EVK-M91 - User guide
10 Common evaluation pitfalls
• Parameters may have the same name but a different definition. GNSS receivers may have a
similar size, price and power consumption but can still have different functionalities (e.g. no
support for passive antennas, different temperature range). Also, the definitions of Hot, Warm,
and Cold Start times may differ between suppliers.
• Verify design-critical parameters; do not base a decision on unconfirmed numbers from
datasheets.
• Try to use identical or at least similar settings when comparing the GNSS performance of
different receivers.
• Data, which has not been recorded at the same time and the same place, should not be
compared. The satellite constellation, the number of visible satellites and the sky view might
have been different.
• Do not compare momentary measurements. GNSS is a non-deterministic system. The satellite
constellation changes constantly. Atmospheric effects (i.e. dawn and dusk) have an impact
on signal travel time. The position of the GNSS receiver is typically not the same between two
tests. Comparative tests should therefore be conducted in parallel by using one antenna and a
signal splitter; statistical tests shall be run for 24 hours.
• Monitor the carrier-to-noise-ratio (C/N0). The average C/N0 of the high elevation satellites
should be between 40 dBHz and about 50 dBHz. A low C/N0 will result in a prolonged TTFF and
more position drift.
• When comparing side-by-side receivers, make sure that all receivers have the same signal
levels. The best way to achieve this is by using a signal splitter. Comparing results measured
with different antenna types (with different sensitivity) will lead to incorrect conclusions.
• Try to feed the same signal to all receivers in parallel (i.e. through a splitter); the receivers
will otherwise not have the same sky view. Even small differences can have an impact on the
speed, accuracy, and power consumption. One additional satellite can lead to a lower dilution of
precision (DOP), less position drift, and lower power consumption.
• When doing reacquisition tests, cover the antenna in order to block the sky view. Do not
unplug the antenna since the u-blox M9 positioning technology continuously performs a noise
calibration on idle channels.
• The Interface Switch is used to select “UART/I2C” or “SPI” port. If the receiver is set to “UART/
I2C” mode and configuration is saved by sending the CFG-CFG save command, all input and
output protocols for SPI are disabled. Powering up the EVK in “SPI” mode will result in the
receiver starting with all SPI input and output protocols disabled. Since SPI input has also been
disabled, it is impossible to recover using the SPI interface. Powering up in “UART/I2C” mode
and sending a CFG-CFG clear command restores the SPI port defaults, which makes SPI usable
again.
UBX-19056858 - R02
10 Common
evaluation pitfalls
Page 20 of 24
