Page 36 - ChipScale_Nov-Dec_2020-digital

setup is very straightforward and far-field OTA measurement setup and a
provides an easy way to correlate with a radiating near-field OTA measurement
5G-compliant bench measurement setup. setup might be significant.
Although the far-field OTA measurement Note that in the near-field region it
setup is excellent for characterization, is no longer possible to do a standard
as discussed in [1], it presents major calibration using the measurement
challenges for test-cell integration in a antenna gain and the path loss as in the
multisite high-volume testing setup due far-field case. Although it is possible to
to its mechanical requirements. make a transformation from the radiating
near-field region to the far-field region
Results with a radiating near-field [6], this requires a complete spherical
OTA measurement setup measurement of power and phase in the
In the radiating near-field setup radiating near field, which is not the case
(Figure 5b), a dual-polarized patch in this OTA measurement setup.
measurement antenna is used on the What the above discussion indicates
socket (as shown in Figure 4 of [1]). is that for OTA testing in the radiating
This measurement antenna is placed at near-field region, the best calibration
a distance of 11mm from the AiP DUT, approach is to use a golden device
Figure 6: Far-field distance computation for the so it is within the near-field region as approach. Also, one should not expect
antenna array. for all tests in the radiating near field
shown in the Figure 6 computations.
antenna to DUT distance is fixed, we The 11mm distance was not selected at to have a straightforward numerical
also know the path-loss through air. random. It was selected based on the correlation between the radiating
These values are then used to calibrate standing-wave effect that is present near-field results and far-field results.
the measured results. Figure 7 shows on any radiating near-field OTA setup What one should expect and try to
the measured error vector magnitude as described in [1,5]. Figure 8 shows achieve is to obtain a set of testing
(EVM) and corresponding constellation the EVM and constellation diagram criteria in the radiating near field that
diagram measurement of the AiP DUT measurement of the AiP DUT in the provides pass/fail correlation with the
using a 28GHz 5G QAM64 waveform exact same conditions as for the far-field far-field measurements.
with 100MHz modulation bandwidth. measurements shown in Figure 7. F i g u r e 9 a s ho w s t w o o t h e r
This measurement is performed with In Figure 8 we show a measured measurement examples done in the
the entire antenna array transmitting 2.76% EVM value for the radiating radiating near-field OTA setup—the
and pointing in the horizontal direction near-field, while the far-field EVM first one is a linearity measurement
to the measurement antenna. Only the measured result shown in Figure 7 was of the phase programming on each
H-polarization field is measured (see 2.74%. Although in this example the antenna on the antenna array (that is,
Figure 3). As previously mentioned, EVM results correlate, for a different the phase-delay elements inside the
this AiP device is intended to be a AiP module with a different antenna Anokiwave evaluation board IC). This
demonstration vehicle; because of its array or a different design of the is a serial measurement where only
simplistic design, one cannot expect measurement antenna and its distance one antenna is measured at each time
good performance. This measurement to the DUT, the difference between a with the others turned off. The second

Figure 7: Measured transmitted far-field EVM and constellation diagram of a Figure 8: Measured transmitted radiating near-field EVM and constellation diagram
28GHz 5G QAM64 waveform (100MHz modulation bandwidth) for H-polarization. of a 28GHz 5G QAM64 waveform (100MHz modulation bandwidth) for H-polarization.

34
34 Chip Scale Review November • December • 2020 [ChipScaleReview.com]