Predict Conducted Susceptibility
Expose an electrical cable to an electromagnetic field from a cell phone, radio transmission, radar, or any other field source, and the cable conductors will pick up energy from the field and carry it into the equipment at the ends of the cable.
Circuitry inside the equipment must be able to withstand the induced voltage and current without being upset or damaged. In a nutshell, that is what conducted susceptibility* is all about.
How do you design circuitry that can tolerate induced noise?
Let’s say you are designing a switching power supply, or maybe you are purchasing an off-the-shelf power supply module.
You need to make sure it will pass CS (conducted susceptibility) testing and will be immune to interference coupled to its cables when in use.
Determine the threat level
CS test levels and methods are specified by several regulatory agencies, so determining frequency range and amplitude of the test signals is usually straight-forward. It is a matter of obtaining the correct specification document and identifying the class or category for your equipment. An accredited EMI laboratory or your procuring agency can help you identify the correct requirement.
Assess Circuit Sensitivity
Often, you will not know the sensitivity of your circuits precisely. Unless you take the time to test each circuit over the required frequency range, and determine its susceptibility threshold at each frequency, you will have to use your engineering judgement, and make an educated guess.
One rule of thumb is to assume that the CS signal that reaches the end circuit will be rectified and will present as a dc offset that is added to the normal circuit signal. The amount of offset that the circuit can tolerate without being adversely affected is the circuit sensitivity.
The difference between circuit sensitivity and the voltage induced on the cable by the injected CS test signal defines the amount of protection required. Usually needed protection varies with frequency.
An EMI filter is the most common method of protection. The filter must attenuate the induced signals to levels that are below the circuit sensitivity.
CS Signals Are Injected on Cables
During CS tests, a current injection probe or current clamp is placed around each cable bundle to inject the specified test signal, usually a modulated sine wave that is swept across a range of frequencies.
At first glance, doing CS calculations seems simple enough. However, CS test signals are affected by several factors.
– Test signal limit and threshold
– Cable resonances and shielding
– Injection point
Limits versus Threshold
Most CS test methods have a limit and a threshold. The way that limit and threshold are specified has a big effect on the noise levels that are actually injected.
For example, MIL-STD-461F, CS114 states that the drive level required to obtain the limit current in a 100-ohm test fixture is to be applied to the test cable, but that current induced on the cable during testing must not exceed two times the current limit.
Limits and threshold are specified differently by different agencies and for different test methods.
Cable resonances and shields
Cables resonate due to many factors, including length, height above ground, and interactions with EMI filers and end circuit components. Resonant effects can be complex, and are often the cause of CS test failures.
Cable shields offer CS protection, and for some circuits a cable shield is all that is needed. Shields have their own resonances, however, and their effectiveness is highly dependent on termination method. Many cables cannot be shielded.
CS test signals cover a broad range of frequencies, typically from a few hertz to several hundred megahertz. At high frequencies, where cable length is greater than about one-tenth of a wavelength, the location of the injection probe along the length of the cable affects cable resonance and therefore affects the noise levels that each the end circuits.
The way the injected signal interacts with the cable conductors and the end circuits affects the outcome of the test.
Sharpen your pencil. The math is complicated.
Without good software models, accurately calculating conducted susceptibility is not practical.
Performing CS calculations correctly requires treating the cables as transmission lines. Adding to the difficulty is the complex impedance of the end circuits, which varies over frequency. If you are analyzing a multi-conductor cable, such as a wire pair over a ground plane, forget about doing the calculations by hand.
Many factors affect radiation
There are so many design elements that affect CS immunity, the only practical way to predict noise levels that reach the end circuits is with software like EMI Analyst.
As a minimum the conducted susceptibility analysis model must account for:
– Cable length and twisting
– Height above ground
– Circuitry at both ends of the cable
– Injection probe placement
– Frequency range
– EMI filter insertion Loss
A Quick Example
The graph below plots voltage induced at the input to a switching power supply, after the CS-induced voltage is attenuated by the input power line EMI filter. Analysis properties are summarized below.
Maximum induced voltage is just 63.3 dBuV (1.46 mV). For a switching power supply this level of interference is negligible, so we can be confident that the input power EMI filter will provide adequate CS protection.
Modeling the cables, EMI filter, and end circuits using EMI Analyst took about 3 hours. Calculations took about 5 minutes.