A Brief Discussion on Bluetooth Wireless Testing

In order to obtain the Bluetooth trademark, Bluetooth product manufacturers must submit their Bluetooth products to BQTF (Bluetooth Certification Test Facility) for testing, and send their test results and products to BQB (Bluetooth Certification Personnel) for review. Bluetooth certification is relatively complicated. Each time a Bluetooth certification is carried out, the applicant must pay a considerable fee to BQRB (Bluetooth Certification Evaluation Committee). Therefore, it is very necessary for Bluetooth product manufacturers to carry out Bluetooth pre-certification. Bluetooth testing includes compatibility testing and interoperability testing. Bluetooth wireless testing belongs to compatibility testing, which verifies whether the RF performance of Bluetooth products meets the Bluetooth RF specification. Many OEM manufacturers directly purchase Bluetooth chips or modules that have obtained Bluetooth certification, and then develop Bluetooth products such as mobile phones, personal digital assistants (PDAs), computers, printers, MP3 players, digital cameras, cars, toys, etc. Due to the needs of different types of products, it may be necessary to replace the antenna, or due to the influence of other wireless modules or clock modules, as well as changes in power supply, these will cause the RF performance of the Bluetooth final product to change. Therefore, the RF performance of the product must be tested during the R&D and production process to ensure that its wireless indicators meet the requirements of the Bluetooth RF specification.

1 Bluetooth wireless test methods and indicators

The current version of the Bluetooth wireless test specification is version 0.91, which defines Bluetooth wireless test indicators and test methods. The Bluetooth wireless test configuration includes a tester and a device under test (EUT), in which the tester is the master unit and the EUT is the slave unit. The two can be connected by a radio frequency cable or by an antenna through air transmission. The tester sends an LMP instruction to activate the EUT to enter the test mode and configure some parameters of the Bluetooth link between the tester and the EUT. For example, whether the test mode is loopback or transmission mode, whether frequency hopping is required, whether the grouping is a single time slot grouping or a multi-time slot grouping, and whether the net number of the grouping is PN9, or 00001111, 01010101. The test mode is a special state. For safety reasons, the EUT must first be set to the "Enalle" state before it can be activated in the air to enter the test mode.

The following introduces Bluetooth wireless indicators and their testing methods.

1.1 Transmitter test

(1) Output power

The tester sets the initial state as follows: the link is frequency hopping, and the EUT is set to loop back. The tester transmits a payload of PN9, and the packet type is the maximum supported packet length. The EUT decodes the packet sent by the tester and uses the same packet type to send the payload back to the tester at its maximum output power. The tester measures the peak power and average power over the entire burst range at low, medium, and high frequencies. The specification requires that the peak power and average power be less than 23dBm and 20dBm, respectively, and meet the following requirements: If the EUT's power level is 1, the average power > 0dBm; if the EUT's power level is 2, -6dBm < average power < 4dBm; if the EUT's power level is 3, the average power < 0dBm.

(2) Power density

The initial state is the same as (1). The tester scans the frequency to find the frequency point corresponding to the maximum power within the 240MHz frequency band, and then performs a time domain scan at this frequency point (scanning time is 1 minute) to measure the maximum value, which is required to be less than 20dBm/100kHz.

(3) Power control

The initial state is loopback, non-frequency hopping. The EUT works at low, medium and high frequencies respectively, and the return modulation signal is DH1 packet of PN9. The tester controls the output power of the EUT through LMP signaling and tests the range of power control step length, which is required by the specification to be between 2dB and 8dB.

(4) Frequency range

The initial state is the same as (3). The tester performs frequency sweep measurement on the DH1 packet with a payload of PN9 sent back by the EUT. When the EUT works at the lowest frequency, the tester finds the frequency fL when the power density drops to -80dBm/Hz; when the EUT works at the highest frequency, the tester finds the frequency fH when the power density drops to -80dBm/Hz. For a 79-channel system, fL and fH are required to be within the range of 2.4 to 2.4835GHz.

(5) 20dB bandwidth

The initial state is the same as (3). The EUT works at low, medium and high frequencies respectively, and the feedback modulation signal is the DH1 group of PN9. The tester sweeps the frequency to find the frequency corresponding to the maximum power, and finds the fL and fH on the left and right sides when the power drops by 20dB. The 20dB bandwidth Df = |fH - fL|, and Df is required to be less than 1MHz.

(6) Adjacent channel power

The initial state is the same as (3). The EUT operating frequencies are channel 3, channel 39, and channel 75, and the return payload is DH1 packets with PN9. The tester scans the entire Bluetooth frequency band and tests the power of each channel. The leakage power of the adjacent second channel is required to be less than -20dBm, and the leakage power of the adjacent third channel and above is required to be less than -40dBm. (7) Modulation characteristics

The initial state is the same as (3), and the EUT works at three frequencies: low, medium, and high. The tester sends a packet with a payload of 11110000 at the maximum packet length supported, and calculates the peak and average frequency offset of the packet sent back by the EUT, which are recorded as Df1max and Df1avg respectively. The tester sends a packet with a payload of 10101010 at the maximum packet length supported, and calculates the peak and average frequency offset of the packet sent back by the EUT, which are recorded as Df2max and Df2avg respectively. The following conditions are required to be met: at least 99.9% of Df1max meets 140kHz < Df1max <175kHz; at least 99.9% of Df2max meets 3115kHz; Df2avg /Df1avg 30.8.

(8) Initial carrier tolerance

The EUT is in loopback state, and sends back DH1 with a payload of PN9 to the tester. The tester first sets the link to non-frequency hopping, and the EUT works at low, medium, and high frequencies, and then the tester sets the link to frequency hopping. The tester calculates the carrier frequency f0 based on the 4 preamble codes, and requires that the difference with the nominal frequency fTX is less than 75kHz.

(9) Carrier frequency drift

The initial state is the same as (3). The EUT works at low, medium and high frequencies respectively, and the return modulation signal is DH1/DH3/DH5 packets with a value of 10101010. The tester first calculates the carrier frequency f0 based on the 4 preamble codes, and then tests the frequency once every 10 bits of payload. The difference between the frequency and the initial carrier frequency is the instantaneous frequency drift. Finally, the tester turns on the frequency hopping and retests the instantaneous frequency drift at all frequencies. The difference between the instantaneous frequency drifts is defined as the drift rate. For the DH1 packet, each instantaneous drift is required to be less than 25kHz. For the DH3 and DH5 packets, the carrier instantaneous drift is required to be less than 40kHz. The specification also requires that the carrier drift rate be less than 4000Hz/10μs.

1.2 Receiver Test

The above introduces the wireless indicators and tests of Bluetooth transmitters. For receiver testing, all indicators are tested based on the statistics of bit error rate, and at least 1,600,000 bits must be counted. As we all know, it is meaningless to count the bit error rate when the frame error rate is large. Therefore, in order to accurately test the performance of the receiver, the tester must be able to test the FER caused by the following six situations: CRC error, incorrect payload length, synchronization word error, HEC error, EUT sends back NACK packets to MT8850A, and does not receive the packets sent by EUT in the expected time slot. The following introduces the test of Bluetooth receiver.

(1) Single time slot sensitivity

The initial state is the same as 1.1 (3). The EUT works at low, medium and high frequencies respectively, and the return modulation signal is the DH1 packet of PN9. According to the requirements of the Bluetooth specification, the tester controls its output power so that the EUT's receiving power is -70dBm. The Bluetooth specification allows the RF signal sent by the EUT to have an initial error of 75kHz and a frequency drift of 40kHz, that is, a total error of 115kHz is allowed. In addition, errors caused by modulation, symbol timing, etc. must also be considered. If the EUT's receiver performance is tested by a tester that outputs a "perfect" signal, its test results are not sufficient to provide redundancy to adapt to the real wireless transmission environment, and the user will get an erroneous result about the receiver quality. Experience tells us that for interference testing, the sensitivity of the Bluetooth receiver will generally deteriorate by 4~10dB, and the specific value is related to the packet length and the type of Bluetooth chip. The tester must support interference transmission (dirty transmitter), see Table 1, add interference to the transmitted Bluetooth signal, one group every 20ms, from the first group to the tenth group, then back to the first group, and repeat continuously. In addition, the Bluetooth baseband signal is modulated by a sine wave. The tester counts the bit error rate, requiring the bit error rate BER < 0.1%. In addition, if conditions permit, it is best to retest in the frequency hopping state.

(2) Multi-slot sensitivity

Similar to the single-slot sensitivity test, but the packet types are DH3 and DH5.

(3) C/I performance

The initial state is the same as 1.1 (3), and the EUT works at low, medium, and high frequencies. The useful signal sent by the tester is the DH1 packet with a payload of PN9, and it also sends a Bluetooth interference signal with a payload of PN15. The power levels of the useful signal and the interference signal are shown in Table 2. The tester performs bit error rate statistics, requiring BER < 0.1%.

(4) Blocking performance

The blocking characteristic refers to the ability of the receiver to receive useful signals when there are large interference signals in other frequency bands. The initial state is the same as 1.1 (3), and the EUT transmits and receives at a frequency of 2460MHz (channel 58). The tester not only sends DH1 packets with a payload of PN9 as useful signals, but also sends continuous wave interference signals with a frequency between 30MHz and 12.75GHz. The power level of the useful signal is 3dB higher than the reference sensitivity. The reference sensitivity refers to the minimum level that the receiver can receive when a certain bit error rate is met. The level of the interference signal is 2dB higher than that given in Table 3. The tester calculates the bit error rate. If BER>0.1%, the tester records the frequency of the interference signal at this time, and the number of frequencies is required to be less than 24. Other conditions remain unchanged, only the level of the interference signal is reduced to -50dBm, and the tester records the frequency of the interference signal when BER>0.1%, and the number of frequencies is required to be less than 5.

(5) Intermodulation performance

Intermodulation characteristics refer to the receiving capability in the presence of two or more interfering signals that have a specific frequency relationship with the useful signal (their intermodulation products just fall within the useful signal band). The initial state is the same as 1.1(3), and the EUT has the same receiving and transmitting frequencies, which are low, medium, and high frequencies. The tester not only sends a DH1 packet with a payload of PN9 as a useful signal, and its power is 6dB higher than the reference sensitivity; it also sends a sine wave interference signal with a power of -39dBm and a frequency of f1, and a PN15 modulated Bluetooth interference signal with a power of -39dBm and a frequency of f2. The difference between 2 times f1 and f2 is exactly equal to the receiving frequency of the EUT, and f2-f1 =3MHz, 4MHz, or 5MHz. The tester calculates the bit error rate, requiring BER < 0.1%.

(6) Maximum input level

That is, the saturation level of the Bluetooth receiver. The initial state is the same as 1.1(3), and the EUT operates at low, medium, and high frequency points. The tester sends a DH1 packet signal with a payload of PN9 and controls its transmission power so that the level at the EUT receiver entrance is -20dBm. The tester calculates the bit error rate and requires BER < 0.1%.

In addition, transceivers need to test out-of-band spurious emissions, that is, according to ETS or FCC standards, test the out-of-band spurious emissions of the EUT in the working and standby states within the frequency range of 30MHz to 12.75GHz, including antenna conducted spurious emissions and chassis radiated spurious emissions.

2 Bluetooth wireless tester

The above introduces the basic indicators of Bluetooth wireless testing. As a Bluetooth tester, it must meet the following requirements: in order to establish a Bluetooth link with the EUT, the tester must have a Bluetooth protocol stack; in order to perform Bluetooth wireless testing, the tester must fully support the Bluetooth test mode protocol; it must be able to perform 1600 frequency hopping per second; and it must support "dirty transmitters".

The Bluetooth wireless integrated tester MT8850A launched by Anritsu fully meets the above requirements. It fully complies with the latest Bluetooth wireless test specification version 0.91 and can test 8 Bluetooth wireless indicators: for transmitters, output power, power control, initial carrier frequency, carrier frequency drift, and modulation; for receivers, single time slot sensitivity, multi-time slot sensitivity, and maximum input level. The main features of MT8850A are accuracy, speed, and flexibility. The test speed of MT8850A is very fast. It only takes 3 seconds to test the four indicators of power, frequency, modulation, and receiving sensitivity on the typical test on the production line, which is very useful for improving the efficiency of the Bluetooth production line. Users can edit the test script on MT8850A, which is a collection of a series of Bluetooth test items. Users can select the required test items and edit the test environment, such as the number of packets, frequency hopping mode, etc. Once MT8850A is configured, a single-button test can be performed. You only need to press “RUN” and MT8850A will automatically establish a Bluetooth link, activate the test mode, automatically perform the test and report the test results.

ME7865A is a Bluetooth wireless pre-certification system launched by Anritsu, and is currently applying for BQTF. It can test 16 wireless indicators specified by the Bluetooth radio frequency specification. ME7865A is mainly composed of two MT8850As, a spectrum analyzer, a power meter, a PC, and a radio frequency switch box. The instruments are connected through GPIB (general purpose interface bus). One MT8850A is used as the main tester, and the other MT8850A generates Bluetooth interference signals. ME7865A will automatically complete the test of 16 Bluetooth radio frequency indicators and automatically generate a test report in Word format. You can submit the test report to BQB.