LUP Student Papers

Probing very close to antennas for manufacturing defect detection in a miniature shielded enclosure

• Mark

Abstract

This thesis investigates probing antennas within a small shielded test enclosure for
IoT, Internet of Things, products for the detection of manufacturing deviations.
The small enclosure forces the transmitter and receiver to be in the near-field.
Shielding of the enclosure additionally generates a multipath environment, which
becomes sensitive to displacements of objects such as wires or metallic holding
pins, which are essential for communication with the DUT and mechanical stability. In this work, a few variants of PCB probes were designed, constructed, and
tested to be placed 10 mm from the transmitter. The frequency range of interest
was initially 0.860 GHz to 6 GHz. After research, detuning of the transmitter was
considered... (More)

This thesis investigates probing antennas within a small shielded test enclosure for
IoT, Internet of Things, products for the detection of manufacturing deviations.
The small enclosure forces the transmitter and receiver to be in the near-field.
Shielding of the enclosure additionally generates a multipath environment, which
becomes sensitive to displacements of objects such as wires or metallic holding
pins, which are essential for communication with the DUT and mechanical stability. In this work, a few variants of PCB probes were designed, constructed, and
tested to be placed 10 mm from the transmitter. The frequency range of interest
was initially 0.860 GHz to 6 GHz. After research, detuning of the transmitter was
considered when designing the probes. The probe designs were open-ended flat
twin-wires designed using Altair FEKO, which in outside measurements reached
S21 values between -30 dB and -16 dB for the frequency band 2.1, 2.8 GHz at 10
mm range. The reflections were simulated using COMSOL Multiphysics to observe
the signal behavior with object displacements. The signal stability was also tested,
along with the effects of absorbers. Measurement results show that the practical
placement of 10 mm from the transmitter causes insignificant detuning for signals
between 2.1 GHz and 2.8 GHz. However, placing the transmitter in a shielded
enclosure causes major detuning. The effects of detuning may also be reflected
within the measurement results. Using a more mismatched probe, the receiver
becomes less affected by detuning from the enclosure. The trade-off, as expected,
becomes a lower received signal level. Simulations showed significant signal attenuation when a large object, representing a PCB, for example, was placed within
the chamber. During measurements, the trend was not visible; however, the S21
response was changed, signifying that different frequencies interacted differently
with the introduced objects. For this work, it has been simulated and verified
that E-field probes close to transmitters can be used in a small test enclosure
environment; however, there are still challenges regarding signal level stability to
be further investigated.

Keywords: Electrical field probe, PCB probe, Detuning, Near-field measurement,
Resonance cavity, Multipath environment (Less)

Popular Abstract

Manufacturers of electronics used in devices for WiFi for need to test their newly
made products to prevent faulty electronics from being released. Initial testing
may therefore be done at the factories to save transportation costs. For electronics with radios and antennas, tests are often conducted within specialized rooms or
metallic boxes, which block the signals from escaping and interfering in the testing space. However, signal isolation using metal will cause reflections, like light on
mirrors. The reflections become problematic when measuring the antennas. During adjustments or reparation of the boxes, if wires or other objects are slightly
moved, the signal will not reflect in exactly the same way as before. These types
of... (More)

Manufacturers of electronics used in devices for WiFi for need to test their newly
made products to prevent faulty electronics from being released. Initial testing
may therefore be done at the factories to save transportation costs. For electronics with radios and antennas, tests are often conducted within specialized rooms or
metallic boxes, which block the signals from escaping and interfering in the testing space. However, signal isolation using metal will cause reflections, like light on
mirrors. The reflections become problematic when measuring the antennas. During adjustments or reparation of the boxes, if wires or other objects are slightly
moved, the signal will not reflect in exactly the same way as before. These types
of changes affect the received signal level. For that reason, it becomes difficult
for the receiver antenna to pick up a consistent signal level over time from the
transmitter.
Smaller testing boxes maximize facility space but challenges arise because the
receivers have to be placed very close to the transmitters. Conventionally, they
are located far apart. The goal of the project was to design and test a compact
receiver antenna for use in a small testing box and very close to the source to see
whether the picked up signal would become more consistent. Another goal of the
project was to gain a greater understanding of how the signals reflect and behave
in the box. Additionally, the study aimed to determine what could be done to
make the signal even less sensitive to object movements within the box.
Limited existing research suggests that unpredictable phenomena occur when an
antenna is placed very close to the signal source. Let’s imagine that you have an
antenna that picks up signals from two radio channels, A and B. A source sending
radio signals was only designed for channel A. So, there is music on channel A
and silence on channel B. If the antenna is placed extremely close to the signal
source, an unexpected phenomenon called detuning may occur. Channel A may
become quieter while channel B may begin to play music instead. This happens
despite the fact that the source was only designed to send signals on channel A.
In reality however, there are thousands of radio channels and predicting at which
channels the source will start sending music is difficult. Therefore, detuning was
considered when choosing the receiver antenna design. The final prototype was an open ended twin wire.
During measurements, it was found that for the designed receiver antenna, when
it was placed at 1 cm from the source which was the closest practical distance, the
source remained unaffected. However, it was revealed that the box itself greatly
detuned the source. How the received signal changes with object movements also
varies unpredictably with frequency. Therefore, it is difficult to determine how
much amplification may be needed at certain frequencies for a chosen power detector. The reason for this is probably that different frequencies reflect differently
within the box. The complexity of the reflections also increases with frequency.
Simulations showed that the signal should become less sensitive to object movements within the box when signal absorbing materials are added to the walls.
Absorbers may also camouflage the box from the source preventing detuning. The
project provided greater understanding of the complexities with measurements of
antennas in the test boxes. (Less)