How to avoid detecting the current signal from the probe housing

Oscilloscope probes have two wires, one for connecting the test circuit to the vertical amplifier of the oscilloscope (called the sensing wire) and the other for connecting the oscilloscope chassis ground and the digital logic ground of the local circuit (called the shield wire). Usually, we only need to consider the oscilloscope's response to the sensing wire voltage. This section analyzes how the oscilloscope responds to the signal on the shield wire.

Any voltage difference between the oscilloscope chassis ground and logic ground can induce current in the shield wire. In Figure 3.17, the shield wire current flowing through the shield wire resistance Rshield creates a voltage drop Vshield. The center conductor of the probe cable, which is the sense wire, does not conduct shield current, so there is no voltage drop across it.

When both the sense and shield wires are connected to the ground of the working circuit, the different voltage drops on the two wires will be reflected as a voltage difference between them on the vertical amplifier of the oscilloscope. We have no way of knowing whether this voltage difference is caused by the actual signal at the far end of the probe cable or by the shield current. Although we hope that the oscilloscope will show no voltage, it can show the shield voltage.

The oscilloscope responds to the shield voltage as if it were a real signal.

The shield voltage is proportional to the shield resistance, not the shield inductance. This is because the shield conductor and the center conductor are magnetically coupled. Any changing magnetic field generated by the current flowing through the shield loop surrounds the shield conductor and the center conductor together, inducing the same voltage on both conductors. The induced voltage exists on both conductors at the same time, while the resistance voltage drop only appears on the shield.

The shield voltage is easily observed:

1. Connect the probe tip of the oscilloscope to the ground wire.

2. Move the probe near the working circuit without touching anything. At this time, only the magnetic induction detection voltage from the induction loop of the probe can be seen.

3. Wrap the end of the probe with aluminum foil and short-circuit the detection contact directly with the metal grounding sheath of the probe. At this time, the magnetic field detection voltage is reduced to near zero.

4. Now put the oscilloscope probe to the logic ground of the test circuit. You should only see the voltage on the shield line. If the shield voltage is very small, it can be ignored.

Shield noise can cause trouble for digital systems that control high-power devices. The huge 60Hz AC current in the device will induce voltage on the digital logic ground, which in turn causes shield noise. If the shield voltage is causing trouble, there are 9 ways to overcome it.

1. Reduce the shield resistance. If the probe is purchased, this is more difficult to do. If you use a homemade coaxial cable probe, then choose a thicker coaxial cable. Change from RG-174 to RG-58, or from RG-58 to RG-8. The hardness of the thicker coaxial cable makes this method unrealistic, but it solves the problem of instrumentation.

2. Add a bypass impedance between the oscilloscope and the logic ground. This causes most of the noise current to flow through the bypass impedance, while a small amount of current flows through the shield. This method is usually not practical, especially for high frequencies. It is almost impossible to pick a good ground point from the test circuit board and connect it to the oscilloscope ground through a conductor with low enough inductance to obtain a significant improvement.

If the bypass conductor is the same length as the probe lead, there is no conductor of large enough diameter that will improve the problem (inductance varies as the logarithm of diameter). If the bypass conductor is much shorter than the length of the oscilloscope probe lead, it may work.

3. Power off the test circuit board, or partially power off. This method is only effective when testing a local circuit. If you suspect that the problem comes from shielding current noise, this is a good test method. This can be used to determine whether the noise is generated by the test circuit or caused by other interference sources.

4. Connect a large inductor in series in the shielding loop, use a large high-frequency magnetic core, and wrap the probe around it 5 to 10 times. This will increase the inductance of the probe shielding layer, thereby reducing the current. This method works better in the range of 100KHZ~10MHZ. Below 100KHZ, a large inductor is required to improve the problem. Above 10MHZ, the effect of the magnetic core will decrease.

5. Redesign the circuit board to reduce electromagnetic field radiation. Change the two-layer board to a four-layer board and add a complete ground plane. Reducing electromagnetic field radiation is the primary way to reduce the tendency of the ground plane to generate noise voltage.

6. Disconnect the safety ground wire of the oscilloscope. Disconnecting the safety ground wire of the oscilloscope invalidates the safety feature of the oscilloscope's AC power system. Once the power supply part of the oscilloscope's power system is connected to the chassis, the oscilloscope's chassis becomes connected to the 110V power supply voltage, which is an unsafe voltage. Normally, if a fault occurs, the safety ground wire bypasses most of the AC power current to the ground and triggers the circuit breaker protection switch of the circuit. It cuts off the power and may well save your life at a critical moment.

Nevertheless, you should be aware of the effect that leaving the safety ground disconnected has on high-frequency signals.

If the oscilloscope case and the safety ground wire are well isolated, the shield ground loop of the oscilloscope probe will be cut off, thereby reducing the shield current. Unfortunately, however, disconnecting the safety ground wire does not provide good isolation.

For most oscilloscopes, there is a 0.01uF capacitor in the internal power supply section that connects the case to each AC power line, which in turn connects it to ground. Even if this capacitor is not used, there will be enough parasitic capacitance on the power transformer to form a high-frequency energy path between the case and the power supply.

At frequencies above 10 MHz, the oscilloscope has enough capacitance to ground anyway, so simply disconnecting the safety ground wire has no effect.

If you want to make your own 21:1 double shield probe, POMONA sells a BNC to double shield conversion adapter that can be used for this purpose. Connect the BNC plug end of the adapter to the BNC socket of the oscilloscope. The other end of the POMONA adapter has a double shield jack, which can connect the outer and middle ground wires internally to a single BNC ground wire. Use a regular double shield plug to terminate this double shield cable and plug it into the adapter. At the circuit board end of the double shield cable, just solder the outer and middle layers together.

8. Use a 1:1 probe instead of a 10:1 probe. A 10:1 probe does not attenuate the shield voltage. Since a 10:1 probe attenuates the actual measured signal, using a 10:1 probe makes the shield voltage appear 10 times larger.

9. Use a differential probe solution. Figure 3.18 shows the correct way to achieve differential measurements. Probe 14 is connected to the signal point and probe 2 is connected to the signal ground. The shields of the two probes are connected together at point G5 and do not touch the circuit board. Use an independent ground wire to connect the circuit board ground to the oscilloscope ground. This independent wire is only necessary when the circuit board has no suitable way to connect to the real ground through other mechanisms.

Set the output of the oscilloscope to be the signal from probe 1 minus the signal from probe 2. This will require some minor adjustments to get good results. Temporarily connect both probes to the same test signal point, and adjust the gain of both probes so that the two signals completely cancel each other out. Next, temporarily connect both probes to ground and observe how much residual detection noise is present. This noise is what we are trying to overcome, so it is worth checking to see if you can catch it.

When using a differential probe, shield currents do not exist because the shield is not touching anything. This is the main benefit of a differential probe. For circuits with floating or ground voltages higher than true earth, a differential probe may be the only choice.

Place the two probes as close together as possible to minimize the size of the magnetic induction detection loop between them. Any magnetic field detected in the loop will induce a voltage between the two probes. Twist or tie the two probe wires together to ensure they are close together.

When using an ordinary oscilloscope probe, keep the ground point close to the test point. The coupling noise coupled into the sensing loop between the two probes through mutual inductance is equal to that of an ordinary single-ended probe.

To achieve a differential effect, the probes must be of the same length and type. Unequal frequency responses or delays of the two probes will cause a common-mode signal to appear on the oscilloscope display.

Some oscilloscopes offer special differential amplifier modules and probes with matching gain and efficiency response characteristics. Usually these modules have special common-mode rejection characteristics, but the bandwidth is generally too low to be very useful for solving adjustment digital problems.

Be careful with 10* probes when making differential measurements. To get good common mode rejection, high frequency compensation adjustments, like DC gain, must be very well matched. This approach is rarely useful when adjusting signals.

7. Use double shielding on oscilloscope probes. Connect one end of the outer shield of this double shield to the oscilloscope case and the other end to the circuit board. The oscilloscope probe wire must be completely wrapped in this double shield. Connect the double shield to the ground point of the oscilloscope probe. At high frequencies, most of the shield current is transferred to the outer shield due to the tidal effect. Because there is no current passing through the inner shield of the probe, there is no voltage drop and noise voltage. This method sounds counterintuitive, but it does work. The double shield can be made of aluminum foil, or the outer shield of an old RG-8 cable can be cut open and wrapped around the probe wire. The exposed length of the probe between the double shield ground layer and the probe contact should be minimized to reduce the coupling of magnetically induced noise in the loop.