Test Definitions#

To communicate with the device under test (DUT) and control various device test modes by automated test software, Silicon Labs provides embedded software applications to allow for both serial communication as well as Over the Air (OTA) communication to automated test software. Both of these interfaces enable configuring a device for receive or transmit modes, turning on peripherals (if applicable), reading Analog to Digital Conversion (ADC) pins on the micro (if applicable), putting the radio and/or microprocessor to sleep, and similar control functions. See Embedded Software Tools for a description of these applications.

The tests can be divided into different types of tests: RF testing, DC testing, and peripheral testing.

  • RF testing is any test specific to the operation and functionality of the radio (for example, transmitting and receiving packets).

  • DC testing is any test related to the voltage and current characteristics of the device or board (for example, active and sleep currents).

  • Peripheral testing is any test not specific to RF or DC, like a sensor or an external crystal.

The following sections describe the tests that make up the potential suite of DUT testing.

Serial Communication Test#

The Serial Communication Test verifies valid serial communication with the DUT before testing. This is a basic check that the device has been programmed correctly. If no communication is present, the DUT fails this test and does not proceed with further testing. This test is important because, if there is no serial communication with the DUT, there is no way to interface with the DUT to put the device into test mode or send commands in a standalone test application.

Supply Current Test#

The Supply Current Test verifies that current consumption is valid for each mode of operation for the DUT. The modes of operation are set through the serial interface and include transmit mode, receive mode, and sleep modes (radio sleep and deep sleep). If there is excessive current draw, the DUT fails this test and does not proceed with further testing. This test is especially important for devices that will be used in battery-operated applications, as these measurements are an effective predictor of battery life.

Guidelines for measuring current consumption are provided in AN969: Measuring Power Consumption on Wireless Gecko Devices.

AN1082: EFR32 Transmit and Receive Current Measurements provides some more insight into the current measurement procedure and configurations.

Quick Verify of Transmit and Receive Test#

The Quick Verify of Transmit and Receive Test quickly verifies that the DUT transmits valid packets to the Reference Node and receives valid packets from the signal generator. This can be tested on any single channel in the available frequency band. If either of these checks fails, the DUT fails this test and does not proceed with further testing. This test identifies hardware that does not require full characterization testing due to a major manufacturing defect.

Transmitter Tests#

Transmit Power Test#

The Transmit Power Test verifies that the power level of the transmitter is at the appropriate level and within a specified range. The power output is measured with the power meter at multiple power levels to confirm power output accuracy at various coded settings. The serial command interface can include a function that enables continuous waveform (CW) or unmodulated tone to be transmitted for ease of measuring these power levels. Silicon Labs recommends that this test be performed over a subset of the frequency band to record trends in power output versus frequency.

The transmit power output can be measured with a spectrum analyzer or a power meter.

Transmit Frequency Offset Test#

The Transmit Frequency Offset Test verifies the crystal accuracy and valid transmission frequency offset of the DUT. The CW tone is again used for this transmission and the crystal capacitor value is set using software.

The EFR32 38.4MHz crystal does not require external loading caps as there is a tunable capacitor bank in the EFR32 that can be used instead. Different capacitor values can be written to registers to observe the corresponding frequency offset, and the CTUNE manufacturing token is used to store the value for use by the application. The optimal CTUNE value for the DUT can be determined in the design characterization stage by sweeping all CTUNE values at the transmit frequency and measuring the frequency offset with a spectrum analyzer. This value can then be programmed as the CTUNE manufacturing token in volume manufacturing, or each board can be tested for optimal CTUNE value, using the value from the design verification stage as a starting point. Note that the measured frequency offset will differ depending on the frequency band, so Silicon Labs recommends determining CTUNE for each frequency band needed for the device.

Additionally, the test can be performed over a subset of the frequency band to record trends versus frequency.

For information on how to set the CTUNE token, please refer to the applicable software tool application note described in section Embedded Software Tools.

Transmit Error Vector Magnitude Test#

The Transmit Error Vector Magnitude (EVM) test verifies that the device’s EVM is within specified limits. The EVM is measured with a spectrum analyzer. Either a transmit packet or transmit stream command is used for this transmission, as the spectrum itself is analyzed. Silicon Labs recommends that this test be performed over a subset of the frequency band to record trends versus frequency.

Some spectrum analyzers are able to measure EVM as a standalone instrument, while other spectrum analyzers require a PC software tool to provide EVM measurement details.

Transmit Sweep Test#

The Transmit Sweep Test verifies transmission of valid packets from the DUT to the Reference Node at all channels or a subset of channels across the frequency band. The Reference Node is put into receive mode while the DUT transmits 100 packets to the Reference Node for each channel, with an attenuation between nodes that translates to a strong signal, approximately 60 dB attenuation between devices. Please refer to the specific radio chip data sheet for more information.

Packet success rate is measured at each channel. The packet success rate percentage is defined as the number of packets received divided by the number of packets transmitted and then multiplied by 100. For the Transmit Sweep Test, anything below 100% packet success rate is flagged as a failure, as this test is conducted at a signal level where all packets should be received. This test confirms that there are no frequency-dependent issues with transmit mode.

Spurious Emissions Test#

The Spurious Emissions test verifies that the transmitter’s unwanted emissions outside the channel bandwidth that result from the modulation process and non-linearity are within the applicable limits set by regulatory bodies. A transmit tone from the DUT can be used to measure the out-of-band emission levels in a spectrum analyzer.

Receiver Tests#

Receive Sweep Test#

The Receive Sweep Test is similar to the Transmit Sweep Test. It verifies reception of valid packets at the DUT from the signal generator at all channels or a subset of channels across the frequency band and at two receiver input power levels (a strong signal level, approximately -50 dBm, and a level closer to the edge of sensitivity performance, approximately -90 dBm).

The DUT is put into receive mode while the signal generator transmits 100 packets for each channel. Packet success rate is measured at each channel/receive input level. Any packets missed at the strong signal level are considered a failure, while the failure threshold at the lower input level can be at a lower percentage, depending on the expected sensitivity of the radio. Please refer to the data sheet of the radio chip for details related to receive sensitivity. This test should be performed over the full operating band to record trends versus frequency.

Silicon Labs recommends using a signal generator to transmit packets to a DUT, as the signal generator allows for easily configuring the receive input power level at the DUT.

Receive Sensitivity Test#

The Receive Sensitivity Test determines the receiver sensitivity of the DUT by measuring the input power level at 0.1% BER (Bit Error Rate) or 1% PER (Packet Error Rate) depending on the radio chip configuration. Please refer to the data sheet for the specific radio chip information on the sensitivity requirements. To measure PER, the DUT is placed in receive mode with 1,000 valid packets being sent from the signal generator for each channel. The power level should begin at some level before the 1% Packet Error Rate (PER) threshold. The PER is measured until the receiver input power level corresponding to 1% PER is determined. This test should be performed over a subset of the operating band to record trends versus frequency. During the characterization testing phase Silicon Labs recommends that the actual sensitivity level be determined, while at high volumes the receive sensitivity specification can be set as the low limit and be used as a single power level for ease of testing.

Similarly, to measure BER, the DUT is placed in receive mode and the signal generator sends valid bits to the DUT for each channel. The BER is measured until the receiver input power level corresponds to 0.1 % BER.

Silicon Labs recommends using a signal generator to transmit packets to a DUT, as the signal generator allows for easily configuring the receive input power level at the DUT.

Receive Waterfall Test#

The Receive Waterfall Test determines the receiver sensitivity of the DUT by collecting data to determine the receiver roll-off curve. The DUT is placed in receive mode with 100 valid packets being sent from the signal generator for each channel. The power level should begin at some level before the 1% PER or 0.1% BER threshold. Please refer to the data sheet for the specific radio chip for more information on the sensitivity. The packet success rate is measured for all input powers selected for testing. Silicon Labs recommends that enough input power levels are selected to ensure that the data collected includes both 100% and 0% packets received. This allows for a complete roll-off curve to be observed. This test should be performed over a subset of the operating band to record trends versus frequency. Silicon Labs recommends this test for characterization testing so that the roll-off is quantified and understood but can be omitted at higher volumes.

Silicon Labs recommends using a signal generator to transmit packets to a DUT, as the signal generator allows for easily configuring the receive input power level at the DUT.

Receive Signal Strength Indicator (RSSI) Test#

The RSSI Test measures the RSSI value for a single channel and known receiver input power level. The RSSI is determined by receiving a valid packet from the signal generator and reading the RSSI value through the serial command interface. The DUT is placed into receive mode while the signal generator transmits a single packet and the RSSI measurement is averaged to determine RSSI value. This single data point is measured to verify that the RSSI pin for the radio chip is connected and that RSSI is reporting a valid level. The RSSI operation of the chip itself is validated at the chip testing level and is not tested here.

Silicon Labs recommends using a signal generator to transmit packets to a DUT, as the signal generator allows for easily configuring the receive input power level at the DUT.

External 32kHz Test#

The operation of the external 32 kHz crystal (if applicable) should be verified through the LFXO tune manufacturing token.

Peripherals Test#

Various peripherals, if applicable, can be tested through the serial port. These include anything that may be accessed through the ADC or GPIO (general purpose I/O) on the micro. For example, an LED is often tied to a GPIO pin for some status to be alerted. The state of the LED can be modified by changing the level of the GPIO pin. Another example is reading an ADC pin for a particular voltage level that corresponds to the status of a peripheral such as a temperature sensor or an accelerometer. Any peripheral accessible through GPIO or the ADC should be tested to ensure valid functionality.