Texas Instruments Uses NI PXI to Reduce FM Transmitter Characterization Cost and Time

"Using the National Instrument PXI platform, we were able to reduce test time by almost six times and develop a cost-effective, long-term RF test bench solution."

- Min Xu, Texas Instruments

The Challenge:

Improving the measurement speed performance of our RF test benches. These test benches are comprised of traditional rack-and-stack instruments involving multiple platforms. While our existing test system is highly functional, we wanted to reduce system complexity and improve measurement speed without sacrificing measurement accuracy.

The Solution:

Using NI PXI instruments to reduce cost and measurement time while improving measurement quality and meeting our goal to reduce test cost and improve characterization time.

Author(s):

Min Xu - Texas Instruments
Ross Kulak - Texas Instruments

 

Combo RF integrated circuits (RFICs) include multiple radios and baseband processors on a single chip. Several benefits derived from multiradio RFICs include increased integration that results in shared common circuitry, power, and size. Combo RFICs enable various radio capabilities such as Bluetooth, GPS, FM signal generation, and WLAN. The benefits of highly integrated multiradio RFICs are easily accessible to product designers, but their increased functionalities make it difficult and challenging for characterization engineers. Greater functionality generally requires additional measurements. Characterization engineers are faced with the challenge of trying to reduce RFIC characterization time, while chip complexity continues to increase. The National Instruments PXI platform proved to be a cost-effective, rapid test bench solution that substantially reduced characterization time and costs for combo RFICs.

    

Application

Our objective for this specific application was to test the FM transmitter portion of a combo RFIC Texas Instruments (TI) chip. The chip functionalities are similar to an automobile FM radio transmitter, and we needed to measure a variety of output characteristics including:

 

  • Signal-to-noise ratio (SNR)                                              
  • Spurious-free dynamic range (SFDR)                              
  • Total harmonic distortion (THD)                         
  • Total harmonic distortion plus noise (THD+N)                
  • Max spur power
  • Max spur frequency         
  • Power in band
  • Peak level

 

When choosing the PXI system, our goal was to find a platform that was fast, cheap, and cost-effective without sacrificing performance. To prove that performance was not sacrificed, we thoroughly compared the PXI measurement results with our existing test bench. The existing test bench used various traditional instruments from Audio Precision, Rohde & Schwarz, and other vendors. The comparison showed similar measurement performance; PXI showed faster measurement time and better measurement throughput.

 

Correlating PXI Measurements With the Existing System

The two platforms we compared were: (1) National Instruments PXI system consisting of an NI PXI chassis with an NI PXIe-5663 vector signal analyzer, NI PXI-4461 digital-to-analog converter (DAC), NI PXI-2596 multiplexer, and an NI USB-5680 power meter; and (2) the standard test bench system consisting of an Audio Precision signal generator, a Rhode & Schwarz RF signal analyzer, and a PC.

 

We used a straightforward test methodology that requires generation of a 1 kHz tone to the IC audio input. We used the RF signal analyzer and power meter to measure the power at the RF output and we demodulated the FM signal using NI LabVIEW software. After demodulating the FM signal, we performed additional measurements such as THD and SFDR on the demodulated 1 kHz tone. The following diagrams show the setup of the two platforms:

 

 

Results

After a thorough evaluation of both the PXI solution and our previous test solution, we discovered that we could reduce test time by almost six times using the NI PXI platform. The full-channel sweep of the standard bench required approximately seven hours in comparison to the PXI platform test time of 76 minutes. Moreover, each channel sampled by the standard bench takes 76 seconds in comparison to the PXI per-channel sample time of 14 seconds. Figure 2 shows the test time comparison for both test bench systems:

 

Based on multiple SNR spectral tests, the NI PXI platform outperforms the standard bench in terms of measurement throughput for an equivalent accuracy (standard deviation). The SNR delta between the two benches of 0.6 dB is due to a known issue in windowing on the standard bench. PXI allowed us to access the data without windowing and is therefore not subject to this issue.

 

Business Benefits

The NI PXI platform provides a cost-effective, long-term RF test bench solution. The standard bench costs approximately $85,000 USD; the NI PXI platform costs approximately $43,000 USD and reduces our system cost by almost half. In addition, the standard test bench system required a six-week downtime to send the equipment overseas for calibration. The PXI platform calibration time is roughly a week, which greatly reduces station downtime. By using the PXI platform and LabVIEW, we reduced test time, shortened time to market, and saved on instrument capital equipment costs, while improving measurement quality.

 

Author Information:

Min Xu
Texas Instruments

Figure 1. The comparison test system used instruments from Audio Precision, Rohde & Schwarz, and other rack-and-stack measurement equipment.
Figure 2. Test Time Comparison