Cable Design Manual

Cable Design Manual

In this chapter, we will discuss how to deal with system cables to solve the problem of electrostatic discharge.

Electrostatic discharge (ESD) will produce electrostatic field effects, charge injection effects and electrostatic discharge current field effects, which have been discussed in Chapter 4. Correctly arranging system cables and shielding design can help solve the problems caused by these three effects.

When discussing the handling of system cables, for convenience, the three electrostatic discharge effects proposed in Chapter 4 were rearranged into the following two effects:
* Radiated noise effect * Conducted noise effect Radiated noise includes static magnetic field, electric field and magnetic field generated by discharging current. Conducted noise includes direct charge injection and current induced by electric and magnetic fields. Of course, under actual circumstances, these effects do not exist independently of each other. But to simplify the discussion, we consider each effect separately.

First discuss the cable shielding or attenuation and conduction issues, the discharge distance of 20kV voltage in the air can reach 2cm. Therefore, in order to prevent charge injection into the electronic circuit, designers can take the following three measures:
A) When designing equipment, take measures so that the operator cannot reach within 2 cm of the electronic circuit, or contact ungrounded metal objects within 2 cm of the electronic circuit;
B) Insulate all electronic circuits with a better insulating substance than air;
C) Provide another charge injection object besides the electronic circuit.
The above measures 1 and 2 will be discussed in the sixth chapter, because this is closely related to the design of the shielded chassis. And measure 3 mainly depends on the design of the system cable. As described in Chapter 2, electrostatic discharge objects must have a path to ground to eliminate electrostatic fields.
The path to ground must be kept low impedance. Otherwise, there may be a lower impedance path formed by the arc through the electronic circuit. In order to obtain a low-impedance path, the most important thing is to ensure that the impedance of the system cable is low. For most systems, the cable impedance is quite low, but at high frequencies the impedance will increase due to the skin effect. The surface area of ​​the cable core can be increased to alleviate this problem (this will be further discussed in the following cable shielding).

Another reason for the increase in cable impedance is the contact impedance at the connection point. Corrosion at the connection point can produce an impedance effect similar to the skin effect. In order to prevent corrosion and make the impedance of the cable lower, the following guidelines should be considered:
A) The objects that are in contact with each other should be compatible in the electrochemical sequence. In a humid environment, electrochemical corrosion will occur between metal objects with different potentials, the extent of which depends on the magnitude of the potential difference of the contacting objects. The greater the potential difference, the more severe the corrosion. Table 1 shows a part of the electrochemical sequence table. It should be noted that the electromotive force listed in each item may sometimes change. For example, beryllium copper can sometimes be used in conjunction with aluminum.
Depending on the environment, the electromotive force may be allowed to fluctuate greatly. In a harsh marine environment, the allowable potential difference should not exceed 0.25V. In general, it is not a problem to keep the potential difference below 0.75V except for overlapping with aluminum. For example, galvanized steel chassis can be easily connected to copper ground conductors with brass lugs.

Table 1 Electrochemical sequence table

Metal EMF (Volts)
Metal EMF (Volts)
Metal EMF (Volts)

Magnesium +2.37
Chrome +0.74
Lead +0.13

magnesium alloy
Iron or steel +0.44
brass

Beryllium +1.85
pig iron
Copper -0.34

Aluminum +1.66
Cadmium +0.40
bronze

Zinc +0.76
Nickel +0.25
Copper-nickel alloy

Tin +0.14
Lead solder
Nickel alloy

stainless steel
Silver solder
Silver -0.80

graphite
Platinum -1.20
Gold -1.50

B) There should be no continuous current flow at the contact. In addition to the electrochemical corrosion discussed above, there is an electrolytic corrosion. Electrolytic corrosion generally occurs when current flows from one metal object to another through the electrolyte. (Lapping in a humid environment can achieve electrolytic corrosion) Electrolytic corrosion can occur even between metal objects where the potential difference does not exist. However, if no current flows, electrolytic corrosion will not occur. Obviously, no constant current should flow through the chassis ground or shield connection, so this rule is easy to follow.

C) Use cathode materials as much as possible. Cathode materials, such as gold, are much more stable than anode materials. Air can easily oxidize anode materials, such as aluminum. On the other hand, everyone knows that gold is not easily oxidized. Obviously, there is some room for the choice of materials, but designers must consider the oxidizability when using materials.

If measures can be taken to control the oxidation and skin effect of the material, the main factor affecting the electrostatic discharge current is the transmission line effect of the cable at the high frequency electrostatic discharge frequency. Ideally, good impedance matching will ensure the least impact on the energy of the system leaking to ground, but in fact this is not possible. The system chassis ground not only includes its internal connection line, but also includes the safety ground of the AC power line in the house and so on. The impedance of the entire ground path must be impedance matched at each connection point, which is why it is impossible to completely achieve impedance matching in actual work.
However, despite this, the low-impedance ground path may allow a large amount of electrostatic discharge charge to flow, which prevents the charge from arcing the electronic device. Because even at the end of a completely open line, if the number of input lines is an odd multiple of a quarter wavelength, then it looks like a short circuit. Similarly, if the number of input lines is an integer multiple of half a wavelength, the end of the short-circuit line is also like a short circuit. Therefore, regardless of the type of terminal matching on the ground path, the impedance is relatively low for many frequency components .
However, for some frequency components, the impedance of the ground path looks relatively high. As mentioned in Chapter 4, it also means that the system chassis ground path must be isolated from the system electronics. Otherwise, a flashover discharge will occur between the chassis of the system and other lines. Isolation is obviously the insulation treatment of the cable. However, at the connection point, there is usually an air gap between the chassis ground and other circuit lines (the connector usually has no airtight gap).

D) The air separation between the chassis ground wire and the connector pin lead and other wires is greater than 2.2mm to prevent arcing discharge.
Unfortunately, the isolation between most connector pins is less than 2.2mm. Because the 2.2mm empirical data is more conservative, this connector can still be used. Most D-type and DIN-type connectors are very good, because the distance between the shell and the pin is about 2.2mm, so the chassis ground wire can be easily connected to the shell. Another method is to use a connector that is completely separated from other connectors in the chassis. In this case, the isolation distance of 2.2 mm is easy to obtain.
After taking the above steps to reduce the problems caused by conduction, you need to consider the problems caused by radiation. The primary radiation problem is the radiation from the chassis itself. The impedance mismatch of the system chassis path will generate standing waves. The standing waves will cause the distribution of electric and magnetic fields along the system cable. Therefore, some methods must be adopted to reduce the internal radiation. The effect of these fields on other signal lines.
Chapter 4 lists nine methods to reduce antenna coupling. Of these nine methods, only method 6 can effectively isolate the cable. Method 6 is to install a shield between the transmitter and receiver. This is the following cable design rule.

E) Use shielded cables and connect their shields to the chassis ground. Due to the skin effect, electrostatic discharge currents flow on the outer surface of the shield (this is one of the reasons why the wires show greater impedance at high frequencies). The shield has a larger surface area than ordinary wires, which will reduce the "skin effect impedance". It can also be assumed that the shield is thick enough that the electrostatic discharge current will flow along the outer surface of the shield, and the inner layer is still a shield body. Therefore, before the field reaches other wires inside the cable, the inner layer of the shield can reduce the field generated by the electrostatic discharge current.

F) The thickness of the cable shielding layer should be at least 0.025mm (in the frequency range of 1MHz-5GHz, the thickness of copper or aluminum as the shielding material does not need to be very thick)
Once measures are taken to shield the cable, it is necessary to give full play to its actual effect. Many designers take the so-called "ground loop" problem too seriously. Therefore, they insist that the cable layer should be grounded at only one end. Unfortunately, although this helps to reduce low-frequency (such as 60Hz) noise problems, it obviously makes the shield useless as an electrostatic discharge path. This requires the next cable handling measure.

G) The cable shield must be connected to the cabinet at high frequency at both ends * If a ground loop is not formed, or the ground loop is not a problem (refer to the following description), the best connection method is to directly connect the metal at both ends . An example where a ground loop usually does not form is the cable connecting the terminal to the keyboard.
* If a ground loop is formed, and problems will occur (refer to the note below). At this time, the cable shielding layer can be metally connected to the casing at one end, and the other end is connected to the casing through a high-frequency capacitor. A typical example is the interconnection between a computer and a printer. Of course, the key point is that both the computer and the printer are connected to a safe ground via an AC outlet ("green line"). This will form a ground loop between the two AC outlets.
Note: If there is no potential difference between the ground points of the electronic devices interconnected in the ground loop, then the ground loop may not be a problem. In short, the better the cable shield is terminated, the less likely the ground loop will cause problems.

H) The cable shielding layer should be connected to the chassis where the cable enters the device, and the unshielded part should be the shortest.
Note that due to the effects of electrostatic discharge current, radiated noise will be generated outside the cable shield. The shielded cable arranged near the PCB board or within the shielding body can be regarded as a Transmitting Antenna for each component in the PCB board or the shielding body. The cable shielding layer only shields the internal conductors from the interference of electrostatic discharge noise. In fact, the shielded cable also radiates electrostatic discharge noise to all equipment outside it. Similarly, a long "braid" connecting the cable shield to the chassis is also a transmitting antenna. Do not surround it near sensitive input ports. Of course, another problem caused by the long braid is to increase the impedance of the shield connection, thereby reducing the shielding effectiveness.
However, how to solve the problem of system design that does not have a chassis to connect? For example, for a keyboard with a plastic case that does not have a conductive chassis, how do you deal with connecting it to the host terminal via a six-core cable?

I) If there is no casing at one end of the cable, the solution may be: connect the cable shield to the logic ground through a high-frequency capacitor.
If the designed unit circuit has no chassis connection points, then it is not a complete design (discussed in Chapter 6). However, if you have no choice, you must logically serve as the chassis ground point at one end of the cable shield. This is by no means an ideal solution, nor is it possible to improve the immunity of electrostatic discharge. In fact, sometimes, this may cause electrostatic discharge problems instead. This method can work, the key is whether the electrostatic discharge radiation from the unshielded cable to each input port will be larger or smaller than the electrostatic discharge noise coupling from the cable shield through the capacitor to the logic ground. In general, the better the logic grounding system, the more likely this method is to improve the electrostatic discharge immunity.
The rated voltage required by the shielded coupling capacitor is directly related to the ratio of its capacitance to the capacitance of the electrostatic discharge source. Remember, for a given charge level, the voltage on an object is determined by its capacitance. The coupling capacitance is determined by the frequency to be coupled. A typical cable already has a capacitance of hundreds of picofarads between the logic ground and the shield. Therefore, the effect of adding a few hundred picofarads is not great. In order to obtain some significant effect, the coupling capacitance from the shield layer to the logic ground is at least 1000pF. On the other hand, a large capacitor has great parasitic inductance, so it cannot couple high frequencies. This means that the typical coupling capacitance should be less than 0.01μF. (A parallel connection of smaller capacitors can be used to obtain a better coupling effect, assuming that the parasitic inductance of the connection can be controlled to be small.)
A ceramic capacitor with a capacity of 3900pF and a withstand voltage of 1kV is most suitable as a coupling capacitor. This capacitor is connected in parallel with the existing capacitor of the cable, and will couple a fairly wide frequency of electrostatic discharge. Moreover, the human body capacitance is around 150pF. When the coupling capacitance of 3900pF is applied, the 20kV charge on the body surface will be reduced to less than 1kV.
The discussion on cable capacitance then brought up the problem of cable inductance filtering (typically ferrite beads). From the standpoint of electrostatic discharge, it is not always a good idea to install a common-mode ferrite filter on a well-designed shielded cable. Because any shielding is not ideal, the electrostatic discharge current on the cable shield of the ferrite will induce a reverse current on the wires in the shield. Of course, even without ferrite, there is a similar problem, but the ferrite increases the mutual inductance between the shielding layer and the inner conductor, which increases this effect. If you need to install ferrite beads on the cable (in order to reduce the emission intensity to meet the FCC and other standards of the United States Federal Communications Commission), you must comply with the following rules.

J) Electrostatic discharge current path should not pass through the ferrite together with other wires (this situation is best if the electrostatic discharge path does not pass through any ferrite)

However, if the cable is not shielded, or if the shielding layer is not effective, ESD noise is induced equally (in a common mode) into all the wires in the cable, common mode ferrites may be useful. In this case, there will be significant inductive interference on all wires in the cable. Of course, common mode ferrites will reduce this common mode interference. In this case, the following rules can be used to deal with.

K) If the ferrite bead is installed on the signal line of the cable, it is the best when it is placed on the receiving end of the signal line, which can filter the noise picked up by the signal line.

L) The extra wires in the cable must be cut off or properly connected. When there are more core wires in the cable than actually needed, use the following two methods to deal with these extra core wires:
* Cut off the excess core wire, so that no suspended wire will pass through the outside of the shield;
* Connect the extra wires in parallel with other core wires in the cable.

M) The flat cable should be provided with a logical ground wire every other wire, and the sensitive signal wire should be set between the ground wires. Even if the shielded flat cable will have some magnetic flux. Therefore, in order to minimize the loop area (refer to the discussion on loop area in Chapter 4), each signal line should be as close as possible to the ground. Moreover, sensitive wires must be as far away as possible from the edge of the cable, because these places are most prone to leakage.

Summary of Cable Design Guidelines A. In the electrochemical sequence, the overlapping voltage of two substances that are far away from each other should not exceed 0.75V.
B. The anode (mostly positive) material must have a larger unwrapped surface than the cathode material.
C. Use shielded cable and connect its shield to the chassis ground. The air distance between the chassis and the leads of other components is at least 2.2mm.
D. The shielding material must be at least 0.025mm thick. (Preferably completely shielded)
E. Connect the cable shield to the cabinet at both ends. It is best to use conductive connections, but if you need to prevent obvious ground loop problems, you need to use high-frequency (capacitive) connections.
F. The cable shielding layer must be connected to the chassis at 4 cm from the device, and the unshielded cable part must be less than 4 cm.
G. The excess cable core must be cut off or connected in parallel with other wires.
H. When the cable shielding layer also serves as the chassis ground, the ferrite beads should not normally be passed through, and the ferrite beads should never be shared with other wires.
I. If ferrite beads are used, they should be installed at the receiving end of the cable.
J. If one end of the shielding layer of the cable cannot be connected to the chassis, it can be connected to logic ground through a 3900pF, 1kV ceramic capacitor.

Transmitting Antenna

Antenna For Automotive Product,Magnetic Antenna Coil Element,Antenna Impedance Matching Transformer,Transmitting And Receiving Antenna

IHUA INDUSTRIES CO.,LTD. , https://www.ihua-coil.com

Posted on