When a device or subsystem is tested for the requirements of MIL-STD-461, it is important to understand the types of tests and measurements which will be made. Each test will determine compliance of a different aspect of the equipment.
CE101 and CE102 Conducted Emissions Testing
In the case of conducted emissions, the two principal tests are CE101 and CE102. Unlike earlier versions of MIL-STD-461 and some other standards, these tests are only on power lines, not on other interconnecting cable bundles. CE101 is a low frequency test, below 10 kHz. The test is performed with a current probe, meaning the measurement will look for audio frequency currents created on the power lines individually. Doing these individually is important since lower frequency emissions tend to be differential mode, meaning that the currents flow in opposite directions in the power lines. To determine if the emissions are differential mode, simply place the current probe on both the power and power return together, and any differential mode energy will cancel out because they are flowing in opposite directions inside the probe.
For CE102 testing, the measurement is made from 10 kHz to 10 MHz. Also, the measurement is from the 50 Ohm port of the LISN placed on the power line. As a result, the measurement is a voltage measurement, as indicated by the units of the limit, which are in dB?V.
Depending on the impedance of the circuit, the emissions results from CE101 and CE102 may be radically different. A low impedance, high current device may generate significant radio frequency currents but due to the low impedance, lower interfering voltage than those from a high impedance device. However, these issues tend to be normalized with the use of the Line Impedance Stabilization Network (LISN) when loaded into a high quality 50? termination.
RE101 and RE102 Radiated Emissions Testing
Most common radiated emission tests are RE101 and RE102. Like the conducted counterpart, RE101 is a low frequency test (up to 150 kHz), whereas RE102 requirements are up to 18 GHz. Also, RE101 will measure mainly magnetic fields, where RE102 will be more sensitive to electric fields and plane waves.
The distinction of these two tests is based on the source of the energy. For magnetic fields, currents are the source, and loops of wires, such as on transformers and inductors, will increase the fields. Also, it is difficult to shield low frequency magnetic fields, since below about 100 kHz, shielding requires a permeable material, and possibly significant thickness for adequate field reduction. For this reason, it is better to avoid the generation of uncontrolled magnetic fields. To do so, use closed core inductors such as toroids, and avoid open cores such as rods or excessively gapped cores. Assure any high current source traces or wires are closely coupled with their own returns.
For electric fields in the radio frequency range, shielding can be easier to obtain, if the chassis is metal, and the mating surfaces of the metal is clean, conductive, and well bonded. Of critical importance is the filtering of all conductive lines in and out of the chassis. Any wire that penetrates the shield will have to be filtered to prevent the coupling of energy from the inside to the outside of the chassis. This may be the most common source and the most neglected cause of radiated emissions.
For an object to efficiently radiate electric fields, the radiating element should be approaching the size of a quarter wavelength of the frequency being transmitted. For MIL-STD 461G, interconnecting cables shall be a minimum of 2 meters long, unless the installation is known to be shorter. For a 2-meter cable, the full wavelength frequency is 150 MHz. The cables become the significant radiation source. Since most devices tested to this standard are smaller than these dimensions, they are more likely to radiate at higher frequencies than the cables do. Also, the chassis of the device are typically solid metal and bonded to the ground plane or shielded enclosure, whereas the cabling is not. Even cables with shields must have the shields well bonded at both ends to assure minimal radiation, but even then can become effective radiators at the half wave frequency. The source cable can be verified by using a current probe on the cable bundles. Often, up to 200 MHz or more, the radiated emission profile can be observed as common mode conducted emissions on the offending cable or cables.
Thus, the filters used on interconnect cables are of great importance and should be designed to handle the very high frequencies involved. Also, their location should be as close to the point of shield penetration as possible to avoid cross coupled energy onto the previously filtered lines.
Reciprocity is an aspect of EMC – what cures emissions will also work for susceptibility testing. Thus, the techniques used to control radiated emissions will also work for radiated susceptibility. The same is true for conducted tests. The areas which are not addressed are impulse and surge tests (CS115, CS116, CS117) and electrostatic discharge (CS118). The former are best controlled through the use of transient suppression to chassis, or by shielded cables which can shunt the currents away from protected cables. For ESD, conductive chassis shielding is best employed, which is also beneficial for RE102 and RS103 testing.