Design Notes and Construction Details

This TEM cell is based on the following publications:

• "Do It Yourself Sheet Metal Open TEM Cell" — Giacomo Giannetti, Norbert Leder, Thomas Weiler, Christian Spindelberger, and Holger Arthaber
• "Simple Semianalytical Septum Design for Improved Matching in Open TEM Cells" — Giacomo Giannetti, Christian Spindelberger, and Holger Arthaber
• "Increasing the Test Volume of Open TEM Cells by Using an Asymmetric Design" — Christian Spindelberger, Giacomo Giannetti, and Holger Arthaber, Technische Universität Wien, Vienna, Austria

Since these publications do not include all design dimensions, particularly those related to the modified non-linear septum transition, this portion of the design was calculated from the information provided in the publications and may differ slightly from the original implementation.

In addition, the PCBs used to connect the septum are my own design. Compared to the original design, they include additional copper polygons and via stitching in the area between the two clamping points to improve electrical continuity.

Construction Details

To minimize screw protrusion at the septum connection, brass M3 threaded inserts were installed directly into the septum and machined flush with the septum thickness. Low-profile screws were then used for assembly.

A possible future improvement would be the use of countersunk (flat-head) screws on the septum side together with PEM-style self-clinching nuts or soldered nuts on the PCB side. This arrangement would completely eliminate protrusions between the septum and the upper conductor.

Despite the existing connection geometry, VNA measurements do not show any significant reflections or losses attributable to these small protrusions.

Mechanical Support

The design includes a plywood base to facilitate handling of the TEM cell and provide additional mechanical support near the connector ends.

One side of the base is specifically designed to accommodate a standard 4" × 4" laboratory jack, allowing support of heavy RF terminations. During testing, the TEM cell was used with both 25 W and 100 W terminations.

Electrical Performance

All measured electrical parameters meet or exceed those reported for the original design.

I do not possess a sufficiently small field probe to directly verify field uniformity within the test volume. However, since all other measured parameters closely match those reported in the referenced publications, it is reasonable to assume that the field distribution and usable test volume are comparable to those of the original design.

Internal Supports

The original design utilized custom 3D-printed supports intended to further reduce their influence on cell performance. However, testing showed that MJF-printed Nylon 12 exhibits dielectric properties significantly different from those of machined Nylon 12, resulting in measurable degradation of cell performance.

The final design therefore uses 1-inch PTFE support columns similar to those employed in the original publication.

The bill of materials also includes four M4 × 10 mm PTFE screws. These were intended to secure the support columns to the septum. Because my laboratory setup allows easy handling of the TEM cell without additional fastening, I chose not to drill the septum and therefore did not install these screws.

Manufacturing

• Aluminum sheet-metal components were manufactured by OshCut.
• Machined components and PCBs were manufactured by JLCPCB.
• All plywood components and PTFE support columns were fabricated by me.

Measurements

The TEM cell parameters were measured using an HP 8753C Vector Network Analyzer.

The software used to visualize the time-domain transform does not correctly display the horizontal axis. As a result, the plots show a frequency scale of 0 MHz instead of the actual time scale in picoseconds.

To provide a more meaningful presentation of the results, the exported S1P files were processed separately and used to generate...