4
\$\begingroup\$

I am a hobbyist and I am intersted in testing some materials for their ability of reflectance/absorbance/transmittance for microwave radiation and I want to know if a spectrum analyzer is suitable for this task and how to do this (setup)?

I posted this question in engineering site and I was recomanded to post here.

\$\endgroup\$
4
  • \$\begingroup\$ What frequency? Also It would be better to get the question migrated then double post. Please delete one of the questions. \$\endgroup\$
    – Voltage Spike
    Commented Jan 23 at 20:24
  • \$\begingroup\$ @VoltageSpike if it is possible i will purchase a 4407b and test from the low frequence up to 26.5 ghz its max. \$\endgroup\$
    – Tintin
    Commented Jan 23 at 20:27
  • 1
    \$\begingroup\$ You'd need a source. If you're looking at EMI, a spectrum analyzer with antennas covering various frequencies would work. If you're looking to characterize the material (reflectance/absorbance/transmittance) you'll need to be able to measure phase. \$\endgroup\$
    – Jason
    Commented Jan 23 at 20:54
  • \$\begingroup\$ @Jason when you say phase you are saying signal which is not the case here. Yes its 6 bands L S C X KU K and lower ones \$\endgroup\$
    – Tintin
    Commented Jan 23 at 21:00

2 Answers 2

4
\$\begingroup\$

Shielding effectiveness is tested by measuring the gain between two antennas, then inserting the shield material between them and measuring the gain again. The measurement must be done in an area where backscatter is strongly suppressed or avoided, or such that the shield seals the two antennas from each other (splitting one anechoic chamber in two).

To perform such a test, a spectrum analyzer, tracking generator, RF power amplifier, and pair of antennas would be needed at minimum. Technically the TG isn't required, and an arbitrary source could be swept, or a noise source used even, with the spec in peak hold mode; the TG is merely faster. A VNA could also be used, which is basically the same thing but with phase reference included. (TG+spec can also be described as a SNA, scalar vector analyzer, so this is no accident.)

The amplifier is required to account for antenna factor, which will most likely be pretty significant to begin with, and obviously goes up by even more with the shield included. A very high SNR TG+spec could also be used; just to say that, you need enough signal level to account for losses, so that you have something to measure above the noise floor even in the shield-in-place condition.

Regarding anechoic chambers: they're rarely perfect, and more likely the reflection is on the order of -20dB, significantly attenuated but hardly an "optical black". The reflected wave (plus whatever scattering occurs) bounces back towards the shield and gets a second go at it, and so on and so forth; thus, there will be a small correction factor due to backscatter. This might be ignored entirely as it most likely amounts to a couple dB; or it might be ignored by specifying the test environment, backscatter included; or the chamber might be measured separately, then calibrated out of the result. With scattering a factor, this isn't the easiest thing to calibrate for (the speckle pattern of the scattered radiation will manifest as random peaks and dips -- speckle in the spectrum, as it were).

Looking up standards on these materials, and their measurement, likely sheds some light on the methods to resolve these issues (and, no, afraid I don't have any offhand, sorry).

\$\endgroup\$
5
  • \$\begingroup\$ Tim it is great to know that it is feasibale and i like the way you explained, so i need a spectrum analyzer, an arbitrary generator, in my case 3 antennas one to transmit and two to receive. One will be ~ parallel with the one transmitting to pick up the reflected waves with an angle that i should figure out. And a third facing the transmitting one that way i can measure the R/A/T microwaves. But how to connect those 3 antennas? And for the power amplifier isnt that included in the arbitary generator? Last thing what i had in my mind is to use a box maybe made of lead that the reason \$\endgroup\$
    – Tintin
    Commented Jan 24 at 14:02
  • \$\begingroup\$ why i headed to engineering site in first time. \$\endgroup\$
    – Tintin
    Commented Jan 24 at 14:02
  • \$\begingroup\$ Lead isn't a good RF absorber... at least until you get to x-ray bands, lol. Measuring reflection is feasible but mind that not all reflection will hit the receiving antenna, and the rest will still backscatter; plus there is some lateral leakage between antennas (i.e. they probably don't have 100% perfect directivity) and without a zero-reflection setup it's hard to calibrate that out as well. Further, there is some polarization loss, depending on relative antenna and shield orientation, and shield material. \$\endgroup\$ Commented Jan 25 at 0:46
  • \$\begingroup\$ So Tim it turned out that 26.5 ghz signal generators are expensive so i need your opinion please about this eval board EV-ADF41513SD1Z as source. There is nothing about gain in the datasheet, for the two antennas in parallel they are going to be separate ofcourse. Tim how to connect two antennas to the spectrum? \$\endgroup\$
    – Tintin
    Commented Jan 25 at 12:41
  • \$\begingroup\$ I'm afraid I can't help with that. Perhaps there is something else about your material that is known and doesn't need to be measured directly, I don't know. You asked about measurements, and no, measurements aren't always easy. You might wish to contact a test lab, which will cost ca. 4 digits USD, but will have all the equipment required to perform the measurement. \$\endgroup\$ Commented Jan 25 at 12:53
4
\$\begingroup\$

I never carried out this type of measurement in my previous work, but I do have a text on the subject. Unfortunately in French: https://www.j3ea.org/articles/j3ea/pdf/2016/01/j3ea160002.pdf

https://theses.hal.science/tel-03034434v1/document source:https://theses.hal.science/tel-03034434v1/document

1- The first method

Was explained by Tim Williams

enter image description here source:https://www.j3ea.org/articles/j3ea/pdf/2016/01/j3ea160002.pdf

2- The second method

Consists of using a mode-stirred reverberation chamber. It's a very complex system, and I didn't have access to this type of equipment. It uses the reverberations to make measurements, unlike the anechoic chamber where they are attenuated to make the measurements.

enter image description here source:https://www.j3ea.org/articles/j3ea/pdf/2016/01/j3ea160002.pdf

3- The conclusion is interesting.

The anechoic chamber makes it possible to measure the high attenuation efficiency of shielding. Only two measurements are needed (without shielding for calibration and with shielding for measurement), but the actual efficiency is overestimated, since the main assumption is a plane wave perpendicular to the shielding. The mode-stirred reverberation chamber provides a better measure of efficiency; multiple polarization is more penalizing in terms of shielding efficiency, and the results, particularly at low frequencies, will be closer to the actual shielding efficiency.

enter image description here source: https://www.j3ea.org/articles/j3ea/pdf/2016/01/j3ea160002.pdf

\$\endgroup\$
1
  • \$\begingroup\$ @Tintin, I said "Unfortunately in French" because we're on an English-speaking forum ;) I'm French. \$\endgroup\$
    – Vincent
    Commented Jan 24 at 15:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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