1
\$\begingroup\$

I need to obtain S11 and S21 to calculate loss tangent of polymer(PVDF) in C-band. I have an access to vector network analyzer, to use it I must somehow attach 2 SMA connectors to polymer and here starts the hard part. Polymer film is roughly 0.1 mm thick with aluminium coating on both sides(around 100 nm thick). Melting temperature of PVDF is 175 degrees C so I can't solder something directly to film. At first I thought that I can just take PCB, solder one SMA to it and glue film to board with second SMA glued to top side of film (with conducting glue). But it wont work correctly because of amount of places where waves will reflect because of glue. Some additional info you should know:

  1. I can't use cavity perturbation method.
  2. I only have this film. I can't obtain this polymer in other form.
  3. I can remove aluminium from film where needed by using PCB marker and ferric chloride.
  4. I can cut the film, I think square shape will be the best solution to ease further calculations.

So you can imagine it as square thin film capacitor. Any ideas?

\$\endgroup\$
  • 1
    \$\begingroup\$ Can you etch a microstrip line in the aluminum (line width would probably need to be somewhere around 0.1 mm, might be tricky)? \$\endgroup\$ – The Photon Dec 17 '16 at 2:40
  • 1
    \$\begingroup\$ Second question, can you get hold of some indium solder? \$\endgroup\$ – The Photon Dec 17 '16 at 2:57
  • \$\begingroup\$ Why can't you use the cavity perturbation method? If the aluminum is removed, it should work the best. That's exactly what it is for. \$\endgroup\$ – Tom Anderson Dec 17 '16 at 5:51
  • \$\begingroup\$ @ThePhoton I don't think it is possible to do with equipment I have, I thought about 50 Ohm strip if that's what you mean. Indium solder is a good idea, I will ask around \$\endgroup\$ – Autiarii Dec 17 '16 at 11:58
  • \$\begingroup\$ @TomAnderson I can't use it for several reasons. This method is mostly used to find properties of dielectric which volume is comparable to volume of rectangular resonator and volume of my film is very very small. I've read someone used this method for films, but he had thicker substrate with film on it. I think I will try this method if everything else will fail. \$\endgroup\$ – Autiarii Dec 17 '16 at 12:17
1
\$\begingroup\$

The most precise way to measure the plastic material properties is to remove all the aluminum and measure the plastic with a split cylinder resonator such as the 85072A 10-GHz Split Cylinder Resonator. (Full disclosure: I work for the company that makes this particular resonator, but I am not speaking for them here.)

The split cylinder measures the dielectric constant and loss of the plastic at many spot microwave frequencies. This is much more accurate than measuring with the aluminum.

To measure with the aluminum coating, it will be important to take into account the effect of the conductivity and roughness of the aluminum. If the aluminum is pure and smooth, the material properties of aluminum can be taken from the values at Matweb. If the plastic is porous, and the aluminum coating process puts aluminum into the pores, in a close-up cross-section analysis it looks like the aluminum has teeth. This type of roughness will cause errors. Cross section analysis is shown in this video.

When using the aluminum coating, I recommend etching a Beatty Structure, which is described in "Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties". To improve the accuracy of the etching process, take advantage of the low-cost stencils available from OSH Stencils or elsewhere.

To connect to the ends of the Beatty Structure, attach the SMA connectors with conductive epoxy. The discontinuity at the connectors will be calibrated out as part of measurement process using the Network Analyzer.

\$\endgroup\$
0
\$\begingroup\$

Make a capacitor from two pieces of single sided PCB that has a surface area wide enough to allow the double sided aluminium coated polymer film to be inserted into. The PCB capacitor gap isn't too important of course but the closer you make the plates the more sensitivity you will achieve when measuring: -

  • The extra capacitance from inserting the polymer
  • The extra losses from inserting the polymer

It might be sensible to make the PCB capacitor have "vertical plates" then, it should be relatively straight forward to "dangle" the polymer and lower it in between the PCB plates. The PCB capacitor plates should, all-round be a little bigger than the polymer although some experimentation might prove otherwise.

Attach a connector to the PCB plates and you should be able to get a fairly decent idea about the polymer's performance. One of the PCB plates can be earthed to make things a little more straight forward if the measurement device produces an unbalanced signal. It might also be a good idea to have a third earthed-plate that is about equidistant from the "hot" plate of the PCB capacitor in order to cut-down on extraneous fields and influences.

\$\endgroup\$
0
\$\begingroup\$

I have figured out how to measure what I need, it is a variation of @Andy 's answer adapted to SHF range. The idea is to take 2 substrates that have stable parameters in microwave (ceramics or special pcb), both substrates have metal coating on 1 side, the second side of each substrate has 50 Ohm strip with a special square plate of known size at the end and SMA connected to it in the beginning. Substrates connect to each other with 4 bolts.

Then I can measure S parameters with and without film like @Andy suggested.

Here I have a quick 3d model for visualisation, golden are SMA connectors and brown is copper. measuring model

\$\endgroup\$

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.