开云体育

I recently wrote a program to renormalize S-parameter impedance. This lets you measure a filter with a 50 ohm VNA even though the design impedance is higher. In the course of testing the program, I needed to plot S-parameters. I searched extensively for suitable viewers, desktop or online. The ones I located were either incomplete, frustratingly awkward, or had bugs.

I decided to write my own Windows S-parameter plotter. It's near the top of the following page. See the bottom of the page for downloading instructions.



The first image compares group delay for two Murata ceramic filters. The next images show the benefit of phase unwrapping.

Brian

I've added impedance measurements to the S-parameter plotter. It now includes the one-port reflection method and the two-port series-through, shunt-through, and Y21 methods. Y21 cancels stray port reactance. The plotter includes reference impedance renormalization. You can measure a component at 50 ohms and see the response at its design impedance. I include a utility to merge files for forward and reverse S11 and S21 measurements to obtain S11, S21, S12, and S22. This lets you use a NanoVNA to fully characterize a component that may have an asymmetrical response, such as a ceramic filter.

The program is at the top of the following page. See the bottom of the page for downloading instructions.



Brian

Don Brant sent me an .s2p file to help test my plotter. It's for an amplifier with several feet of cabling. When I looked at the phase response, I knew I had found a new poster child for the virtues of phase unwrapping.

Brian

Off Topic...
Brian,
I have fond memories of your Antenna modeling program - from back in the day.. I would load parameters and let the computer run all night. I didn't have a co-processor nor could afford the $99 then $69 for one, as time went buy. Life happened, my job took me Detroit to The DC area, then up to Boston area, then out here to Wichita Kansas, and then retirement. 24 years ago my wife Ann passed and left me a 10 year old boy to raise.. That ended ANN-tennas - a swap & shop business that has been dormant since.

However as your antenna modeling software took a back seat, I went on to EZNEC and enjoy that even today.

So thanks for the memories and the contributions to Ham Radio.

If you ever get to "cow town" Wichita, Kansas let me know I have some no fertilizer (can't afford it) no hormones, (why?) grass fed born and raised on my 25 acres - irradiated (100watts) under my homebrew antennas, steaks we can put on the grill. I also have some Canadian beer as my grandkids live in NY and I bring it back from Buffalo. Maybe we can start a beer can antenna of your design..

Larry W8LM
ARRL Life member Licensed 59 years.
past President 2 terms, VFW post 3115 Amateur Radio Club, W?VFW, Wichita, KANSAS

Larry, what a wonderful message. I don't think this is the place to chat so I'll send you an email.

Brian

These plots compare reference impedance renormalization using S11, S21, S12, and S22 (yellow) with using just S11 and S21 (cyan). Shown are results for two different 10.7 MHz Murata ceramic filters renormalized from 50 to 330 ohms. The signal passband is 10.6 to 10.8 MHz. The plots demonstrate that when using a NanoVNA that can measure only S11 and S21, you can benefit by reversing the filter and measuring S22 and S12 as well.

Download the S-parameter plotter with renormalization and Y21 method:



Brian

On Mon, Mar 3, 2025 at 07:44 AM, Larry Macionski wrote:

I have fond memories of your Antenna modeling program - from back in the day..
I would load parameters and let the computer run all night. I didn't have a
co-processor nor could afford the $99 then $69 for one, as time went buy.

Larry,

I still use Brian's excellent antenna simulation programs today. In fact, just about every week. (Different story not for here.)

But, I run them on both a very new macOS computer and a Windows 10 computer using this app:



It works very well. There's a couple features with YO that I haven't gotten working right because of the various calls into and out of a controlling batch file, but overall it's a great experience, still. The best part is that simulations run very quickly - minutes at most instead of overnight.

73,

Clarke K1JX

I finally managed to implement renormalization for unequal port reference impedances. The images show unequal resistances for a 10.7 MHz Murata ceramic filter. I optimized them by hand for minimum group delay variation over 10.6 to 10.8 MHz, the passband of an FM broadcast signal. The optimization also flattened the passband and lowered the spurs.

I'm not quite ready to release this. The new renormalization algorithm I'm using can also handle a reactance shunting either port. This would let you include the standard 10 pF Murata test load. I think it might be useful for other kinds of filters and transformers as well. The problem is that I'm running out of screen real estate, at least on the bottom line. I'll figure out something. In the meantime, would there be any use for a shunt inductance? I don't recall seeing this, but it would be easy to implement.

Brian

Brian,

When I go to this link:
I get the message that ham-radio.com domain has expired.

Is the above link you posted correct.
73,
Geoff - W8GNM

Geoff, this will download the program and documentation:



Brian

I seem to have gotten everything working. The plots show a 280 kHz Murata ceramic filter with an 82 pF load on its output. The dB and group delay curves are similar to those published by Murata for a 230 kHz filter.

I spent a lot of time looking for bugs when I noticed that S12 > 1 for some capacitive loads. I finally found two references that explained how this was possible for a passive device for a complex port reference impedance. It does not indicate a violation of conservation of energy as it would for a real reference impedance. I am using pseudo-wave analysis instead of the older power-wave analysis. They yield identical results for real reference impedances, but only pseudo-wave can handle complex impedances.

I'm going to spend some time checking for bugs as I changed a lot of code. I'll post a message when the new version is available.

Brian

Thanks Brian,
That link worked OK
73,
Geoff

This will download the updated S-parameter plotting program:



I was surprised to discover that the final Touchstone 1.1 file specification provides for multiple, real-valued reference impedances. The plotting program will handle these files.

The maximum reference impedance is 9999 ohms. This seems reasonable given VNA accuracy limitations for high impedances. Some interesting devices are beyond this limit (e.g., 27k ohms for Collins mechanical filters). Shunt capacitance can go to 9999 pF.

The program is not intended to replace NanoVNA-Saver or NanoVNA-App. It provides a convenient way to display all S-parameters, including S12 and S22, and to use the full set for reference impedance renormalization. It handles a different reference impedance for each port, with or without shunt capacitance. No need to build a special, impedance-matching test fixture to measure a filter with nonstandard, unequal impedances. The program provides the stray-cancelling Y21 series-through impedance measurement as well as the usual S21 and S11 series and shunt methods.

Brian