Many of you may know, as an EMC consultant, I’ve been partnering with one of the best EMC test labs in Colorado, EMC Integrity, in Longmont, north of Denver. EMC Integrity was founded by Vince Greb in 1993 and now owns two of only three 10m semi-anechoic chambers in the state (the other is owned by Hewlett-Packard in Ft. Collins). They specialize in both commercial and military EMC testing. Check them out at www.emcintegrity.com. They were recently featured in an article in the Boulder County Business Report (March 15-23, 2013). Check it out here…
I’ll be presenting a comprehensive two-day EMC design and troubleshooting short course there April 2-3 (sold out), with a possible follow-up course later this year.
Technician, Casey Lockhart running a radiated emission test at one of the 10m chambers. Photo courtesy BCBR (© Jonathan Castner).
Always on the lookout for useful, but inexpensive test equipment, I recently ran across the Triarchy Technologies USB spectrum analyzer, model TSA5G35. The one thing that really struck me was the whole thing was built into a USB dongle, just a little larger than a memory stick. What’s more, the advertised frequency range was 1 MHz to 5.35 GHz. This, I had to see for myself.
So, is a spectrum analyzer no larger than a pack of chewing gum that you can carry in your pocket good enough for EMC analysis and troubleshooting? For a total cost of $599 (through their store on eBay), I decided to take a chance and run this remarkable PC-based analyzer through the ringer.
Figure 1 – Photo showing the analyzer with supplied USB extension cable and 30 dB attenuator.
Figure 2 – A screen capture of a series of 16 MHz oscillator harmonics.
SPECIFICATIONS – Basic specifications include frequency coverage of 1 MHz to 5.35 GHz, resolution bandwidths of 50 through 500 kHz (not selectable), frequency spans from 1 MHz to 1 GHz, input level range of -110 to +30 dBm (using the supplied 30 dB attenuator for the higher power levels), and typical noise levels of -80 to -100 dBm (depending on the span and RBW). The maximum power level is +20 dBm for 1 minute (without the external attenuator) and +/- 25 VDC, which is excellent protection for this little instrument. The reference level range is -60 dBm to 0 dBm (no external attenuator) or -30 dBm to +30 dBm (with the external attenuator. the usable display range is 80 dB with a noise floor of -115 dBm at a 5 MHz span and -60 dBm reference level at 1 GHz. Amplitude accuracy is specified at less than 3 dB. All in all, not to bad for this little guy.
For more of this hands-on review on Test & Measurement World, click here…
These unusual EMI sources may be used to perform pre-compliance testing (radiated or conducted immunity) to help determine the immunity characteristics of your circuits or product.
1. Chattering Relay (120VAC powered) – can produce strong broadband emissions all the way out to at least 1 GHz.
2. 3 VDC Motor – produces strong emissions out to about 750 MHz.
3. Pocket Plasma – produces broadband frequencies up to 10 MHz.
Click here for more details…
I recently received the following question on how to calibrate current probes and thought you’d be interested.
Question: Good morning. I read your article, “HF Current Probe: Theory and Application”, but now I have a question I’m hoping you can help me answer. I am attempting to measure the transfer impedance of a current monitor probe using a probe calibration fixture or jig. To keep the setup simple, I am using a signal generator and a power meter. As in your example, I am setting the generator to source 0dBm and I will verify it with the power meter and sensor through an adapter; I will then connect the generator’s output and current probe to the calibration fixture and measure the probe’s output using the same meter and sensor. This all works fine until the input SWR of the calibration jig reaches about 1.3 at 100MHz. From that point up to 400MHz, the SWR of the jig reaches 3.4. It appears that one would be measuring both the current probe’s insertion loss and the calibration jig’s mismatch loss. Would it then be best to establish the reference by measuring the output of the signal generator while it is connected to the calibration fixture (without the probe inserted), so as to include the jig’s mismatch loss in both the reference and measurement sweeps?
Answer: You’re on the right track. You need to normalize out the effect of any mismatch from the jig setup. There are several methods for calibrating current probes. If you have the jig, that’s great. Basically, what you’re trying to do is measure the current accurately versus frequency – a not so trivial task to keep the current fixed as frequency changes. The problem is that any parasitics (R, C, L) in the wire to be measured can greatly influence the current value. That was the problem I was running into when measuring the wire in the referenced article. I tried to keep the value of current fixed by inserting a small resistor in series and measuring the voltage drop, keeping the this voltage drop steady by adjusting the RF generator output. It’s much better to use the 50-Ohm jig, but there will still be mismatch errors, which may be somewhat alleviated through the use of 6 to 10 dB attenuators. The goal is to measure the current through the probe versus the voltage at the probe terminals. Dividing the terminal voltage by the current gives you the transfer impedance. I’ve attached a few references.
Here’s a recent article from Interference Technology.
Teseq also has a calibration procedure within the instructions for their test jig, and look under the “downloads” tab.
Dr. David Pommerenke, of Missouri University of Science and Technology (MST), authored a paper with Ram Chundru and Sunitha Chandra on “A New Test Setup and Method for the Calibration of Current Clamps“, which runs through the historical calibration methods and then suggests an improved method.
As EMC engineers, we use many types of antennas – many broadband, these days. As a traveling EMC troubleshooter/consultant, I reply on small collapsible DIY antennas for troubleshooting, as described in an earlier blog posting.
In order to characterize these adjustable antennas versus frequency, it’s useful to be able to measure them with different element lengths extended, so that you know about where to set the length for the specific harmonics of interest. Using the new Rigol DSA815TG spectrum analyzer with tracking generator, and VSWR (voltage standing wave ratio) option, you can determine both the resonant frequency and VSWR, or how well the antenna is matched to the 50-Ohm coax cable. more…
One of my favorite companies, Mini-Circuits, developed a (really!) low-noise, broad band, preamp (model PSA4-5043+) earlier this year for use as a front-end amplifier for mobile telecom applications, such as GSM, CDMA, LTE and WiMax. However, it would also be ideal to amplify spectrum analyzers and the low-noise feature would lower the effective noise figure of the analyzer, allowing you to see low-level signals better. This would not only be useful for amplifying near-field or current probe outputs, but would work well to boost the antenna output in semi-anechoic chambers – especially if there is a long run of coax cable.
Lets take a closer look and sweep the gain on our spectrum analyzer. more…
In my never-ending quest to search out useful reference books on EMC, I recently ran into the 3rd edition of “The ARRL EMC Book” (ISBN 9780872590915). First published in 1999, the new 3rd edition was released in 2010. For those unfamiliar with the ARRL (Amateur Radio Relay League), this is the national organization representing amateur radio operators (“hams”) in the U.S. Most hams are members of the league, which also publishes a range of useful operating, design and general radio reference books. Because hams are allowed to operate their two-way radios at up to 1.5kW, on occasion, this may be the cause of local interference to poorly-designed or poorly-shielded consumer products. Thus, several years ago it was decided to publish a reference book on RFI (radio frequency interference). In this article, I’ll describe the most important content and why you might want to buy it. More…
There are a number of engineering tools that work well for simulating EMC scenarios. One of these is MathCad, a tool that’s been around for a number of years. The neat thing about MathCad is that you can define a set of equations in free-form layout and plot out the analysis in several engineering-type graphs, including logarithmic and polar. These plots are easily copied and may be pasted into documents or technical papers.
As an example, I’ve calculated the harmonic content of a square wave given user-defined fundamental frequency, rise times, pulse widths and duty cycles. It’s very instructive to run through several scenarios prior to building hardware. By changing any of the user-defined parameters, you can quickly judge the outcome. More…
Many spectrum analyzers include tracking generators, but some new engineers my not completely understand what the purpose is and how they can help with certain measurements. In a nutshell, tracking generators are variable, or swept, RF generators that “track” with the spectrum analyzer sweep frequency. Some tracking generators have a limited frequency sweep as compared with their mating spectrum analyzer, so you need to know any frequency limitations in advance. In this article, I’ll use the new Rigol DSA815TG and the built-in tracking generator to measure a 150 MHz bandpass filter and a 10 dB attenuator. More…