7 tips for selecting PXI test equipment for wireless applications

Author : Sheri Detomasi, Keysight Technologies

25 April 2016

Figure 1 - Simplified power amplifier test block diagram

With the increased complexity in RF components used in wireless products such as cellular phones, tablets, and wireless routers design engineers are under increasing pressure to add more functionality, while keeping design and manufacturing costs low. 

(Click here to view article in digi-issue)

Selection of test equipment and software can impact the performance and efficiency of the wireless device design as well as the cost to produce the device. The following seven tips will help design validation and manufacturing test teams select hardware and software that will accelerate design, increase test throughput, and ultimately lower the cost of test.

New multi-radio and multi-function devices are requiring wider transmission bandwidths, more complex modulation schemes, and even multiple transmit and receive chains, which significantly increases device complexity and test need to verify the device’s performance. The following are some typical challenges faced by RF device test engineers.

Challenge 1: Increasing device complexity 

The simplified block diagram in Figure 1 shows a typical power amplifier (PA) manufacturing test setup which enables a test engineer to quickly determine if the device under test (DUT) meets the required modulation performance at the desired power level. In this example, a vector signal generator (VSG) is connected to the DUT through cabling, switching and signal conditioning. Switching is often used to support testing of multi-band modules. Many new PAs support multiple operating modes, along with multiple frequency bands that allow for higher performance from the user equipment. As the complexity of these devices increases, the ability to get measurements through the additional switching and signal condition becomes more difficult.  

Challenge 2: Reduce test time 

Reducing test time by even a few ms can have a significant impact on the cost of test. In design validation, multi-radio tests can take anywhere from hours to days, but in manufacturing, seconds count. For example, a typical PA manufacturing test (see figure 1), includes the time it takes to adjust the modulated input power from the signal generator to get the desired output level from the DUT. In a typical servo test, a control loop is used to determine the final gain of the PA. After the DUT output level is measured by the signal analyser, the new value of the output power is calculated based on the difference between the measured power and the desired power. The signal generator is then adjusted to the necessary output power to achieve the correct output power from the DUT. Only after the DUT output level is set to the correct value can the specified parameters be tested. The time spent adjusting the signal generator to get the correct DUT output power can be one of the major contributors to overall test time which increases the cost of test. 

Challenge 3: Scalable instrumentation for multi-channel measurements 

Newer wireless designs implement multi-antenna techniques like carrier aggregation and MIMO to increase data throughput and cell capacity. Testing these designs often requires multiple signal generators and/or signal analysers to simulate and analyse the multi-antenna signals (see Figure 2). This test setup gets very complicated and requires multi-channel synchronisation and new measurement techniques that allow for the hardware and software to demodulate and analyse the multiple data streams simultaneously. 

Figure 2 - MIMO testing setup diagram

Tips for selecting modular test equipment

The following 7 tips help RF test engineers select hardware and software that will accelerate design, increase test throughput and ultimately lower the cost of test.

Tip 1: Select a signal generator with good modulation performance, even at high output power levels, so the highest quality signal can be delivered to the DUT

Many PAs require the digitally modulated signal, such as LTE, to be at an input power level of 0 to +5 dBm at the DUT. A signal generator that can output signals as high as +15 dBm is suggested in order to overcome any losses between the signal generator and the DUT which may arise due to switching or other signal conditioning. Shown in Figure 3, is an adjacent channel power (ACPR) measurement at high output power levels from a PXIe VSG with excellent modulation performance. At +10 dBm there is little to no degrading of the ACPR and at +15 dBm the ACPR is still near 60 dBc. New PXIe VSGs can provide digitally modulated signals up to +19 dBm output power with ±0.4 dB level accuracy, resulting in the highest quality signal being delivered to the DUT at a lower cost.

Tip 2: Select instruments with fast measurement speed to reduce test time and lower the cost of test

A signal generator with fast frequency and amplitude switching time enables quick modulated input power adjustments to get the desired output level from the device under test. When the output level of the signal generator cannot be predetermined for use with list mode, rapid frequency and amplitude switching speeds, along with superior linearity, repeatability and resolution specifications will significantly shorten the time it takes to converge on a desired output power and to reduce test time and the overall cost of test. On the measurement side, selecting a signal analyser that can perform power, ACPR, EVM and harmonics tests quickly and accurately, and that can switch between different measurements in minimal time will ensure that accuracy is not sacrificed for speed.

Power measurements with newer PXIe VSAs are captured in the digitiser in real time and returned in a single measurement value to the application program. No additional computation of the power measurement is required in the controlling PC. These VSAs provide power measurements with acquisition times from 10 µs to 1 ms. When combined with the power level switching speed of the PXIe VSG, the step time for a power servo loop test can be less than 1 ms, dramatically reducing the amount of time it takes to converge on the desired output power level in the servo test. Reducing test time, increases device test throughput, and the overall cost per device goes down.

Tip 3: Select a flexible platform that easily incorporates different instruments for testing throughout the product lifecycle 

Figure 3 - ACPR measurement at high power

In the design or design validation phase of the product development cycle, it may be necessary to look at the RF device’s performance out-of-band. Any harmonics or spurs generated by the device can affect the quality of the output signal. If out-of- band signals are present, this can cause interference in the wireless network and compliance issues with FCC regulations. It’s helpful to have the flexibility to include signal analysers with higher frequency ranges earlier in the design phase to measure out-of-band spurs and harmonics. If, for example, the device supports LTE bands 40/41, the design engineer may want a signal analyser that that can measure the 3rd harmonic up to 8.1 GHz or even the 7th harmonic up to 19 GHz. Once the product moves to manufacturing, a lower frequency, lower cost signal analyser can be used. 

Tip 4: Select test equipment that uses the same measurement software from R&D to manufacturing for consistent, reliable test result validation 

When traditional benchtop and PXI instruments are used for device test, use of common software with the same measurement algorithms and measurement science will ensure consistent, reliable results, regardless of instrument form factor. For example, measurements made in the lab with benchtop equipment can be quickly validated on PXI instruments in a manufacturing test environment. The ability to use the same programming commands and a consistent user interface further reduces test development time.

Tip 5: Select test equipment that has warranted specifications for greater measurement integrity 

Poor measurement integrity can lead to test escapes or false failures in devices under test, which increase repair costs and can lead to higher overall manufacturing costs. Calibration directly impacts measurement quality. An instrument’s specifications are guaranteed through calibration. PXI instruments may be calibrated at the individual module level or as a bundle of multiple modules when they are combined to make a single instrument, such as a PXIe VSA. Some vendors only calibrate individual modules. When this is done, it is very difficult to specify warranted instrument level specifications. It is important that the test equipment selected include calibration routines at an instrument level. 

Tip 6: Select a test platform to minimise system downtime

Supportability and quick repair turnaround times (TAT) are critical factors in keeping a system up and running. It can be very costly when unplanned maintenance or equipment failures are encountered in manufacturing or in critical R&D applications. Just one test system unexpectedly going down for repair can have a devastating impact on DUT shipments. In addition, the cost for repairs, re-calibration after repair, and re-installation into the system adds up quickly. Be sure to select test equipment that has low meantime between failures, fast TATs and longer product warranties.

Tip 7: Select a test platform that is scalable for future needs 

We live in a dynamic world. Wireless bandwidths continue to increase and test equipment needs to support new standards and test requirements. Selecting a PXI or AXIe platform that can evolve as standards and test requirements evolve can help future-proof a test system. Some companies offer license key upgrades for hardware options to let engineers keep their equipment operational, in house. Be sure to select equipment that can be upgraded later as test requirements change.


Contact Details and Archive...

Print this page | E-mail this page