Calibration for measuring return loss

[flysight]

I have been using the SA44B/TG44A to measure return loss with a Mini-Circuits ZFDC-10-5-S+ directional coupler:

http://www.minicircuits.com/pdfs/ZFDC-10-5.pdf

The directional coupler is only good out to 2 GHz, so I’m treating results past that with some suspicion. However, I’m getting some strange results below 2 GHz as well. I am following the instructions in section 3.2 of the user manual here:

https://signalhound.com/sigdownloads/TG44A/TG44A-User-Manual.pdf

I am measuring the return loss of two Mini-Circuits ANNE-50+ 50 ohm terminators:

https://www.minicircuits.com/pdfs/ANNE-50+.pdf

The results of the test are here:

http://imgur.com/OwMVT3W
http://imgur.com/WimdA4T

What surprises me is that, although the terminators have a specified return loss of at least 30 dB from DC to 4 GHz, I’m seeing a return loss of about 20 dB at 2 GHz.

I have a few questions:

• What might be causing this discrepancy?
• Would there be any value in using a more thorough calibration, e.g., using a short and/or a 50-ohm terminator for the scalar network analyzer feature?
• Is there any way to do this in the Spike software?

Thanks for your help!

• This topic was modified 9 years, 8 months ago by flysight.
• This topic was modified 9 years, 8 months ago by flysight.

Flysight,

You can see the “typical” directivity of the ZFDC-10 rolls off to about 20 dB at around 1900 MHz, and from personal experience I can tell you it is horrible above 2 GHz. We do offer a better directional coupler that goes to 6 GHz.

That being said, there is additional impedance mismatch from the SA44B and TG44A. Adding a 3 to 6 dB pad (SMA fixed attenuator) to the SMA port of each can improve your return loss a great deal, and should get you very close to the performance of the directional coupler.

Since we do not have phase data, we cannot do an open-short-load calibration, but a good directional coupler and a couple of 3-6 dB pads should yield 25-30 dB of directivity.

Thanks, Justin! I hadn’t realized the additional calibration was for vector analyzers only.

I wonder if you could elaborate on the cause of the impedance mismatch at the SA/TG. Looking at the manuals for the SA44B and TG44A, section 2.1 indicates that both SMA connectors have an impedance of 50 ohms.

I learned recently that it’s a good idea to add a small attenuator at one or both instruments, but but I’m afraid I’m not clear on why this improves the result. I would love to understand a bit better.

They are 50 ohms nominal impedance, but their VSWR tends to be higher than a typical spectrum analyzer. Let’s say the VSWR of your SA44B had a typical return loss of 16 dB (probably even worse than this at high frequencies). If you add a 3 dB pad to this, it increases to 22 dB, assuming it’s a good pad with excellent return loss (signal goes through pad, reflects, then through pad again). Add a good 6 dB pad, and return loss would be 28 dB. You could even try 10 dB pads, which would give you maybe 36 dB return loss, but then your signal might be noisy enough that the extra return loss isn’t worth it…

Thanks again for your excellent support, Justin!

Hi…i am a new user here. As per my observation there is additional impedance mismatch from the SA44B and TG44A. Adding a 3 to 6 dB pad to the SMA port of each can improve your return loss a great deal, and should get you very close to the performance of the directional coupler.

[Author]

Posts