Incoming inspection and final testing of GSM/UMTS phones
(Editor: Achim Grolman, Product Marketing Communications Manager for Willtek Communications)
With UMTS coverage available in many centres and more than a dozen dual-mode GSM/UMTS phones available on the market, repair centres are now asking what to test on these phones both when they are reportedly defective and before the repaired phones are returned to the customer. A new Willtek application note details the tests to be performed; this article is an essence of the application note.
Before starting with the measurements, it is important to ensure that the phone or the measurements are not affected by radio signals from other phones or a nearby base station. Similarly, the RF power transmitted by the phone must not interfere with the real network! – Figure 1 shows an example of a test setup that includes a box for RF shielding.
Figure 1: Test setup example with 4920 RF Shield Box
Transmit power measurements
In any radio network based on CDMA, control of the transmit power is important as it affects the capacity of the radio cell. Generally, each transmitter adds to interference, so the capacity of the cell is affected even by calls in the neighbouring cells. A large capacity can only be assured if the transmit power of each wireless terminal is minimised to the extent that the signal strength just allows for the required quality of service.
Therefore power control is of great importance. There are two different ways to control transmit power: open loop power control during call setup and closed loop power control during an ongoing connection. The latter implies that the base station controls the phone’s transmit power by way of the TCP (Transmit Power Control) bits. The transmit power is defined over a dynamic range between minimum and maximum; the lower end is fixed for all wireless terminals at -50 dBm. The upper end is defined by the power class for the phone at hand. To ensure that the phone can communicate both nearby and very remote from a base station, measuring the minimum and maximum output power can be important for the final test.
The transmit power is controlled in steps of 1 or 2 dB relative to the previous level, in contrast to GSM where the base station controls the absolute power level of the phone. The power level change must be applied in the time slot immediately following command reception. This process is called “inner loop power control” and requires accurate measurement of the level changes in the final test or incoming inspection (see Figure 2).
Figure 2: Inner loop power control
Modulation quality
The quality of the WCDMA signal can be determined with the error vector magnitude (EVM) and the peak code domain error. Both are equivalent methods of assessing the modulation quality.
The error vector can be represented in the I/Q diagram of the WCDMA signal, where amplitude and phase are shown. The modulation error is supplied by the difference between the measured and the calculated ideal modulation vector, and evaluated for each symbol. Quality assessment is simplified by regarding the RMS-averaged EVM only. This single parameter describes the modulation quality of the complete signal.
Alternatively, the modulation quality can also be described in the code domain. The entire transmit power is split into the individual code channels characterising the CDMA system. Code channels that are not allocated do not transport data but only noise which can be considered as cross-talk from the allocated channels. Code channel crosstalk is caused when the real-life transmitter injects additional noise into other channels so that orthogonality is affected or lost. For each code channel, the test instrument displays the channel power relative to the entire power. The code channel most interfered with and the power injected are of high interest. The ratio between each unallocated code channel and the power in the allocated code channel is calculated, giving the code domain error; the largest one is considered as the peak code domain error (PCDE). It is important to evaluate this error over a longer period of time as it varies, so this measurement is less suitable for a quick final test than the EVM measurement!
Finally, the frequency error describes the difference between the used and the assigned carrier frequency. The actual frequency may deviate by a maximum of 1 ppm from the frequency assigned by the base station. In the European and Asian frequency band, for example, this means a limit of ±198 Hz.
Frequency spectrum
Spectrum measurements for WCDMA include that of the occupied bandwidth (OBW). 99% of the entire power should be spread within a range of no more than 5 MHz around the carrier frequency. In the example measurement in Figure 3, the occupied bandwidth is 4.17 MHz.
Figure 3: Occupied bandwidth measurement
Also included in the group of spectrum measurements is the adjacent channel leakage power ratio (ACLR). This measurement determines the spectral power in the neighbouring channels relative to the power in the allocated channel. The limits should not be exceeded to avoid interference with neighbouring channels. For this measurement, the phone usually transmits at maximum transmit power, which depends on the actual power class (out of four).
In the spectrum emission mask (SEM), the signal spectrum outside the allocated channel is measured. The resulting display is split into two parts. In a spacing between 2.5 and 3.5 MHz from the carrier frequency, the signal is measured at a resolution bandwidth of 30 kHz. A 1 MHz filter is used above 3.5 MHz up to 12.5 MHz from the carrier. The specification gives upper limits depending on the frequency; these limits are pre-programmed and marked red on the 4400 display as shown in Figure 4.
Figure 4: Spectrum emission mask test
Receiver measurements
Bit error and block error rate measurements (BER, BLER) help to assess the receiver of digital transmission systems. The test set transmits defined test sequences to the phone’s receiver; these are pseudo-random bit sequences (PRBS). The transmitted are compared with the received bits (BER) or data blocks (BLER), and thus the error rate is determined. The reference sensitivity level describes the lowest receive power at which the bit error rate does not exceed 0.1%. Proper UMTS phones achieve bit error rates below 0.1% at levels down to –106.7 dBm.
|
