This brings back memories of the microwave lab in school which held the donated network analyzers ($$$) and was behind a keycard door within another lab that was behind its own keycard door.
Still amazed that someone was able to come up with this tool
How times have changed. You can now get a two port (S11/S21) battery-powered color touchscreen NanoVNA vector network analyzer that can do Smith charts, SWR, and all kinds of other things for under $40. I use one to analyze ham radio antennas and it's not perfect, but works well enough for my needs and is an incredible value. The firmware for it is open source, too.
Its performance is nowhere comparable to a real VNA, those VNAs made by Hewlett-Packard in the late 80s are still the golden standard for many labs.
But NanoVNA is great value for the money for anyone interested in radio electronics! For $50, almost nothing beats a NanoVNA, and it's much more effective than the traditional sweep generator setup. You get complex S11/S21 and 900 MHz bandwidth [0], good for measuring complex impedance, the insertion loss of the coax, the frequency response of filters, SWR of antennas, and even basic Time-Domain Reflectometry. The firmware is available under GPLv3 for hacking as well.
I recently used it to experiment different ferrite RF transformers for my homebrew radio receiver, and to characterize frequency response of inductors.
The only thing to beware - since it's a "almost" free hardware design (i.e. FOSS firmware + a block diagram for hardware, only the PCB layout is not available), there are a number of low-quality clones that use low quality components and don't include proper shielding. I recommend hugen79's version ("NanoVNA-H"), he is not the original designer (it was designed by edy555 [2]), but hugen79's version is currently the most common source with reasonable quality, see the picture for comparison. [1]
[0] > 600 MHz uses higher-order harmonics, and is less reliable, but still better than nothing.
There is a forthcoming NanoVNA version 2 that will have improved dynamic range due to better down mixers and added audio op amps.
I think the whole NanoVNA eruption (including the clones and alternatives) is outstanding. I believe there is a large unsatisfied demand for low performance instruments that are sufficient for experimentation below ~3 GHz. There are a lot of used instruments around that are large, heavy and usually need service, and there are a number of built-down-to-a-price entirely proprietary instruments from Asia, but nothing beats a device that fits in your hand, runs free code from GitHub and costs so little that you can fry it accidentally and not care much.
I've already heard about it, looking forward to the new one.
> I believe there is a large unsatisfied demand for low performance instruments
Yeah, there's a huge demand for low performance instruments of all types.
And a surprising amount of such equipment can be built using low-cost parts thanks to progress in semiconductors - very limited performance, often uncalibrated, sure, but valuable for experimenters. For example, a 1 GHz active oscilloscope probe for $50 [0].
I think the critical part is not simply the possibility of building them, but making them easily available. If you read the old homepage on the web, or old electronics magazines, you can see a lot of similar DIY projects. But a lot of those DIY projects are not very repeatable.
But in recent years, I think the rise of the web and the popularity of free and open source software and hardware design is a game changer in this aspect. Today, I can download a PCB design from GitHub and send it to a factory for assemble simply by a few mouse clicks, and a week later, I can get a prototype. Even if the original designer stopped working on it, others in the community can contribute easily and carry on.
> but nothing beats a device that fits in your hand, runs free code from GitHub and costs so little that you can fry it accidentally and not care much.
Smith charts are really versatile. You can use them for all sorts of RF circuit design tasks, as a complex number calculator, and perhaps more importantly a way to visualize impedance. It is almost how people think in their native language, you can visualize how impedances will interact using the smith chart in your head.
Is this comparison based solely on the idea that both of these tools are outdated? If you have something deeper in mind, could you fill in some details of this metaphor? I haven't seen Smith charts before, but a cursory read of TFA doesn't really evoke punch cards.
Smith Chart is both a visualization tool (as a chart) and a calculation tool (as a nomogram), you can use a printed Smith Chart to do RF-related calculations (impedance matching, return loss, etc) with a ruler and compass on a graph paper, pretty much like using a slide rule to calculate logarithm and trigonometry. I think it's why the OP compared it to a punch card.
It's true that today it's rarely used as a calculation tool, but it's still useful as a visualization. I don't think it's obsolete like punch cards.
Punch cards are made thoroughly obsolete many orders of magnitude over by modern storage and IO technologies. What replaced Smith Charts? The >$100k only-a-few-years-old VNAs next room over still plot S11 and S22 on Smith Charts and I'd bet you an entire VNA that this will still be the case in 10 years.
Preemptive: No, 3D Smith Charts do not obsolete Smith Charts, just like 3D line charts don't obsolete 2D line charts.
Well in my EMAG class in college we were introduced to Smith Charts. We studied them for about a week or two and used software to check our work. So software has replaced them. There is nothing a Smith Chart can do that a simple software application can't manage in 1/16'th the time.
A severely misguided analogy. Every VNA made to this day supports a Smith chart display, and every credible course on RF engineering covers them.
Meanwhile, nobody who's not involved in running a computer museum uses punch cards for anything, and there's little or no pedagogical value in continuing to teach IT or CS students about them.
I've found two uses for smith charts: causing students pain and impedance matching. They're good learning tools but imo it's less of a headache to do impedance matching in code.
For actual measurements: there is much less ambiguity in a standard magnitude/phase response, since you actually can see the values of your independent variable (frequency).
They're quite similar to slide rules in that respect. Learning how to use one gives you an intuitive grasp of the operations you might otherwise miss (or not see as clearly), but there are far better tools for doing the actual computation.
On a couple of tests, we actually had to analytically draw our smith charts on graph paper, then use it for then next problem.
Say you want 15 dB return loss at some frequency, well there are two impedances points for that. One with the impedance locus outside Z0, and another with the locus encompassing Z0. The latter has broader bandwidth, but that intuition is lost without a Smith or polar plot.
Alan Wolke (W2AEW) has a good video tutorial (along with his other good radio electronics tutorials) on YouTube about the basic applications of Smith Chart, understandable to most radio hobbyists without an electrical engineering background.
>Smith approached a number of technical magazines for publication of his transmission line
diagram; however the acceptance process was slow. Finally, after two years, Smith’s article
describing his chart was published in the January 1939 issue of Electronics magazine. In a
second article, published in the January 1944 issue of Electronics, Smith incorporated further
improvements into his chart, including its usage alternatively as an impedance chart or an
admittance chart.
In the late '60s, the introduction of PL/I, which had complex numbers as a primitive type and a math library of functions of complex numbers, made transmission line calculations much easier.
No way is the Smith Chart outdated. I use it daily in simulation software and on the VNA. If a Smith Chart is not shown in a meeting, then that meeting is a waste of time.
I have a collection of various RF slide calculation tools. The prized one is rotary sliding Smith Chart.
That makes sense for forward response, when you are optimizing for bandwidth or phase matching, or tuning for frequency response. Smith Chart is pointless for S21, though polar isn’t, though the VNA well let you do the former.
When I’m working on antennas, I’ll have both Smith Chart and log return loss plots stacked on top one another.
Say you want 15 dB return loss at some frequency, well there are two impedances points for that. One with the impedance locus outside Z0, and another with the locus encompassing Z0. The latter has broader bandwidth, but that intuition is lost without a Smith or polar plot.
36 comments
[ 2.9 ms ] story [ 104 ms ] threadStill amazed that someone was able to come up with this tool
But NanoVNA is great value for the money for anyone interested in radio electronics! For $50, almost nothing beats a NanoVNA, and it's much more effective than the traditional sweep generator setup. You get complex S11/S21 and 900 MHz bandwidth [0], good for measuring complex impedance, the insertion loss of the coax, the frequency response of filters, SWR of antennas, and even basic Time-Domain Reflectometry. The firmware is available under GPLv3 for hacking as well.
I recently used it to experiment different ferrite RF transformers for my homebrew radio receiver, and to characterize frequency response of inductors.
The only thing to beware - since it's a "almost" free hardware design (i.e. FOSS firmware + a block diagram for hardware, only the PCB layout is not available), there are a number of low-quality clones that use low quality components and don't include proper shielding. I recommend hugen79's version ("NanoVNA-H"), he is not the original designer (it was designed by edy555 [2]), but hugen79's version is currently the most common source with reasonable quality, see the picture for comparison. [1]
[0] > 600 MHz uses higher-order harmonics, and is less reliable, but still better than nothing.
[1] https://github.com/hugen79/NanoVNA-H/blob/master/doc/clone.j...
[2] https://github.com/ttrftech/NanoVNA
I think the whole NanoVNA eruption (including the clones and alternatives) is outstanding. I believe there is a large unsatisfied demand for low performance instruments that are sufficient for experimentation below ~3 GHz. There are a lot of used instruments around that are large, heavy and usually need service, and there are a number of built-down-to-a-price entirely proprietary instruments from Asia, but nothing beats a device that fits in your hand, runs free code from GitHub and costs so little that you can fry it accidentally and not care much.
I've already heard about it, looking forward to the new one.
> I believe there is a large unsatisfied demand for low performance instruments
Yeah, there's a huge demand for low performance instruments of all types.
And a surprising amount of such equipment can be built using low-cost parts thanks to progress in semiconductors - very limited performance, often uncalibrated, sure, but valuable for experimenters. For example, a 1 GHz active oscilloscope probe for $50 [0].
I think the critical part is not simply the possibility of building them, but making them easily available. If you read the old homepage on the web, or old electronics magazines, you can see a lot of similar DIY projects. But a lot of those DIY projects are not very repeatable.
But in recent years, I think the rise of the web and the popularity of free and open source software and hardware design is a game changer in this aspect. Today, I can download a PCB design from GitHub and send it to a factory for assemble simply by a few mouse clicks, and a week later, I can get a prototype. Even if the original designer stopped working on it, others in the community can contribute easily and carry on.
> but nothing beats a device that fits in your hand, runs free code from GitHub and costs so little that you can fry it accidentally and not care much.
Can't agree more.
[0] http://blog.weinigel.se/2016/02/26/ghz-differential-probe.ht...
It's true that today it's rarely used as a calculation tool, but it's still useful as a visualization. I don't think it's obsolete like punch cards.
Preemptive: No, 3D Smith Charts do not obsolete Smith Charts, just like 3D line charts don't obsolete 2D line charts.
You weren't kidding. From https://en.wikipedia.org/wiki/Punched_card_input/output :
> If all columns of an 80 column card encode information this translates to approximately 2,500 characters per second (CPS).
According to https://www.techradar.com/reviews/samsung-970-evo-plus a 970 EVO Plus SSD is about 1,400,000 times faster. Wow. I would love to go back a few years and blow young me's mind.
Meanwhile, nobody who's not involved in running a computer museum uses punch cards for anything, and there's little or no pedagogical value in continuing to teach IT or CS students about them.
For actual measurements: there is much less ambiguity in a standard magnitude/phase response, since you actually can see the values of your independent variable (frequency).
Say you want 15 dB return loss at some frequency, well there are two impedances points for that. One with the impedance locus outside Z0, and another with the locus encompassing Z0. The latter has broader bandwidth, but that intuition is lost without a Smith or polar plot.
https://www.youtube.com/watch?v=TsXd6GktlYQ
>Smith approached a number of technical magazines for publication of his transmission line diagram; however the acceptance process was slow. Finally, after two years, Smith’s article describing his chart was published in the January 1939 issue of Electronics magazine. In a second article, published in the January 1944 issue of Electronics, Smith incorporated further improvements into his chart, including its usage alternatively as an impedance chart or an admittance chart.
I have a collection of various RF slide calculation tools. The prized one is rotary sliding Smith Chart.
When I’m working on antennas, I’ll have both Smith Chart and log return loss plots stacked on top one another.
Say you want 15 dB return loss at some frequency, well there are two impedances points for that. One with the impedance locus outside Z0, and another with the locus encompassing Z0. The latter has broader bandwidth, but that intuition is lost without a Smith or polar plot.
https://github.com/chartjs/chartjs-chart-smith
http://math-fail.com/wp-content/uploads/smith-chart-tattooo....
https://www.microwaves101.com/uploads/SmithCake-NITT.jpg
f(z) = (a z + b) / (c z + d).
The key feature is that this is a conformal transformation which means it preserves angles (between curves).
Reading it a bit more I guess thats clear.