As the name implies, the Universal Serial Bus (USB) provides a widely accepted digital port for connecting computers to other electronic devices, typically printers and memory sticks but sometimes test equipment. By relying on an external computer and the appropriate software, USB-driven test equipment can provide quite a bit of performance and functionality in a small package, often a fraction of the size of the more traditional benchtop version of the unit. But there may be times when having that older, larger unit in the test lab may provide more comfort and confidence than the simple convenience of adding measurement capability to a laptop computer by plugging in a USB connector. How do the two formats stack up when comparing workhorse units like vector network analyzers (VNAs), spectrum analyzers, and power meters/sensors? Comparing the two options may help when it comes time to choose between them for a particular test application.
The most glaring difference between USB-controlled units and standard benchtop units is usually size. A USB unit fits next to a laptop computer on a small desk while a benchtop unit occupies the better part of a laboratory work area or 6U or more of a standard 19-in.-wide test equipment rack. A USB unit may weigh about 3 lbs. and fit into a briefcase for ease of transport while a benchtop unit may weigh 10 times as much and require a second set of hands to help when shift it around in the lab.
Another difference between a USB unit and a traditional benchtop unit is the computer. For USB equipment, the computer is part of the test setup, whereas for a benchtop unit, it is an option. For USB test equipment, the choice of personal computer (PC) can contribute to the effectiveness of the measurements since it must have the speed and compatible microprocessor and operating system (OS) to run the test software for the USB units. In a benchtop unit, the controlling computer is built into the equipment, contributing to its larger size.
An external computer can always be added to a benchtop unit, usually as part of an automatic-test-equipment (ATE) system with other units linked to a system-level software program. But a benchtop unit will run quite well without an external computer since it usually has an embedded computer of its own. A high-resolution display screen, whether integrated with a laptop computer or as a separate external display screen, must also be selected for any USB test setup, but is usually built in as a small but high-resolution display screen for a benchtop test unit.
Investigating Test Equipment
How does test equipment in the two formats compare when they have been made to perform the same measurements? The Tektronix TTR506A is a USB VNA from a company long associated with bringing portability to spectrum analysis. VNAs are typically large benchtop units that are rarely considered portable, but the model TTR506A is a compact 11.25 × 8.125 × 1.75 in. (28.58 x 20.64 x 4.45 cm) and weighs less than 3.5 lbs. (1.59 kg). with two test ports and an internal RF/microwave frequency synthesized test source. It can make all four S-parameter measurements from 100 kHz to 6 GHz, owing its compactness and measurement capabilities to a single-board design and proprietary application specific integrated circuit (ASIC). It integrates a bias tee for each port to supply power to an active device under test (DUT). The VNA’s own power supply converts 100 to 240 VAC at 50/60 Hz to +5 VDC; the USB test equipment draws maximum current of 4 A during operation, performing forward and reverse S-parameter measurements on transistors, ICs, and other active and passive devices.
An external computer is an important part of any USB-driven test setup and the TTR506A requires a PC with at least an Intel Core™ i3 microprocessor, 8 GBytes memory, and Microsoft Windows® 7 or newer 64-b OS. The compatible computer interface is USB 2.0. Since the VNA does not have a display screen, the screen for the PC or a separate external screen shows the results of measurements. By making the USB connection, the result is a VNA with frequency range of 100 kHz to 6 GHz and 1-Hz resolution, dynamic range of 112 dB to just under 1 MHz, 124 dB to just under 200 MHz, and 122 dB to 6 GHz, with a broadband noise floor of -130 dBm. Measurements are made via Tektronix VectorVu-PC™ software running on the controlling PC.
When checking the performance of the model TTR506A USB VNA with a benchtop network analyzer such as the Keysight E5071C ENA network analyzer, the first word that comes to mind, other than size or weight, is control. The E5071C is available with many options, such as two- and four-port versions, and many frequency ranges, including 100 kHz to 6.5 GHz (and 1-Hz resolution) with bias tees and 9 kHz to 6.5 GHz without bias tees. Tremendous measurement versatility is possible by controlling different parameters although adding to measurement setup time and complexity compared to a USB VNA. The increased control comes at the expense of larger size, weight, and cost. The model E5071C is more than twice the size of the TTR506A USB VNA, at 42.6 × 23.5 × 47.2 cm, and many times the weight, with two-port versions at 40.04 lbs. (18.2 kg) and four-port versions at 43.78 lbs. (19.9 kg).
Parameters that can be controlled include the intermediate-frequency (IF) bandwidth, from 10 Hz to 1.5 MHz, which will impact the amount of noise detected during a measurement and the subsequent dynamic range. The dynamic range through 6 GHz is typically 123 dB with a 10-Hz IF bandwidth, increasing to 98 dB through 6 GHz with a 3-kHz IF bandwidth. Test port power can be controlled, maintaining amplitude linearity within ±1.5 dB during sweeps, over a range of -55 to +9 dBm through 6 GHz. The E5071C analyzer incorporates a GPIB control port (via 24-pin D-sub connector), 1-Gb/s LAN, and a USB 2.0 host port. It shows test results on a 10.4-in. TFT color LCD touch screen with 1024 × 768-pixel resolution and can drive an external monitor.
The Tektronix RSA513A was developed as a line of USB-controlled portable spectrum analyzers. It runs for about 4 hours on a rechargeable 14.4-VDC lithium-ion battery and can make mobile measurements from 9 kHz to 13.6 GHz (1-Hz resolution) when connected to a battery-powered PC. It achieves a 70-dB spurious-free dynamic range for a 40-MHz real-time acquisition bandwidth and, with its low noise floor (typical noise is -80 dBc without preamplification), can detect signals over a wide range of power levels with built in 0-to-51-dB attenuation, adjustable in 1-dB steps. In the field, built-in GPS capability simplifies mapping and interference location, and its high sensitivity makes it a candidate for signal intelligence (SIGINT) gathering. In the lab, it is fully capable of EMI/EMC compliance testing. The USB spectrum analyzer measures 10.68 × 11.78 × 2.65 in. (271.3 × 299.1 × 67.3 mm) and weighs 7.5 lbs. (3.40 kg) without the battery and 8.5 lbs. (3.85 kg) with the battery. It operates by means of the company’s SignalVu-PC software but requires a PC with memory storage rates of at least 300 MB/s.
For comparison, the Rohde & Schwarz FSV13 is a larger benchtop model signal and spectrum analyzer that provides resolution of 0.01 Hz over an AC-coupled range of 10 Hz to 13.6 GHz and a DC-coupled range of 10 MHz to 13.6 GHz. Its typical displayed average noise level (DANL) is low across the unit’s bandwidth, about -130 dBm at the lowest frequencies and -145 dBm or better above 100 kHz. It shows measurement results on an 8.4-in. color TFT display with 800 x 600-pixel resolution. The FSV13 measures 16.22 × 7.76 × 16.42 in. (412 × 197 × 417 mm) and weighs 22.7 lbs. (10.3 kg). Although nominally a benchtop unit, it is available with a host of options, including a 12-VDC lithium-ion battery pack for portable operation.
Power measurements may be where the USB format is most noticeably different than the traditional method of reading power levels with a sensor and a meter, simple because with USB power measurements, the meter appears to be missing. In fact, it is integrated with the sensor into a since sensor-sized device with cable assemblies to connect to a DUT at one end and the USB port of a computer at the other end. As an example, the Keysight U8487A USB thermocouple power sensor appears to be a single power sensor, but the metering circuitry is inside the sensor housing and the controlling computer is connected via USB port.
The model U8487A looks more like an accessory than test equipment, measuring 127.70 × 46 × 35.90 mm and weighing just 0.22 kg. It is small but capable. Working with a PC and the company’s BenchVue software, it achieves a power measurement range of -35 to +20 dBm with ±0.5% linearity at room temperature (+25°C) at frequencies from DC to 50 GHz. Since the PC is part of the test setup, when compared to a more traditional combination of power meter and sensor, such as the Keysight E4418B power meter with a Keysight 8487A thermocouple power sensor, the size/weight differential is not as great as with the other USB units.
A model E4418B with a model 8487A thermocouple sensor covers the same power and frequency ranges as the USB model U8487A, but in an enclosure weighing 4 kg and measuring 212.6 × 88.5 × 348.3 mm. When comparing the two approaches, the single word that comes to mind is flexibility, since the E4418B power meter has a total frequency range of 9 kHz to 110 GHz, depending upon sensor, and works with both thermocouple and diode power sensors.
There is no doubt that USB test equipment is finding their places on test benches, in the field, and even in home laboratories. They offer accuracy and ease of use but rely on a PC for measurements. Older, larger benchtop versions of essential test units will not be going away anytime soon, since they do provide levels of accuracy, control, and flexibility simply not possible with USB test equipment.