To be precise one should take into account the velocity factor which is not equal to 1 in case of a non-air-filled coaxial line like in this example. That means that wavelength is shorter and the phase shift is greater than expected.

it been a while... need a refresher course on this... itt we went over smith charts

and please add coaxial chokes to that, too. although they are pretty close to each other

And if you could, please include how to measure CMRR in a Balun. I have been looking into this and found some info but there seems to be multiple methods and nothing I found explains it in a away I fully understand. Thanks for all your great videos. I have learned a lot from them.

You are definitely right. I always add calibration standards to test fixture PCB hosting the DUT. TRL calibration using these embedded standards placed at the desired reference plane performs quite well, provided that frequency range satisfies phase variation between 20 and 160 degrees for phase unambiguity as per TRL requirements.

@@paolomonai9511 Regarding the frequency ranges: utilize multiple TRL lines with center frequencies according to predetermined frequency ratio. Most of the time ratio 4. I.e. Line 1: 0.25to1GHZz, line 2: 1to4GHz, line 2: 4to16Ghz.

@@stefanhummelink3662 Yes, of course the use of muliple lines is the right thing to do for a very wideband characterization, for example multioctave bandwidths. Regards from Italy.

And the moral of that is: ALWAYS get a sample first and measure it yourself!

@@hectorpascal If you want to be sure that a job is well done, do it by yourself...it is the only way!!! Regards from Italy.

Well, I can't afford phase stable cables either, so I don't have any to demonstrate/show.

I've got several other videos that describe a lot more about the Smith chart.

Check Gregg Messenger's channel

@@RobertResearchRadios Thank you. I also find Gregg's channel informative. But although I am quite new to this I feel that more could be done than measuring common mode rejection. For instance can I use the nanoVNA to measure the characteristics of an unknown balun without tearing it apart?

They're on my "long list" of future video topics.

@@w2aew Thanks, till then we wait and enjoy your excellent video !

They’re a pair of m-f adapters (connector savers) from Mini-Circuits. Calibration was done after adding them.

@@w2aew Thank you sir!

@@w2aew I forgot to mention why I requested a calibration kit video. I'm expecting the NanoVNA-F soon and there is a beta firmware to increase the frequency to 2.7gigs. Cheaper kits may be okay up to 900 M but calibration may be off at 2.7. Thanks!

For those interested: www.minicircuits.com/pdfs/SF-MQK50+.pdf

Maximum power transfer from the amp to the antenna system (antenna plus feedline) will occur when the input impedance to the antenna system is equal to the complex conjugate of the amp's output impedance.

The answer is a qualified yes - meaning that there are a few ways to measure impedance, and each method has different limits of impedance it can measure accurately. Some involve some calculations. I hope to do some videos on it - it is a very popular question.

Nope - I don't play guitar - I have no musical abilities whatsoever.

I don't have an "F" model, but I assume that the calibration process is the same. Setup your stimulus frequency range, clear the existing calibration, run through the cal process applying the open/short/load at the end of the lead, APPLY the calibration and save it.

I believe that all of the NanoVNA-H4 models come with a housing.

Port extension assumes a nice, lossless extension of the 50 ohm environment. Depending on your frequency of operation, 5m of RG58 may introduce a few dB of loss, thus giving some misleading results when using port extension. Better off building your own OSL cal kit to use at the end of the cable if your goal is to establish your measurement plane there.

Awesome, glad this video was helpful! Nice to hear from you, old friend!

I think that you'll find that the NanoVNA lacks the required dynamic range to properly tune a repeater duplexer. This video shows this: kzfaq.info/get/bejne/fc-gdrl22M_ViJc.html

@@w2aew Yeah, it was kinda funny right after I asked the question I clicked on your video and found you had beat me to it haha, Thank you for getting back to me though :) 73's

Performing the calibration at the far end of the added connector, fixture, etc. is the PREFERRED thing to do, this will take the extra length into account. You would use Port Extension *only* when it isn't possible to do this (maybe due to mis-matched connector types, etc.).

@@w2aew If I am using a small coax cable from my VNA that does not have a standard SMA connector on the DUT end but rather a stripped piece of wire that I solder to my board, can I just use a 0 OHM, 50 OHM and no resistor to perform the calibration ?

@@eduardorey8238 often times yes, but gets trickier as you go higher in frequency.

@@w2aew I also need to solder a coax to the PCB to tune the Pi network of a Bluetooth LE ceramic antenna. Is there a way to verify that the 50 Ohm resistor behaves like a resistor (inductive/capacitive part stays negligible) and that the calibration is somewhat valid?

@@marcspeck If you complete the calibration to the ports of the VNA, then solder the same type of 50 ohm resistor to the end of the coax, and connect that to the VNA, it should still look like a 50 ohm resistor (because a properly terminated 50 ohm line will look like a 50 ohm resistor). Then you'll know that the far-end termination is OK. Otherwise, you can use Port Extension.

There are safety concerns regarding the shipment of Lithium-based batteries - some people would rather not take the chance. And some hobbyists already have these on hand, and can save money by receiving a cheaper unit without one.

The connector savers are also simply known as a male-female adapter. Here is a good example: www.minicircuits.com/WebStore/dashboard.html?model=SF-SM50%2B

Yes! Maybe one with a HDMI output ?

I have had instructors in the past that, for me for example, made everything so easy to understand but for others flew right over their head. Your style makes it easy for even the slow Learners to understand but not boring for someone, for example, like me. Thank you so much for doing this for us!

If you need a larger display, you can connect it to a computer and use a program like NanoVNASaver.

Me too

The one that I am using is a 4" screen. It is the NanoVNA-H4. Another 4" version is the NanoVNA-F.

If the antenna is balanced or has a proper counterpoise, then the coax will not change the SWR. The coax *will* transform the impedance, but not change the SWR. So if your goal is to learn/measure the SWR at your transmitter, then simply calibrate the VNA locally and make the measurement. If you wish to learn the actual feedpoint impedance of the antenna itself, then either calibrate at the end of the coax (then connect it to the antenna), or use the electrical delay setting to move the reference plane to the antenna (assuming you know the exact electrical length of the coax).

Even though they don’t measure phase, the amplitude of the signals on a line with reflections is still affected by phase.

@@w2aew Thank you!

There really isn't an optimal cable length - SWR does not change with cable length.

This is the NanoVNA-H4. In the US, here is a good source: www.randl.com/shop/catalog/product_info.php?products_id=75145&osCsid=h25v4pc5ilnipaj13ddg8bs1p0

Not really very practical - mainly because XY operation isn't very straighforward. It can be done though...

@@w2aew Alan, here is something interesting I've just now noticed after watching your videos for years; at 57, this new hobby reminds me of my 8~10 year old days. Had to hurry and take it apart and look inside. I.E Nice fly reel, 4 screwes later... Boing! Cahboom!... So I asked about a curve tracer crap portable tv. If it was easy, everyone... But, if some crazy person made a small b&w tv curve tracer, would it only be black and white? No cool green colors?

Mine is the NanoVNA-H4 running the firmware it came delivered with (v0.5). S11 is a reflection measurement - which tells you how much of the applied signal is being reflected back towards the source, and the phase shift induced by the reflection off of the DUT input impedance. It is also known as the reflection coefficient. 0dB means 100% reflection, and negative dB values indicate a smaller reflection. A perfect match would be a very large negative number. Return Loss is just changing the "sign" of the Reflection coefficient. Thus if the magnitude of S11 is -10dB, then the Return Loss is +10dB. The greater the number, the more return signal that is "lost" at the DUT input (i.e. more signal accepted by the DUT). S21 is the transmission coefficient - i.e. how much of the applied signal passes through the DUT. If the DUT is an amplifier or active mixer, etc., the the signal coming out of the DUT can be greater than the applied signal. In this condition, S21 would be a positive number and would represent gain. When there is attenuation through the DUT, some of the applied signal is lost, and the S21 will be a negative number.

Unfortunately, flying leads with alligator clips just don't work well with a VNA because there is no controlled impedance - it varies as you move the leads around. The only way to get somewhat consistent results is to affix the leads to a stable physical location and then calibrate. Even with this, it will be difficult to get consistent results.

@@w2aew Maybe I should have get the BNC/Banana binding post instead, they are "fixed"... Thank you for the precisions and awesome videos !

Are you calibrating to the end of the coax leads (putting the open, short, load at the end of the leads when calibrating)? That should remove the phase shift and resulting rotation on the Smith chart with nothing connected.

Yes, if you can perform the SOL calibration at the balun end of the cable, this will take out the effects of everything up to that point, and will be superior than using port extension. Just be sure to perform that calibration over the same frequency range as your intended test.

@@w2aew Awesomely presented & timely videos on the whole VNA topic, which I am just starting to explore. I hope to apply the insights gained to my HAM Antenna DIY projects. Thank you so much for posting these videos. I have been using the MFJ-269 as my 'goto' tester; and expect to receiver a nanoVNA -F shortly. After going thru the long list to nanoVNA variations found on Amazon, I picked the '-F' version; mainly for the larger display and the ability to store field values. Do you have any comparative comments on the '-F' version of the nanoVNA, vs the one you use in this video ??

@@cache4pat I've never used the -F, so my comments are based only what I've read. The metal case on the -F is nice. Also, I like the fact that the -F *properly* labels the ports as Port 1 and Port 2, instead of CH0 and CH1 like the -H/H4. The -F also has vertical axis labels which are nice. But, I think the -F doesn't go all the way to 1.5GHz like the -H4. I think the RF performance of the -F is little better.

Yes you can. This tip is useful for situations where it might not be possible or practical to run the calibration with the extension included.

@@w2aew .I just ran through this with my Nano V4, Really easy.. And I understand why this would be helpful without disturbing your calibration settings. THANKS AGAIN ALAN !! *73* N8QZH

One assumes a perfect 50 ohm extension and one doesn’t.

@@w2aew Thank you.

They are reset to zero when to turn the device off. When the electrical delay has been added, it is shown in white text in the upper left of the display (Edelay value)

@@w2aew Thank you so much, once I got off my lazy butt and did it myself it became apparent, you are a muse, god bless.

You really can't - it does not measure power directly.

Sometimes it isn't possible to use the OSL standard at the measurement plane of interest. One example is testing the impedance of a printed circuit antenna that is sitting at the end of a microstrip transmission line on the circuit board. The VNA would connect to the input of the transmission line, but the desired measurement plane is at the antenna end of the line, where the matching network would be applied. Port Extension would be used to "move" the measurement/reference plane from the input of the line, down to the far end (antenna end) of the line.

It's actually *always* a square wave. It only tunes natively to 300MHz. Operation from 300-900MHz uses the 3rd harmonic of the squarewave, and operation from 900-1500MHz uses the 5th harmonic of the squarewave.

@@w2aew ahhh got it, thanks! Actually, I'm having a hard time calibrating it in the 900-1500 MHz band. The reponse here is not flat afterwards. Have you experienced this also?

@@mstjerning8919 I seem to get a decent result in 900-1500MHz. Not perfect, but not bad. It isn't going to be as good as the sub-300 and 300-900MHz bands for several reasons, the primary of which is because the stimulus signal is so much lower. It is 15-20dB lower than the fundamental (180 to 300MHz)

In the case I showed in the video, you could. However, in real life, it may not be practical to apply the calibration standards at that point (it might be buried in a circuit board trace or inside of an assembly).

You can think of Group Delay as a plot of electrical delay vs. frequency. Often times, the delay through a circuit will vary with frequency, and this can cause distortion in wideband RF signals.

@@w2aew how can electrical delay and group delay cause distortion in the RF signal? When using S21 gain or S21 phase I don't see the distortion in the RF, how can you tell an view the distortion on a VNA or Spectrum Analyzer? The Distortion will look like what

I'm very sorry to see you go - I have no control over this. Are you saying that if you click on the "bell" icon that you are no longer seeing notifications?

Good question, as I have been doing the same thing and it seems to be working just fine. To me, I thought calibration took everything into consideration to include internal PCB delays and all to include cable length to the terminating end?

Hell yea! W2AEW shows the method in the video to work great of you do not have the option to calibrate at the end of your system connections! For lower frequencies this method by W2AEW works perfect, but if we as RF designers need absolutely accurate data you shall and cannot mitigate the correct and full 2-port TOSM or TRL calibration to calibrate your VNA system to either the end of your SMA(or i.e. smp) cable if your only interestes in the behaviour of your DUT with its connectors in total as a black box, OR to calibrate the VNA system to a plane right next to you DUT in a PCB for instance. For the latter your require a PCB callkit that contains the same connector transition to the actual PCB and should have a open and short standard at the plane location of interest. The callkit should also contain an through, which are simply two open standards connected back2back (no exactly true since the open standard it self should compensate for the fringing field at the open end).... The KIT also contains a match/load (perfect 50 ohm termination).

@H Higgins yea, sounds familiar. Luckily I get to use SMA/B/P interfaces for all my RF applications 😎.

For sure. Just connect the cables and perform calibration

Performing the calibration at the far end of the added connector, fixture, etc. is the PREFERRED thing to do, this will take the extra length into account. You would use Port Extension only when it isn't possible to do this (maybe due to mis-matched connector types, etc.). You have been doing it right!