. But that is part of the MATCHING PROBLEM. Our first challenge is to find the accurate Z at the antenna driving point. So that's where the little high performance matchbox tuner comes in. If we adjust this tuner to obtain a super return loss at its 50 ohm port where our VNA is connected, then the TUNER contains the information we seek, namely the antenna impedance! =================================== I'm not understanding how to connect and use the nano to determine the L and C values of this little L-network. I assumed one would need an LC meter for that. ============================================ WELL, you are not. However, you are using the VNA to assist in the construction of this well performing tuner whose readout of C and L values (assume a simple L network) are ACCURATE. So you need to carefully characterize the C and L's used in the tuner construction to eliminate as much parasitic from their design. Remember, we are not designing a tuner to handle any power! We are designing a tuner to have stable impedance characteristics. Hence small array of C's or piston caps and an array of tapped air wound L's would be suitable. Recall, the tuner is being driver by the nanovna which has feeble power! ================================================== Knowing the L and C values required to achieve the good return loss permits you to calculate the antenna Z. ================================================== WHEN the L and C's of the tuner are set to produce a high performance return loss as measured by the vna, then in essence, if the tuner were terminated (where the vna was positioned) with 50 ohms and we were to look into the TUNER where the antenna was connected, we would see the ANTENNA Z CONJUGATE. Wow, that's a mouth full. The best was to see this is to do an example problem and a simulator like LT Spice is a nice tool to learn. Or there are other SMITH GRAPHIC programs that are quite helpful to assist in this process. Standby and I will see what I can assemble. Alan
|
On Sat, Nov 16, 2019 at 5:59 PM Oristo <ormpoa@...> wrote:
A green (not blue) ~US$10 reflection bridge
e.g. from eBay
with matched SMA references (want a matched pair to calibrate nanoVNA CH1)
Why is a green reflection bridge better than a blue one ? (Presumably it's not just a matter of preferred colour ..)
|
Hi Adrian - A green (not blue) ~US$10 reflection bridge
e.g. from eBay
with matched SMA references (want a matched pair to calibrate nanoVNA CH1) Why is a green reflection bridge better than a blue one ?
(Presumably it's not just a matter of preferred colour ..)
Blue are bad clones of transverters-store design and are wired wrong.
|
Thanks. I had a blue one on order but have a green one on order too, now.
At least they're cheap !
toggle quoted message
Show quoted text
On Tue, Nov 19, 2019 at 12:44 PM Oristo <ormpoa@...> wrote: Hi Adrian -
A green (not blue) ~US$10 reflection bridge
e.g. from eBay
with matched SMA references (want a matched pair to calibrate nanoVNA
CH1)
Why is a green reflection bridge better than a blue one ?
(Presumably it's not just a matter of preferred colour ..)
Blue are bad clones of transverters-store design and are wired wrong.
|
Oristo
Kurt Poulsen
