Say you have an antenna (2.4 GHz) for which you want to know the return loss using a VNA. The connection between your VNA port 1 and the antenna is through a coax cable (e.g., Murata). The cable's own return loss is specified at 14 dB (VSWR \$\approx 1.5\$).

When I measure the return loss with the VNA for the whole setup (cable + antenna), I get about 25 dB, but I am unsure of how the cable's own return loss bias this measurement one way or another. How do I account for the cable's return loss? How would I remove the cable's own return loss such that I know the actual return loss from the antenna on its own?

I am aware the insertion loss (IL) of the cable will bias the measurement by 2*IL (because of the round-trip). So if the cable has an IL = 1dB, the return loss for the whole setup is going to seem better than reality by 2 dB. I'm just curious about the cable's return loss impact on this.

  • 3
    \$\begingroup\$ no measurement cable ever would have so terrible specs. Get a proper connector, get a proper, shorter measurement cable. (this doesn't invalidate your question at all! It's just to say: don't make your measurement unnecessarily bad; a better cable would be cheap, especially at that benign frequency.) \$\endgroup\$ Jul 18 '20 at 21:24
  • \$\begingroup\$ @MarcusMüller hey! That is actually the worst case VSWR. This is the actual cable: datasheet.octopart.com/… \$\endgroup\$
    – Big6
    Jul 18 '20 at 21:28
  • 2
    \$\begingroup\$ Most VNAs offer calibration routines to eliminate the effect of the cables used for measurements. In most cases you need a 50 Ohm termination, a short and a through to perform the calibration. \$\endgroup\$
    – w7sbc
    Jul 18 '20 at 21:36
  • \$\begingroup\$ @Big6 I don't see how you'd be seeing 14 dB of loss with that cable? VSWR of 1.5 would still be pretty bad for a cable that is supposed to fulfill a specified impedance. \$\endgroup\$ Jul 18 '20 at 21:47
  • 1
    \$\begingroup\$ @MarcusMüller Return loss (match) \$\neq\$ insertion loss. 14dB of return loss isn't very bad, and neither is an insertion loss of 1dB. It's not metrology grade with those specs at 2.4GHz, but it's not terrible. \$\endgroup\$
    – Shamtam
    Jul 18 '20 at 21:58

TL;DR - The cable won't have any affect on the measurement if you take the measurement properly.

If you have a VNA, you should be calibrating it before taking a measurement. Typically done with a SOLT (Short, Open, Load, Thru) calibration kit for coaxial cables (and just SOL for a 1-port measurement). The VNA measures the response of the standards in the calibration kit, and uses them to determine characteristics of the "stuff" between the VNA and the device you'll be testing. These characteristics are typically constants determined by various measurements and used in a 12-term error model. Once the calibration measurements are done, the calculated error terms allow the VNA to mathematically remove the effects of your connectors, cables, etc., from the measurement of your actual device.

With a modern VNA with incredibly high dynamic range (on the order of 90+ dB!), even a "terrible" cable with 1dB loss at 2.4GHz will have no effect on an accurate measurement

Now, of course, I agree with @Marcus in the comments to just get a better cable. But that said, you'll still need to calibrate to get a proper measurement. See edit below.

You can search using keywords like "12-term error model/correction" and "VNA calibration" to read more about this. It's a very fundamental and extremely important concept in microwave metrology. Even if you have a VNA that doesn't allow calibration (which I'd be.... very surprised if it was the case), you can take measurements of standards and calculate your own error terms, then measure your device and do the math to "de-embed" the raw data down to the reference plane of your device. Of course, you'll need to get a calibration kit if you don't already have one, but a VNA is all but useless without a calibration kit.

Edit: On further discussion in the comments, I now understand what you're trying to do. The "cable" you linked in the original question is probably more accurately considered a probe specifically intended for in-situ measurement of your circuit. Your connector is a switched connector that provides isolation from the RF circuit and the antenna when the probe is connected to it, facilitating in-situ measurement. It seems Murata also provides Calibration Adapters which adapt the board connector to SMA, which would allow you to use your SOL standards to be connected to the connector end of the probe and calibrate to that reference plane. Unfortunately, it looks like that would still introduce a little bit of error (since the calibration would be at the SMA reference plane, not at the probe connector), but that's likely a non-issue if you're only interested in the magnitude of \$S_{11}\$ (i.e. the VSWR) and not the phase.

  • \$\begingroup\$ Thanks for the answer. I do have the standards and calibrated the VNA with those. Now, to connect to the antenna on the PCB, I used the cable. So I did not calibrate at the end of the cable if that is what you mean, I actually calibrated at the VNA ports with the short, load, and open standards. I can do port extensions at the ports but do you mean I need to calibrate with the cable included? \$\endgroup\$
    – Big6
    Jul 18 '20 at 21:43
  • \$\begingroup\$ Yes, your calibration needs to be at the cables. Calibrating at the VNA ports doesn't do you any good (and in fact, likely looks very similar to an uncalibrated measurement, since the source and receivers in the VNA are likely "by default" measuring at the reference plane of the test ports on the front of the instrument). Look up the concept of a "reference plane," if you're not already familiar. It's crucial to understand your reference planes when making a measurement for modeling purposes to know what is and isn't accounted for in your measurement. \$\endgroup\$
    – Shamtam
    Jul 18 '20 at 21:44
  • \$\begingroup\$ You could also use port extensions, but for your purpose, that just sounds like a bad idea that will introduce more uncertainty in the measurement for no good reason. Port extensions are useful in certain cases when it's impossible to calibrate at the required reference plane for one reason or another. \$\endgroup\$
    – Shamtam
    Jul 18 '20 at 21:46
  • \$\begingroup\$ Thanks. I guess the problem with this approach is that the standards are SMA connectors to the VNA. In order to calibrate with the cable, and to the PCB at the location of the antenna, I would need to create a 'proper' short, open, and a load on the PCB itself at the point where the cable plugs in. Not the easiest thing to do since it's hard to re-create true shorts or opens at those frequencies, but I will give that a try. \$\endgroup\$
    – Big6
    Jul 18 '20 at 21:47
  • \$\begingroup\$ Possible XY challenge: why do you need to know the antenna's characteristics independent of the cable? Presumably you need the cable to connect the antenna to whatever it will be used for, so consider the cable as part of the antenna? \$\endgroup\$
    – Shamtam
    Jul 18 '20 at 21:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.