In today’s EMC testing world testing is very often automated, using well calibrated test equipment. The operator sets up the EUT and support equipment, sets up the test equipment, fires up the software on the computer designed to automate the test, clicks on “Go” and sits back to watch the blinking lights. After a while the test is complete and a file is generated containing the results of the test. All very straightforward and “simple” to perform.
How was testing performed years ago when calibrated equipment might not be available, or only some of the equipment was calibrated? What if only the antenna was calibrated and the operator had a signal source that was also calibrated, but the receiver wasn’t, nor were the cable loss characterized, etc? Could you still perform a measurement and trust the accuracy of the amplitude information?
The answer is YES!
Back in the “old days”, before automation was common and fully calibrated receivers were not readily available (at least not where the author worked) a different measurement technique was used for all amplitude measurements for emissions. This was called, as you might guess from the title of this paper, the substitution source method. Take the measurement, note the amplitude of the signal on a display and then connect a signal source in place of the antenna and duplicate the amplitude reading. As long as you had the calibration factors for your antenna or other transducer (current probes?) and a calibrated signal source you could accurately measure the amplitude of a signal. It was time consuming, but it worked well. Frequency determination was still at the mercy of the receiver, but you could tell closely enough to make a pass/fail determination.
How was this done?
Substitution Source Method
The method for substitution source measurements was simple to explain, but laborious to perform. The tester first connected the antenna or other transducer (such as a current probe) to the end of the cable connecting it to the receiver. The antenna was placed appropriately for the test and the frequency range of interest was scanned. When an emission from the EUT was found its amplitude on a meter on the receiver, or on an attached oscilloscope, was noted. Very often where the author worked the IF gain of the receiver would be adjusted to provide an easy to remember reading on an oscilloscope display. Any belief that the amplitude indication of the receiver was calibrated went right out the window with a variable IF gain control. The frequency indicated by the receiver was written down. Depending on the resolution of the display of the receiver this might only be one or two significant digits to the right of the decimal
point. 100.15 MHz or 100.1 MHz, for example. The antenna or other transducer was then disconnected from the end of the transmission line and the calibrated signal source was connected in its place. The amplitude of the signal from the signal source was then adjusted to match the amplitude noted on the meter or oscilloscope. The amplitude indicated by the signal source was then corrected by the calibration factor for the antenna or transducer and written down. Note that the calibration of the amplitude indication on the receiver or oscilloscope was not needed, nor was it necessary to know the insertion loss of the cable to the receiver as both measurements (transducer and signal source) were performed at the source end of the cable. Only the antenna factor (or transducer factor) and output level of the signal source were needed.
For example, if the signal source used was an impulse generator and read 65 dBuV/MHz when the amplitude indication on the receiver matched what was seen from the emission and the antenna factor was 20 dB/m, the amplitude of the emission would be 85 dBuV/m/MHz. Simply add the two numbers. Very simple and straightforward, just time consuming to perform.
This process would be repeated for each emission found from the EUT. As many times as necessary to measure all noted emissions from the EUT.
The author has seen this process automated at a commercial EMC laboratory. The signal generator and RF switch were located in a shielded box in the floor of a 10 meter RF semi-anechoic chamber. The switch and signal generator were remotely controlled to allow the process to be automated.
That was a lot of work!
You bet it is. In addition to all the connecting and disconnecting a signal cable from the antenna and signal source there was a bunch of manual data crunching. All that connecting and disconnecting of cables took its toll on RF connectors on cables. A side benefit was that laboratory personnel became very adept at installing new BNC or N connectors on cables! Today’s automated testing using fully calibrated test systems is so much faster and easier on laboratory equipment and cables. However, somewhere along the line you might have to make a measurement or two when all that you have in the way of calibrated test equipment is an antenna and a signal source. Isn’t it nice to know that you could accurately measure the amplitude of the signal with only that equipment?A