Review: Windfreak Technologies SynthNV RF generator (Part 1)
Figure 1 - The Windfreak Technologies SynthNV RF generator weighs just a few ounces and easily fits in your hand.
For months, I’ve been seeking a small RF generator that could replace the 40-pound monster I keep under my workbench. What really caught my eye initially was that the generator could AM modulate the RF output - perfect for radiated immunity pre-compliance testing! In addition, it will pulse modulate the output - perfect for testing to the MIL-STD-461 and DO-160 standards. The RF output level is sufficient to drive a near field probe with enough field strength to investigate susceptibilities within a product’s internal circuitry.
But wait, there’s more! Here’s rundown of the features of this palm-sized jewel. Some of these additional features will be reviewed in Part 2 of this series.
• RF sweep generator (34.4 MHz to 4.4 GHz at up to +19 dBm output)
• Network analyzer (34.4 MHz to 4.4 GHz)
• VSWR analyzer (using external power coupler)
• RF power meter (real time)
The generator is USB powered and can run on most Windows operating systems, including Windows 8. It also includes an external power adapter input, so it can be programmed into a given state and then disconnected from the PC and run standalone as a local oscillator or RF generator. There is a port that can source or receive an external 10 MHz clock, as well as an RF input port for measuring power. This port is also used for the network analyzer function.
Figure 2 - The basic user interface for the generator is based on National Instruments Labview.
The well designed user interface (Figure 2) is based on National Instruments Labview and the provided software includes the runtime engine for those who don’t own the full Labview software. It installed and ran just fine on my Macbook Pro with Parallels 9 and Windows 8. There are several tabs along the left half of the panel. These select the major functions of the instrument controller. When in manual mode, the large knob tunes the frequency in preset steps of 1kHz, 10kHz, 100kHz, 1MHz, 10MHz and 100MHz. The user may also enter frequencies directly by typing in the data blocks or by using a keyboard control in place of the large knob. The nominal +19 dBm RF output power is controlled by the rightmost panel. There are two buttons controlling the preset output. The High Power button will switch between the default high power or when pressed decreases the overall power output by about 55 dB (Low Power). A second button turns the RF on/off. The slide control further adjusts the output power by up to 31.5 dB. Note that the power scale is “dB’ and not the actual output power in “dBm”. This can be confusing at first, because the natural inclination is to assume the scale corresponds to the actual power output. This requires some mental calculations (or confirmation measurement) to set the precise output level. I suspect one slick way to confirm the desired power level is to run the output (through attenuators) to the “RFin” port to make that measurement. One improvement might be to redesign the slider to conform to the actual output power level - changing the scale according to the three preset power levels.
Figure 3 - The typical RF output spectrum shows relatively low phase noise (measured at 100 MHz and 100 Hz RBW).
Figure 4 - Typical 100 MHz fundamental and harmonic output of the generator. The third harmonic is about 10 dB down from the fundamental.
I believe the relatively high harmonic levels (Figure 4) are due to the fact the output waveform below about 500 MHz is more of a square wave with spikes on the leading edge. Above 500 MHz, the signal converts to a nice sine wave. I’ll be investigating this further in Part 2. For the purposes of radiated immunity testing, this is not that important, as the next highest harmonic (3rd order) is 10 dB lower (1/10th the power). Still , if cleaner signal is desired for other applications, filtering will be required.
Figure 5 - A power sweep of the maximum RF output in dBm. Note the average is about +19 dBm and is fairly flat out to 1.5 GHz (the limit of my spectrum analyzer).
I let the generator sweep slowly from 34 to 1500 MHz (Figure 5) to check the output flatness. I was impressed that it held up so well. Above 3 GHz, however, the output gradually falls off by about 8 dB. This is still plenty of power to drive the near field probe.
Figure 6 - Set up the AM modulation by clicking on the Mod and then AM tabs.
For the usual modulated AM immunity test, click the “Mod” tab , then “AM” tab (Figure 6) and set the AM frequency to 1000 Hz (the closest the synthesizer can get is about 1008 Hz), the number of cycles to 100 (this cycle of 100 will continue to repeat in the software, so is not too critical), and the amplitude at 80%. This will set up an 80% AM modulation to the RF output (Figure 7).
Figure 7 - Spectral display of the 80% AM modulation at 100 MHz as measured on an Agilent MSO-X 3102A 1 GHz BW oscilloscope. There appear to be some digital switching artifacts superimposed, but I don’t believe this will affect the testing outcome. If anything it makes the test more rigorous.
Next, set up an automated or manually stepped sweep with the start, stop and step frequencies. Set an appropriate step time for automated sweeps and then press either the Sweep Once or Sweep Continuously buttons. It would be nice for the purposes of pre-compliance testing if the control panel allowed for manually controlling the sweep up or down, as well as manually taking control once immunities are discovered. Currently, the generator only sweeps upward in frequency. Perhaps a later update will address this.
Figure 8 - A measurement of the approximate H-field level in V/m using the larger Beehive model 100C probe closely coupled to the wire antenna of a field strength meter. This serves to simulate coupling energy directly to a circuit trace or cable. There appears to be saturation at 9.3 V/m, so this may be anomalous data. The more important issue is to ensure a large field level coupled to the circuitry under test. Absolute levels are not as important when troubleshooting, so long as the relative level can be changed.
In the usual troubleshooting or pre-compliance process I generally connect a near field probe (either H- or E-field) directly to the output of the generator and, while observing for disruptions, carefully sweep the probe around on the operating circuit board or cables of the product under test (Figure 9). It’s helpful if there is already compliance test failure data available, so the failing frequencies can be pre-tuned in advance. Otherwise, slowly sweep from 80 to 1000 MHz (or 2000 MHz depending on the standard). Once areas of circuit disruption are discovered, the power may be decreased in order to zero in on the offending circuitry more closely. The field levels at the tip of the probe were measured to be between 4 and 9 V/m between 34 and 1000 MHz - plenty of signal to induce most product failures.
Figure 9 - Probing a RaspberryPI embedded processor for radiated immunity.
Overall, I’m impressed with the capabilities of this generator. I love the clean design of the control panel and the fact I can now carry my entire radiated immunity test setup literally in my pocket. This will be a big help when traveling to a client’s location to help troubleshoot immunity issues. The unit may be ordered directly from the company web site via PayPal and delivery was quick. I also appreciated the great technical support - both over the phone and via email.
Besides the previous software control enhancements, I’d be more comfortable if there was a small LED “RF On” indicator on the device itself. This would confirm the RF was really on during lengthy testing for immunity. I guess I don’t altogether trust software! Highly recommended.
Part 2 will cover measurements using the network analyzer and real time power analyzer.
For more information and products mentioned:
Windfreak Technologies (SynthNV RF generator)
Beehive Electronics (near field probes)
Rigol Electronics (DSA815TG spectrum analyzer)
Agilent Technologies (MSO-X 3102A oscilloscope)