Author Topic: Smitih Charts, still relevant in 2024?  (Read 3787 times)

0 Members and 1 Guest are viewing this topic.

Offline W4PJBTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: us
Smitih Charts, still relevant in 2024?
« on: June 18, 2024, 04:44:20 pm »
A little backstory on this question. Recently I met with a seller to look at an HP 8714 ET network analyzer. The seller was a retired engineer from, let's just say, a large radar company that we've all heard of.

I asked if the unit had a Smith Chart, and he literally laughed at the idea of still using one. According to the seller, nobody in his former employer used Smith Charts for decades, and even remarked it would be laughable among his peers to use one.

Now I'm certainly no EE, just an idiot learning electronics who also likes playing with radio and antenna systems. I'm sure there is a lot a VNA can do I'm not familiar with (thus the reason to get one, to learn) beyond what a Smith Chart can produce. So, are Smith Charts really a thing of the past in most professional environments? Or do they still find common use by those who actually do this for a living?

Of course, it's also possible he was simply downplaying the fact that his unit didn't have one, so I would still buy it (I didn't). But it did get me to wondering  :-//

 

Offline ZigmundRat

  • Regular Contributor
  • *
  • Posts: 130
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #1 on: June 18, 2024, 05:38:05 pm »
Why wouldn’t it be? Tabular data does not give you the visual representation and intuitive insights that the Smith Chart provides. You can find complex impedance and reflection coefficients at a glance. I think your engineer was a snooty know it all type that wanted to hold their ‘superior’ knowledge over you.
 
The following users thanked this post: W4PJB

Offline paul@yahrprobert.com

  • Regular Contributor
  • *
  • Posts: 91
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #2 on: June 18, 2024, 05:42:51 pm »
The Smith chart is a mapping from the complex reflection coefficient (the rectangular grid) to the complex impedance Z=(ZL-Z0)/(ZL+Z0).  In the bad old days of slide rules it would take you quite a while to do that complex arithmetic by hand, and so Smith charts were very handy.  Nowadays with computers you don't need Smith charts.  Some people still like them as a way of visualizing things, but if you're handy with Matlab, Octave, or Python you can do things quicker and way more accurately.
While we're at it, lets get rid of VSWR and just talk about reflection coefficients (voltage or power).
« Last Edit: June 18, 2024, 10:19:02 pm by paul@yahrprobert.com »
 
The following users thanked this post: AgiRigSig, W4PJB

Online Wallace Gasiewicz

  • Super Contributor
  • ***
  • Posts: 1381
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #3 on: June 18, 2024, 07:44:03 pm »
HP 8714 HAS a Smith Chart
 
The following users thanked this post: W4PJB

Offline Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: Smitih Charts, still relevant in 2024?
« Reply #4 on: June 18, 2024, 09:01:22 pm »
The "engineer" you spoken to was a clueless moron. Likely was a paper pusher, a mechanic or a janitor at the "large radar company that we've all heard of".
Facebook-free life and Rigol-free shack.
 
The following users thanked this post: nctnico, egonotto, tautech, 2N3055, ch_scr, W4PJB

Online tautech

  • Super Contributor
  • ***
  • Posts: 29812
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Smitih Charts, still relevant in 2024?
« Reply #5 on: June 18, 2024, 09:20:43 pm »
The "engineer" you spoken to was a clueless moron. Likely was a paper pusher, a mechanic or a janitor at the "large radar company that we've all heard of".
Precisely.

Smith tells us what other sweeps cannot.
From ~900 MHz Yagi:

Avid Rabid Hobbyist.
Some stuff seen @ Siglent HQ cannot be shared.
 
The following users thanked this post: egonotto, W4PJB

Offline W4PJBTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #6 on: June 18, 2024, 09:31:40 pm »
HP 8714 HAS a Smith Chart

The ET model (reflection/transmission)? Or only the ES model (S parameter)? We looked through the menu briefly and could not find it. Obviously, he would have never used it if it was there anyway.
 

Offline W4PJBTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #7 on: June 18, 2024, 09:37:22 pm »
The Smith chart is a mapping from the complex reflection coefficient (the rectangular grid) to the complex impedance Z=ZL-Z0/(ZL+Z0).  In the bad old days of slide rules it would take you quite a while to do that complex arithmetic by hand, and so Smith charts were very handy.  Nowadays with computers you don't need Smith charts.  Some people still like them as a way of visualizing things, but if you're handy with Matlab, Octave, or Python you can do things quicker and way more accurately.
While we're at it, lets get rid of VSWR and just talk about reflection coefficients (voltage or power).

I think this was his point. I have no experience with any of those, so that's where my ignorance lies.  :-//
 

Online Wallace Gasiewicz

  • Super Contributor
  • ***
  • Posts: 1381
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #8 on: June 18, 2024, 09:47:16 pm »
Mine is an HP 8713 which I upgrades to 8714 with a EPROM change, You should read a little about these things before you buy.Here is a short Keysight Summary of the 8714 ET it includes Smith Charts:     

https://www.keysight.com/us/en/product/8714ET/rf-network-analyzer-tr.html
 
The following users thanked this post: egonotto

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: Smitih Charts, still relevant in 2024?
« Reply #9 on: June 18, 2024, 10:04:19 pm »
If was regularly doing RF impedance matching, then I would be regularly using a Smith chart.
 
The following users thanked this post: egonotto, Howardlong, W4PJB

Offline W4PJBTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #10 on: June 18, 2024, 10:07:07 pm »
Mine is an HP 8713 which I upgrades to 8714 with a EPROM change, You should read a little about these things before you buy.Here is a short Keysight Summary of the 8714 ET it includes Smith Charts:     

https://www.keysight.com/us/en/product/8714ET/rf-network-analyzer-tr.html

Thanks! I passed on this unit (price was right and it was local, only reason it was on my radar) and actually purchased an 8753C w/ 85046A S-Parameter test set last night. Seemed like a much better deal and was already upgraded to an LCD screen. The screen on the 8714 I looked at was pretty dim and burned in.
 
The following users thanked this post: egonotto

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 12277
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #11 on: June 19, 2024, 12:30:37 am »
...
Some people still like them as a way of visualizing things, but if you're handy with Matlab, Octave, or Python you can do things quicker and way more accurately.
...

I don't see why you couldn't have your cake and eat it too.  Or, why you wouldn't.    I wouldn't want to trade in an XY graph on an oscilloscope for some digital readouts.  Same for the SA, logic analyzer....   I would rather see the waveform and for complex impedances, Smith.

Offline Georgy.Moshkin

  • Regular Contributor
  • *
  • Posts: 212
  • Country: hk
  • R&D Engineer
    • How to become a Tech Sponsor
Re: Smitih Charts, still relevant in 2024?
« Reply #12 on: June 19, 2024, 03:23:51 am »
When we need to visualize complex number, we can use polar plot.
E.g., in Octave
Code: [Select]
handle1=plot(real(s11), imag(s11), 'Color',[1.0,0,0]);
Simply using real part of s11 for x coordinates, and imaginary part for y coordinates.
This graph also can be called a Smith chart, we can add constant resistance circles, stability circles, etc.
There is no need to eliminate Smith chart, as well as polar coordinate system  :-DD

Smith chart is used in a lot of books and recognized by many people. It may help to remember some concepts. E.g., see here is Smith chart visualization of patch antenna inset length tuning compared to patch dimensions tuning: https://youtu.be/v8uxhbu36Ug?si=GXQMQTdlT_ddQQCy&t=32

When we tune inset length, we jump between resistance circles (S11 curve radius changes)
And when we change patch dimensions, we tune resonant frequency (point moves along S11 curve)

Maybe seller meant that Smith charts are no longer used in paper form to design matching circuits. A lot of CAD programs provide polar plots and call them Smith charts. Some hatred toward Smith charts is caused by unnecessary complication: drawing all those circles at different offsets and pushing into student heads. Maybe teachers should have used simple polar plots without all those circles and terminology first. So, if Seller doesn't like Smith charts, it's not his fault!

Smith chart would be especially handy if I design series-feedback oscillator, tunable impedance block (measuring S22 for FET while tuning Vgs bias).
« Last Edit: June 19, 2024, 04:07:55 am by Georgy.Moshkin »
Disappointed with crowdfunding projects? Make a lasting, meaningful impact by becoming a Tech Sponsor today. Visit TechSponsor.io to Start Your Journey!
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 29812
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Smitih Charts, still relevant in 2024?
« Reply #13 on: June 19, 2024, 07:37:25 am »
When we need to visualize complex number, we can use polar plot.
E.g., in Octave
Code: [Select]
handle1=plot(real(s11), imag(s11), 'Color',[1.0,0,0]);
Simply using real part of s11 for x coordinates, and imaginary part for y coordinates.
This graph also can be called a Smith chart, we can add constant resistance circles, stability circles, etc.
There is no need to eliminate Smith chart, as well as polar coordinate system  :-DD

Smith chart is used in a lot of books and recognized by many people. It may help to remember some concepts. E.g., see here is Smith chart visualization of patch antenna inset length tuning compared to patch dimensions tuning: https://youtu.be/v8uxhbu36Ug?si=GXQMQTdlT_ddQQCy&t=32

When we tune inset length, we jump between resistance circles (S11 curve radius changes)
And when we change patch dimensions, we tune resonant frequency (point moves along S11 curve)

Maybe seller meant that Smith charts are no longer used in paper form to design matching circuits. A lot of CAD programs provide polar plots and call them Smith charts. Some hatred toward Smith charts is caused by unnecessary complication: drawing all those circles at different offsets and pushing into student heads. Maybe teachers should have used simple polar plots without all those circles and terminology first. So, if Seller doesn't like Smith charts, it's not his fault!

Smith chart would be especially handy if I design series-feedback oscillator, tunable impedance block (measuring S22 for FET while tuning Vgs bias).
Again:

Smith tells us what other sweeps cannot.
From ~900 MHz Yagi:


A Smith and a Polar, add a marker and which provides more info ?
Avid Rabid Hobbyist.
Some stuff seen @ Siglent HQ cannot be shared.
 
The following users thanked this post: Someone

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 930
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: Smitih Charts, still relevant in 2024?
« Reply #14 on: June 19, 2024, 09:22:38 am »
Out of the 215 videos I have on the Rohde & Schwarz YouTube channel, the most viewed, most commented on, and most liked video is the one on the Smith Chart

And yes, a lot of our customers still use the Smith Chart :)

Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: Smokey, egonotto, newbrain, Bicurico, radiohomebrewer2000

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 930
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: Smitih Charts, still relevant in 2024?
« Reply #15 on: June 19, 2024, 09:33:40 am »
Of course, it's also possible he was simply downplaying the fact that his unit didn't have one, so I would still buy it (I didn't).

It's very hard for me to imagine a VNA that doesn't have a Smith Chart display.  R&S is credited with manufacturing the first commercial network analyzer (the Z-g Diagraph) over 60 years ago and even that instrument had a Smith Chart display :)

2291947-0
Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: egonotto, Kean, KE5FX

Offline Georgy.Moshkin

  • Regular Contributor
  • *
  • Posts: 212
  • Country: hk
  • R&D Engineer
    • How to become a Tech Sponsor
Re: Smitih Charts, still relevant in 2024?
« Reply #16 on: June 19, 2024, 09:45:41 am »
Hey, tautech, I think you've got me wrong. I see polar and Cartesian plots on your images.. And what I was saying is that polar plot = Smith chart. And it's probably how it's implemented on most of VNAs firmware.
Disappointed with crowdfunding projects? Make a lasting, meaningful impact by becoming a Tech Sponsor today. Visit TechSponsor.io to Start Your Journey!
 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 12277
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #17 on: June 19, 2024, 12:53:09 pm »
Hey, tautech, I think you've got me wrong. I see polar and Cartesian plots on your images.. And what I was saying is that polar plot = Smith chart.

Using your logic, polar plot = Smith chart = Cartesian plot.   

Quote
And it's probably how it's implemented on most of VNAs firmware.

I have an HP8754A from the 1970s.   It has no firmware.  There are overlays that you insert over the CRT to get the different graph types.   There is a mechanical switch that selects polar or rectangular.

When I wrote the software for it, I map everything to an X-Y graph.  Like the 8754A, I use overlays that map over the top of the X-Y graph.   A technique I still use with my current software. 


Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 930
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: Smitih Charts, still relevant in 2024?
« Reply #18 on: June 19, 2024, 02:10:42 pm »
Coincidentally, I was just using a Smith Chart this morning for a video I'm working on ("Measuring Antennas with the ZPH") so I thought I would share a picture on my "work" Instagram account as well.

https://www.instagram.com/p/C8ZmxqSgWN2
Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: egonotto

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 2393
  • Country: us
    • medpants.com
Re: Smitih Charts, still relevant in 2024?
« Reply #19 on: June 19, 2024, 07:04:13 pm »
When I first got introduced to using smith charts, I found them to be one of the best examples of data visualization in charts that I knew of.  Data visualization, even without analysis, let's us - humans evolved for visual pattern recognition - pull a lot more meaning out of a dataset than when presented as a table, a series of numbers, or some other basic formats.  Your basic trend chart or scatter plot engages our understanding so much more than the raw values, and it's astounding how much useful information is easily gleaned from a smith chart, especially given the relatively modest requirements for plotting data points on one.  A plot that shows impedance matching over frequency that can be normalized to virtually any system impedance, shows meaningful information out towards the extremes, and doesn't just give you match/mismatch information but tells you how so that you can make adjustments to the DUT.

Incredibly practically useful for anyone needing to match RF elements all available at a glance, and the absolute basics of relative match vs. mismatch can be see at a glance by someone with no understanding of reactance or how points are plotted on it.  Really top tier visualization/user interface technology.
 
The following users thanked this post: egonotto, newbrain, pdenisowski

Online tautech

  • Super Contributor
  • ***
  • Posts: 29812
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Smitih Charts, still relevant in 2024?
« Reply #20 on: June 19, 2024, 08:29:45 pm »
Hey, tautech, I think you've got me wrong. I see polar and Cartesian plots on your images.. And what I was saying is that polar plot = Smith chart. And it's probably how it's implemented on most of VNAs firmware.
Trace/Sweep, yes.
Just as Mag Loss and SWR sweeps are very similar yet in all cases addition of markers gives additional info at marker frequencies.
Smith and Polar are the same but their use depends on the info you need from them where Smith is superior.

To get even more precise detail scaling can be adjusted like using the screen overlays that Joe showed for his old analyzer.
In modern analyzers we just adjust the display scaling to see the detail.
Avid Rabid Hobbyist.
Some stuff seen @ Siglent HQ cannot be shared.
 

Offline radiohomebrewer2000

  • Contributor
  • Posts: 39
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #21 on: June 20, 2024, 11:58:15 pm »
I think Smith Charts are still valuable especially for someone learning radio electronics.  It will help you learn complex impedance and what it is made of: inductive reactance, capacitive reactance, and resistance.   

I recently got some Smith Chart paper to draw some Smith Charts. Found a lot of videos of Smith Charts on YouTube.  I did this so it would give me a better understanding of the Nano VNA I got.    But I also want to get into designing my own impedance matching networks.  So, I think knowing Smith Charts will help.

I am the type of person who prefers to draw his schematics with pencil on graph paper and then take a photo of it instead of drawing it via some software.    But, I am an amateur - not an engineer.
 

Offline EggertEnjoyer123

  • Frequent Contributor
  • **
  • Posts: 335
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #22 on: June 21, 2024, 12:44:40 am »
It's way easier to do impedance matching with a Smith Chart tool than without one.
 

Offline mtwieg

  • Frequent Contributor
  • **
  • Posts: 353
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #23 on: June 21, 2024, 11:54:33 am »
I'm wondering if the engineer in the OP was referring to the pencil-and-paper approach to matching network designs by scribing arcs on a smith chart. That certainly is rare nowadays. But I've never seen a VNA that doesn't have a smith chart display.

Or maybe they only worked with scalar network analyzers?
 

Offline bson

  • Supporter
  • ****
  • Posts: 2497
  • Country: us
Re: Smitih Charts, still relevant in 2024?
« Reply #24 on: June 22, 2024, 02:18:26 am »
I don't understand what a Smith Chart really helps with - and never felt I needed one.

Certainly not impedance matching?  I mean, say I have source impedance of 50+j0Ω, a line of 50-j7Ω, and a load of 45-j20Ω.  So matching them requires 5+j27Ω which is a resistor and an inductor.  This is kind of elementary, no?  If I use the VNA to measure the impedances, isn't the impedance discrepancy obvious?  Of course, some devices like antennas don't have a fixed impedance but vary with frequency, usually shifting from inductive to capacitive at the resonant frequency - so can't be matched over a wide band, but how would a Smith Chart help with this?

I feel like I've always missed something here...
Explain it like I'm 5 please?
« Last Edit: June 22, 2024, 02:21:59 am by bson »
 
The following users thanked this post: egonotto


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf