From: "Ed Price"
Subject: Re: Cheap, home made TEM cell?
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
Date: Sat, 23 Nov 2002 11:44:54 GMT
NNTP-Posting-Date: Sat, 23 Nov 2002 06:44:54 EST
Organization: Cox Communications
"Tony (remove "_" from email address)" wrote in
> For a long time I've contemplated making a TEM cell for precompliance
> testing (or maybe that should be pre-precompliance testing). For my own
> didn't need to worry about infringing patents etc, so I was relatively
> copy any commercial product.
> I liked the GTEM design with its simple 50 ohm feed and generally cheap
> construction, but it seemed that the 50 ohm impedance came from the high
> capacitance from septum to shell, whereas I'd get more V/m from a higher
> impedance design and a step-up transformer.
> Enter the Eurotem 2 - compact, balanced 200 ohm load and a superior test
> But the enclosure needs to be fully lined with expensive ferrite tiles.
> Another approach seems to be to combine the two - a long cell similar to
> GTEM (with resistive septum loading and absorber across the mouth), but
> 1 more slowly tapered (both to reduce problem reflections if I don't get
> exactly right, and to maximise the test space for a given mouth area),
> 2 containg four striplines like the Eurotem; narrow and further in from
> shell, to get the impedance back to 200 ohms, to get the most from a small
> linear amplifier.
> 3 Also like the Eurotem, line the mouth with ferrite tiles instead of
> to maximise the test volume (still expensive, but much cheaper than lining
> whole thing) and to extend the bottom frequency limit, albeit with
> lower performance.
> Which leads to the questions - can anyone direct me to shareware or
> that can calculate the impedance of a twinaxial transmission line (the
> quad arrangement is actually symmetric) in which the septa are striplines
> degree angles?
> Also, does anyone know where I can buy cheap ferrite tiles?
> Finally, have I got it all wrong?
FWIW, as I don't have a TEM or GTEM myself.
The impedance of the cell is defined by the geometry of the cell and the
conductivity of the cell walls, septum and insulators. There have been a
number of designs for parallel plate cells and stripline antennas. Remember
that the GTEM cell is an extrapolation of the TEM (Crawford) cell, which
itself grows out of classic waveguide design. I think that 377 ohms must be
the upper impedance limit for a cell, but I'm not smart enough to remember
The classic TEM cell has an attractive simplicity, since a tapered impedance
matching section is used on both sides. That lets you start with a 50 ohm
generator, transition to a larger volume for exposure, and transition back
to a 50 ohm load. If you already have the test equipment, the cell becomes a
sheet-metal exercise. Of course, you are aware of the TEM's frequency &
specimen size limitations, and the GTEM was the best response to those
Unfortunately, the GTEM added complexity, with its off-center septum and its
termination technique. Everybody who makes a GTEM has their own ideas about
the best geometry for the resistive termination elements and the RF absorber
material. If you copy an existing GTEM, you will at least have a decent
head-start for your experimentation.
The Eurotem looks pretty interesting; maybe you might want to build that
first. Looking at the design, I keep wondering about how all those ferrite
absorbers keep from being power hogs. If you have absorbent material lining
the walls of a TEM or GTEM, you will waste power into the absorbers. Maybe
the Eurotem needs those absorbers to "smooth" the field gradient, despite
Cheap ferrite tiles? Sorry, but the tile fairy doesn't exist!
No, you're certainly not all wrong. But it takes a lot of imagination to
finesse the electromagnetic laws, and break-through new antenna designs are
rare. I do think building a copy of a Eurotem is more risky for a first
project, but that's the best way to learn about it. Maybe you should try to
build a very small Eurotem style system, since you could get away with the
least amount of absorber tile.
Good luck, and publish your experiences.