[TowerTalk] Test Fixture for Common Mode Chokes

[Jeff Blaine]

Brian, would you consider the N2PK VNA in the "fancy" category rearding 
your calibration comment?

Tks

73/jeff/ac0c
alpha-charlie-zero-charlie
www.ac0c.com

On 1/23/2026 6:27 PM, Brian Beezley wrote:
> AC0C said:
>
> "It looks like the Y21 method is a bit better, especially with 
> increasing frequency, but the S21 is a decent proxy as is mentioned on 
> the web page article already."
>
> Whether the Y21 method is needed depends entirely on the test fixture. 
> Because it's more convenient to use the VNA's built-in S21 method, I 
> recommend generating an .s2p file for the S-parameter plotter and 
> check the difference. If the error is tolerable, use S21 in the VNA.
>
> I forgot to mention that both the S21 and Y21 methods assume the port 
> impedances are 50 ohms. NanoVNA calibration compensates for any 
> deviation from 50 ohms for port 1 but not for port 2. Here's an 
> example of the wide deviation of the port 2 impedance for a NanoVNA-H4:
>
> https://k6sti.neocities.org/vnaz
>
> The impedance error will directly affect the results.
>
> Recent firmware for the NanoVNA-H and NanoVNA-H4 offers partial port 
> impedance compensation with something called Enhanced Response. It 
> seems quite good. You can download the latest firmware here:
>
> https://github.com/DiSlord/NanoVNA-D/releases/
>
> If you have a fancier VNA that measures S11, S21, S12, and S22 and does
> 12-term calibration, any port 2 impedance error will be calibrated away.
>
> Finally, a common-mode choke is a distributed device. Common-mode 
> errors can arise due to coupling to the VNA cables and the VNA itself. 
> You can minimize the problem by not handling or touching the VNA while 
> making a measurement. Use the plastic stylus that came with it. Don't 
> connect a USB cable during measurement. Using the Y21 method helps 
> reduce CM as well.
>
> Brian
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk