What is meant by: “Short ? calibration of the coil and conversion to “Z: Reflection ??

[Ted Jackson]

In application note AN4974

https://www.st.com/content/ccc/resource/technical/document/application_note/group0/94/99/80/37/c8/08/46/7f/DM00347152/files/DM00347152.pdf/jcr:content/translations/en.DM00347152.pdf

Section 7.3 - Verification of the Q factor in the frequency domain (page 40)

Line 4 is a bit ambiguous:

4. "Short" calibration of the coil and conversion to "Z: Reflection"

Can anyone tell me what is being said there a bit more clearly, please? I'm familiar with the open/short/50ohm calibration procedure of the VNA and some aspects of Smith charts. But that line stumps me.

[Travis Palmer]

Hello Ted,

In OPEN or SHORT response calibration, calibration data are measured by connecting an OPEN or SHORT standard, respectively, to the desired test port. For frequency response, these calibrations effectively eliminate the reflection tracking error from the test setup in a reflection test using that port. It is also possible to carry out isolation calibration with a LOAD standard during OPEN/SHORT response calibration. An isolation calibration will eliminate the directivity error from the test setup in a reflection test using that port

The detailed description, how to perform a "short" calibration should be mentioned in the user-manual of your VNA.

BR Travis.

[Ted Jackson]

Hi BR,

It sounds like you're just telling me to do the normal open/short/50ohm calibration procedure, which is simple enough and which I already mentioned I know how to do and which has nothing (in and of itself) to do with any coil per se.

*Attach the open standard termination and calibrate that.

*Attach the short standard termination and calibrate that.

*Attach the load (50 ohm) standard termination and calibrate that.

*Then I'm ready to do reflection measurements at least, though not transmission measurements which require two additional calibration steps.

I'm asking if "Short calibration of the coil" is any different from that simple procedure, and if so, what steps I'm supposed to do and why use that odd terminology anyway. Sounds like you just told me what I already know, just with a bit more jargon and background. Unfortunately, despite your effort (thanks), you added little clarity for me. I'm sorry. I must be dumb. Me bad. You're not talking to an experienced master chef. You're talking to a novice cook trying to read a recipe at home for the first time. Are you sure you understood my question? Thx

[Travis Palmer]

Hello Ted,

this is the purpose of this community: To clarify questions and help you guys to do awesome designs.

Typically when using a VNA a OSL calibration is done (open-short-load). Some VNA also support additional calibration methods. Our E5061B VNA from Keysight also supports such "short" calibration - besides the OSL calibration. The idea is to calibrate away the characteristics of the pickup coil, and be able to only sense the characteristic of the reader or tag antenna - since the antenna is still connected during the "short" calibration.

A OSL calibration would only calibrate the VNA port + cable. The characteristic of the pick-up coil would overlap with the characteristic of the reader or tag antenna.

I would recommend to consult the user manual of your VNA to see if such a function is available.

BR Travis

[Ted Jackson]

Here's documentation for the E5061B. I can't find a damn thing about short calibrating a coil. You're evidently familiar with it and know it has that capability. Can you find any mention?

https://ena.support.keysight.com/e5061b/manuals/webhelp/eng/

My nanoVNA includes only the OSL calibrations and Isolation and Through calibrations. So, unless the short calibration of the coil is a function of those and/or can be achieved through measurement arithmetic, I suppose my nanoVNA just doesn't support that calibration. But here's documentation for the nanoVNA. I'm tired of endlessly searching:

https://groups.io/g/nanovna-users/wiki#For-Beginners-and-Experienced-Users

[Travis Palmer]

Hello Ted,

The information of my post from the 14th of June was taken from the

Agilent E5061A/E5062A ENA Series RF Network Analyzers User’s Guide Eighth Edition available on keysight.com. Chapter 4, page 91.

Most likely it refers to "Response calibration" mentioned in your ena.support link.

My miniVNA pro only supports "Reflection" and "Transmission" mode. In none of these modes i can do the response calibration or get reasonable readings with using a calibration coil.

As an alternative you can also use a Scope and calculate the Q-Factor by measuring the rise and falling times as explained in chapter 7.2.

BR Travis

[Travis Palmer]

Hello Ted,

i had another look with my colleague.

If you replace the short during OSL calibration by the antenna, you can sense the resonance frequency of the tag.

I can crosscheck the results next week, once I have access to our VNA again.

Below a picture of my setup + measurement result.

 

br Travis

[Ted Jackson]

Hi Travis.

That would seem to make sense. Hence the term "short calibrating the coil". Go figure. Obvious enough in hindsight, I suppose. I look forward to what you find in your crosscheck. I've really appreciated your help on this. Now there's only the conversion to "Z: Reflection" part that remains unclear.

Since you mention oscilloscopes... Know of a nifty little widget scope that has the bandwidth and bit resolution so as to be useful for RFID applications in the 13.56 MHz range, similar in price and size to a nanoVNA or miniVNA that you'd recommend?

I noticed that AN4974 was recently amended/superceded (DM00347152), just some innocuous grammar changes to the language therein (?), some changes to the data in the NFC tuning circuit calculator and the addition of a brief section at the end.

One might also suggest to the author/amender the following:

• Add your simple solution of replacing the short calibration with short calibration coil attached, as you recommend to Section 7.3 - Verification of the Q factor in the frequency domain (page 40).

• The choice of 8 for targeted Q factor was mentioned on p20 of AN4974, but no reason for that number was given therein. Just seems an arbitrary choice as presented. In DM00347152, a new Fig 36 in the new Sec 7.5 shows the 8. It IS explained in UM2042 "User manual Discovery kit for the ST25R3911B high performance HF reader / NFC initiator", Sec 2.2.4 p8 in association with VHBR. A sentence or two or a reference to UM2042 might help someone struggling with AN4974 like I have been.

• I assume these terms are all synonymous: 'system Q', 'target Q', 'matching Q' and 'desired antenna Q factor'. A little more definition, clarity or consistency as to those terms in the text might be helpful, although you needn't do so here.

• Fres (the resonant frequency) appears in the equation for the calculation of Q on page 17, but Fwork is actually used.

Sincere regards,

Ted

[Travis Palmer]

Hello Ted,

I typically use Keysight scopes with quite some memory to observe a few frames in the analog graph.

My software colleagues use a Picoscope 3000 series to capture mixed signal traces (SPI communication + RF field capture.

They also use digital analyzer to only trace the SPI communication.

I have captured two traces: first one is taken with our Keysight VNA:

The second one is captured with the miniVNA pro:

You can use this pictures to compare the two results. Both use the same antenna and same tag on top.

The miniVNA is missing the Z-conversion. But i assume you can export the S-parameter and do the conversion externally.

br Travis

[nmeunier]

Hello Travis,

I also have trouble measuring the Q factor of an antenna with a MiniVNA Pro. In your last message, you provide 2 screenshots for which I would like some precisions :

• The first one is, according to the display, S22. So when does the conversion to Z:Reflection mentioned in the AN4974 steps in ? I also assume that S22 is equivalent to S11 in your setup, is it true ?
• The second one is the Return Loss which is related to the S11. You advise to convert it to Z-parameter, but to do so you need S21, S12 and S22 (see formula below).

 

This last point on S21 is true for both. According to the Agilent E5061A/E5062A ENA Series RF Network Analyzers User’s Guide, Zreflection is a function of Sab (see image below), which is S21 to me.

My question is : did you perform a transmission measure ? How ?

Finally, isn't the measure of Q on time domain with a scope more precise ?

Thanks for your answers,

Nicolas Meunier.

[Travis Palmer]

Hello Nicolas,

Yes the conversion with the above formula is do able. Anyhow our VNA has a build in conversion function which does the job.

Yes Chapter 7.2 of the above mentioned application note AN4974 guides you, how to calculate the q-factor from rise and fall time between 90% and 5% of the signal.

BR Travis