June 4, 1998
Sniffer probe locates sources of EMI
Bruce Carsten, Bruce Carsten Associates Inc
A miniature EMI "sniffer probe" and an oscilloscope can help
to locate and identify magnetic-field sources of EMI. Small size with relatively high
sensitivity and electrostatic shielding enable the probe to provide much more information
than other available probes.
Electromagnetic interference (EMI) can be difficult to locate and
correct in electronic equipment, but a small "sniffer probe" can help. The EMI
sniffer probe's original and primary use was to diagnose sources of EMI in switch-mode
power converters. However, the probe is also useful for high-speed logic systems and other
electronic equipment. For applications that have many adjacent and similar sources of EMI,
you can use a related "high-discrimination" sniffer probe to further localize
the EMI-generating currents.
Although a number of other H-field probes are available, they are
generally large, ranging from 1 cm in diameter upward, and do not always have sufficient
shielding against spurious pickup of electric fields.
The principal advantages of the miniature multiturn probe over other
H-field probes and simple pickup loops include a spatial resolution
of approximately 1 mm; relatively high sensitivity for a small coil; a 50ohm source termination, which minimizes cable reflections with unterminated scope
inputs; and Faraday shielding, which minimizes sensitivity to electric fields.
Sources of EMI
Rapidly changing voltages and currents in electrical and electronic
equipment can easily result in radiated and conducted noise. Switch-mode power converters,
for example, generate the most EMI during switching transients, when power transistors
turn on or off. You can readily use conventional oscilloscope probes to see dynamic
voltages, which are the principal sources of common-mode-conducted EMI. High dV/dt signals
can also feed through poorly designed filters as normal-mode voltage spikes and may
radiate fields from a circuit without a conductive enclosure.
Dynamic currents produce rapidly changing magnetic fields, which radiate
far more easily than electric fields, because magnetic fields are more difficult to
shield. These changing magnetic fields can also induce low-impedance voltage transients in
other circuits, resulting in unexpected normal-mode- and common-mode-conducted EMI.
Voltage probes can't directly sense these high dI/dt currents and the resultant fields,
but the sniffer probe can. Although current probes can sense currents in discrete
conductors and wires, they are of little use with pc traces or for detecting dynamic
magnetic fields.
Construct the probe
The basic probe consists of a miniature pickup coil of typically 10 turns in the
end of a small shielded tube with a BNC for attachment to a coaxial cable (Figure 1). The sniffer probe's output voltage is essentially
proportional to the rate of change of the ambient magnetic field and thus to the rate of
change of nearby currents.
A high-discrimination probe has a similar construction, differing primarily in
its use of a dual coil and its use of a shield tube without "flux-entry" saw
slots (Figure 2). For this probe, all flux comes in from
the open end of the shield tube.
Few of the miniature probe's construction details are critical; you can
modify the probe's coil-wire size, diameter, and number of turns without a significant
impact on perform-ance. However, three aspects are important:
outer-probe insulation to avoid short circuits while probing; one (or preferably two) thin
saw slot in the coil end of the shield tube to allow entry of the magnetic field at high
frequencies; and a ground connection between the shield tube and the exposed end of the
pickup coil near the shield end, which minimizes E-field pickup.
The 50ohm internal terminating resistor is not essential. However, if you don't include
this resistor, you must then use the probe with a 50ohm load
termination to avoid transmission-line resonance, which completely distorts the pickup
waveform.
Probe response characteristics
The basic sniffer probe is sensitive to magnetic fields only along the
probe's axis. This directionality is useful in locating the paths and sources of high
dI/dt currents. The resolution is usually sufficient to locate the pc-board trace or
component-package lead that conducts the EMI-generating current.
For isolated single conductors or pc-board traces,
the probe's response is greatest just to either side of the conductor where the magnetic
flux exists along the probe axis (Figure 3a). The probe's
response may be a little greater with the axis tilted toward the center of the conductor.
A sharp response null occurs in the middle of the conductor, with a 1808 phase shift to
either side and a decreasing response with distance. The response increases at the inside
of a bend where the flux lines crowd together, and the response is lower on the outside of
a bend where the flux lines spread apart.
When the return current is in an adjacent parallel conductor, the
probe's response is greatest between the two conductors (Figure
3b). A sharp null and phase shift occur when the probe is over each conductor.
When you move the probe outside the conductor pair, the peak response goes down and
continues to decrease with distance.
The response to a trace with a return current on the opposite side of
the board is similar to that of a single isolated trace except that the response may be
greater when the probe's axis tilts away from the trace. A ground plane below a trace has
a similar effect, because there is a counterflo wing
"image" current in the ground plane.
Test the sniffer probe
You can functionally test the sniffer probe using a small coil and
current-sensing resistor (Figure 4a). The 12.4ohm resistor in parallel with a termination impedance of 50ohm provides close to a 10ohm current shunt, with
the impedance rising about 10% (+1 dB at 50 MHz). For this test, you center the EMI probe
in the coil for maximum response and calculate the field intensity using the following
equation:
B=H=1.257×N×I/l,
where N is the number of turns of the test coil, I is the current in the
coil, and l is the length of the coil.
For testing between 1 and 500 MHz, you can use a 1.5-turn coil in the
center conductor of a coaxial transmission line between a generator output and a 50ohm oscilloscope or voltmeter input. For a 1.27-cm-long, 20-turn test coil, the flux
density is about 20 gauss/A. At 1 MHz, the sniffer probe's voltage is 19 mV p-p ±10% per
100 mA p-p for a 1-Mohm load impedance. A 50ohm load impedance reduces this voltage by half.
The setup in Figure 4a allows you to
test the probe's frequency response to a uniform magnetic field (Figure
4b). Because of large variations in field strength around a conductor, you should
consider the probe only as a qualitative indicator and make no attempt to calibrate it.
The response's roll-off near 300 MHz results from the pickup coil's inductance of 75 nH
driving the total terminated impedance of 100ohm; transmission-line
reflections cause the mild resonant peaks, with a 1-Mohm scope
termination, at multiples of 80 MHz.
The response of the high-discrimination probe is quite different (Figure 2b). The response has a sharp peak when the probe
centers on the trace with one coil to each side of the center line. This probe has an
extremely small response to ambient magnetic fields, so it is of limited use in the
initial localization of EMI sources.
How to use the sniffer probe
Using the sniffer probe requires at least a two-channel oscilloscope:
one channel to view the noise whose source you want to locate, which may also provide the
scope trigger, and the other channel for the sniffer probe. The probe's response nulls
make it inadvisable to use the probe's channel for triggering.
A third scope trigger channel or an external trigger can be useful,
particularly if triggering on the noise is difficult. Transistor-drive waveforms or their
predecessors in the upstream logic are ideal for triggering; these waveforms are usually
stable and allow you to see events that are immediate precursors of the noise.
Start with the probe at some distance from the circuit, with the probe's
channel set at maximum sensitivity. Move the probe around the circuit, and "sniff around" for something happening in precise synchronization with
the noise transient. A precise time-domain correlation between EMI noise transients and
internal circuit fields is fundame ntal to the diagnostic
approach. The probe's waveform is not identical to the noise transient, but the waveform
usually has a strong resemblance to the transient.
As you locate a noise-source candidate, move the probe closer while
decreasing the oscilloscope's sensitivity to keep the probe's waveform on-screen. Then,
you should be able to quickly bring the probe down to the pc-board trace (or wiring) where
the probe signal seems to be a maximum. This spot may not be near the point of EMI
generation but should be near a pc-board trace or other conductor carrying the current
from the EMI source. You can verify the location by moving the probe back and forth in
several directions; when the probe crosses the appropriate
pc-board trace at roughly right angles, the probe output goes through a sharp null over
the trace, with an evident phase reversal in probe voltage on each side of the trace.
You can follow this EMI "hot" trace--like a bloodhound on the
trail of a scent--to find most or all of the EMI-generating current loop. Trace out the
noise-current path as much as possible, and identify the current path on the schematic. If
the trace is hidden on the back of or inside the board, mark its path with a felt pen, and
locate the trace on disassembly, on another board, or on the artwork. From the current
path and the timing of the noise transient, the source of the problem usually becomes
almost self-evident.
Typical dI/dt EMI problems
The sniffer probe diagnoses several problems--some of which are more
common than others. Familiarizing yourself with these problems and their suggested fixes
may help you to more quickly solve your own EMI problems.
In power converters, reverse recovery of rectifiers
is the most common source of dI/dt-related EMI; the charge that P-N junction diodes store
during conduction causes a momentary reverse-current flow when the voltage reverses. For
diodes with a "snap" recovery, which is more likely in devices that have a PIV
rating of less than 200V, this reverse current may stop in less than 1 nsec.
Alternatively, the reverse current may decay more gradually with a "soft"
recovery (Figure 5).
The sudden change in current creates a rapidly changing magnetic field,
which causes the radiation of external fields and induces low-impedance voltage spikes in
other circuits. This reverse recovery can shock parasitic L-C circuits into ringing, which
produces oscillatory waveforms with varying degrees of damping when the diode recovers. A
series R-C damper circuit in parallel with the diode is the usual solution.
Output rectifiers generally carry the highest currents and are thus the
most prone to this problem. Designers often anticipate this problem and design sufficient
snubber networks for these rectifiers. Thus, catch or clamp diodes with no snubber
networks can be more of an EMI problem. The fact that a diode in an R-C-D snubber may need
its own R-C snubber is not always self-evident, for example.
You can usually identify the problem by placing the sniffer probe near a
rectifier lead. The signal is strongest either on the inside of a lead bend in an axial
package or between the anode and cathode leads in a TO-220, TO-247, or similar type of
package (Figure 5).
Using softer recovery diodes is a possible solution, and Schottky diodes
are ideal in low-voltage applications. However, unlike a diode with a snap recovery, a P-N
diode with soft recovery is also inherently lossy because the diode simultaneously
develops a reverse voltage while still conducting current. The fastest possible diode,
which results in the lowest recovered charge with a moderately soft recovery, is usually
the best choice. Sometimes a faster, slightly snappy diode with a tightly coupled R-C
snubber works as well as or better than a soft but excessively slow recovery diode.
If significant ringing occurs, a quick-and-dirty R-C snubber design
approach works fairly well; place increasingly large damper capacitors across the diode
until the ringing frequency reduces by half. The total ringing capacity is now quadruple
that of the original ringing frequency, or, to put it another way, the original ringing
capacity is one-third of the added capacity. The necessary damper resistance is
approximately equal to the capacitive reactance of the original ringing capacity at the
original ringing frequency. Then, you connect the frequency-halving capacitor in series
with the damping resistance and place the pair across the diode as tightly coupled as
possible.
Snubber capacitors must have a high pulse-current capability and low
dielectric loss. Temperature-stable-disc, multilayer-ceramic, silvered-mica, and some
plastic film-foil capacitors are suitable. Snubber resistors should be noninductive;
metal-film, carbon-film, and carbon-composition resistors work well, but you should avoid
using wirewound resistors. The product of the damper capacity, switching frequency, and
the square of the peak snubber-capacitor voltage provides an estimate of the maximum
snubber-resistor dissipation.
Snubbers on passive switches (essentially diodes) or active switches
(essentially transistors) should always have close coupling--as much as is physically
possible--with minimal loop inductance. This arrangement minimizes the radiated field that
arises when the current path changes from the switch to the snubber. Close coupling also
minimizes the turn-off-voltage overshoot that's necessary to force the current to change
paths through the switch-snubber loop inductance.
Leakage-inductance fields
Transformer leakage-inductance fields emanate from between primary and secondary
windings. Single primary and secondary windings create a significant dipole field, which
you can see by placing the sniffer probe near the winding ends (Figure
6a). If this field is generating EMI, there are two
principal fixes: split the primary or secondary in two to sandwich the other winding or
place a shorted, copper-strap electromagnetic shield around the complete core and winding
assembly (Figure 7). Eddy currents in the shorted strap
largely cancel the external magnetic field.
The first approach, which creates a quadrapole instead of a dipole
leakage field, significantly reduces the distant field intensity. This approach also
reduces the eddy-current losses in any shorted-strap electromagnetic shield you use, which
may be an important consideration.
External-air-gap fields
External air gaps in an inductor, such as those in open bobbin-core
inductors or those with E cores spaced apart, can be a major source of external magnetic
fields when significant ripple or ac currents are present (Figure
6b). The sniffer probe can also easily locate these fields; the response is a
maximum near an air gap or near the end of an open-inductor winding.
Open-inductor fields are difficult to shield, and, if they present an
EMI problem, redesigning the inductor to reduce external fields is usually necessary.
Placing all of the air gap in the center leg virtually eliminates the external field
around spaced E cores. If eddy-current losses are low enough, the shorted-strap
electromagnetic shield of Figure 7 can minimize fields
due to a possibly intentional residual or minor outside air gap.
A less obvious problem can occur when you use inductors with open cores
as second-stage filter chokes. The minimal ripple current may not create a significant
field, but such an inductor can pick up external magnetic fields and convert them to noise
voltages or can create an EMI-susceptibility problem.
Poorly bypassed high-speed logic
Ideally, all high-speed-logic designs should include tightly coupled
bypass capacitors for each IC, and all multilayer pc boards should have power and ground
distribution planes. Unfortunately, poor design practices still exist, such as using just
one bypass capacitor at the power entrance to a logic board and routing power and ground
to the ICs from opposite sides of the board. This faulty distribution scheme creates large
spikes on the logic supply voltage and produces significant electromagnetic fields around
the board.
With a sniffer probe, you can determine which pins of which ICs have the
largest current transients that are synchronous with the supply-voltage transients. In one
case, using the sniffer probe set some logic-design engineers straight. The engineers
accused the power supply of creating the noise. However, the probe helped to determine
that the supplies were fairly quiet; it was the poorly designed logic power-distribution
system that was the problem.
A sample test setup uses the sniffer probe with a line-impedance stabilization
network (LISN) (Figure 8). The optional LISN ac-line
filter reduces ac-line-voltage feedthrough from a few hundred millivolts to microvolt
levels, simplifying EMI diagnosis when a suitable dc voltage source is unavailable.
The electrostatic Faraday shielding of the sniffer probe is excellent, despite
the open end of the probe. This end of the pickup coil connects to ground to enhance
shielding. The spurious capacitive pickup is only about 4 fF, or 0.004 pF, based on the
measured capacitive feedthrough and the shielded coil inductance of 75 nH (Figure 9a). The effect is so slight that you can ignore it
in virtually all applications. The capacitive response is actually very difficult to
measure; the test requires a special test jig to minimize pickup of associated capacitive
displacement currents in the vicinity while maximizing the true capacitive coupling with a
flat surface--typically a 0.18-in.-diameter disc--against the open end of the probe tip.
Because of the inductive loading of the pickup coil--at least below 200
MHz and even with a 50ohm termination--the capacitive response is not proportional to the derivative of
the voltage, dV/dt, but to the second derivative of the voltage. You can see this effect
in the probe's response to a voltage slewing 600V with a 90 to 10% fall time of 5.2 nsec (Figure 9b). The probe's capacitive response is proportional
to the second derivative, or curvature, of the voltage wave, and not to the first
derivative, or slope. This response is greatest at the bottom "corner" of the
input-voltage waveform, where the probe shows its peak response. Slower voltage changes
result in a dramatic decrease of the response.
The spurious capacitive response of the probe can often appear much
higher because of capacitive displacement currents near the probe. Displacement current is
proportional to dV/dt, and the probe is sensitive to the displacement current, dI/dt,
which is proportional to the second derivative of the voltage. Thus, the
displacement-current response is the same as that for true parasitic capacitive coupling.
Finally, you may wonder about the sniffer probe's ability to inject a
signal. You can use some EMI-sensing probes to test for EMI susceptibility by injecting a
current into the probe and placing it near potentially sensitive circuits. However, this
miniature probe is not particularly suitable for this application because of its small
coil and limitation to low drive levels; more than a 1/8W input, which is equivalent to
2.5V rms, can overheat the internal terminating resistor. |