While cleaning the CRT faceplate of the computer color monitor display, a person received a painful electric shock at the other hand which was touching the top of the monitor plastic enclosure. This person lost several days of work recuperating from the shock.
While listening through headphones to an audio disc in a CD-ROM drive, a person started cleaning the CRT faceplate of the computer color monitor display. That person received a painful electric shock in both ears! While carrying a monitor across a room, a person got a shock through the openings in the enclosure. (The monitor was disconnected from everything!) While talking on the phone, a person absentmindedly began cleaning the color monitor screen, and received a startling shock into the ear. These are true incidents. While they all involve color CRT, there are no geographic, weather, product model, manufacturer, or other coincidences.
Many people have received shocks from CRT faceplates. Almost everyone attributes the shock to static electricity.
Static electricity is defined as stationary electric charges, such as those resulting from friction. If the shocks are due to static electricity, what is the mechanism which generates the static electricity in the first place?
Three of the incidents involved cleaning the CRT faceplate while the monitor was on. This involves rubbing a cloth across the screen, which could be the source of the friction. However, if we perform the same cleaning operation on a window, there is no static electricity. If we perform the same cleaning operation with the monitor off, there is no static electricity.
If the phenomenon is friction-generated static electricity, then why is a charge generated only on a CRT which is on, and not from a CRT which is off or not from an ordinary window? How does the friction-generated static electricity hypothesis explain the shock incident where the monitor was off, and was being carried across the room? How does the static electricity hypothesis explain the phenomenon as only occurring on color CRT and not on monochrome CRT?
Clearly, traditional static electricity concepts do not explain the observed phenomena.
Insulation can be modeled as a parallel circuit of a capacitor, a resistor, and a spark-gap.
The principal characteristic of an insulator is that, normally, the values of resistance, capacitance, and spark-gap breakdown voltage are sufficiently high (low for the capacitance element) that they can be ignored as circuit elements. Another characteristic of an insulator is that, often, it has a conductor at only one terminal of the model; the other terminal of the model is a “phantom” conductor. Nevertheless, the insulator ALWAYS behaves as if a conductor actually exists at the “phantom” terminal AND that the “phantom” terminal is connected to some other circuit element which completes the circuit.
As is the case for insulation resistances, capacitances, and breakdown voltages, usually the circuit elements at the “phantom” terminal can be ignored. However, in the case of electric shock from the CRT faceplate, if we ignore the “phantom” terminal of the CRT faceplate insulation, and if we ignore the remaining circuit, then we cannot explain how the shock occurs.
Circuit for electric shock from CRT faceplate
Clearly, for an electric shock to occur, there must be a circuit — a voltage source and a current path. For an electric shock from the CRT faceplate, the circuit must include a voltage source, the faceplate insulation, AND a circuit path to ground.
Furthermore, we have two, DIFFERENT circuits:
- The circuit WITHOUT the body as a circuit element.
- The circuit WITH the body as a circuit element.
The voltage source for a shock from a CRT must be the CRT anode voltage. There is no other voltage source that is in the region of the CRT faceplate. The anode of a CRT is the inside surface of the faceplate glass. The anode voltage is regulated (constant) dc, with a variable ac current depending on what is being displayed. If the display is black, then there is no ac or dc current, but voltage is present at the anode. The circuit (without the body) is comprised of two insulators in series: One insulation is that of the faceplate glass, the other is that of air!
For the purposes of this analysis, we can ignore the insulation resistances of both the glass and the air. And, we can ignore the spark-gap of the glass. (The values of the glass insulation resistance and of the glass spark-gap are sufficiently high that they can be ignored.)
This leaves us with two capacitances in series, and a spark-gap in parallel with the capacitance of the air. For now, we will ignore the air spark-gap.
First, let’s look at the relative values of the two capacitors. The dielectric constant of glass is about 6.5, and the dielectric constant of air is nearly 1. Assuming the areas and distances of the two plates of each capacitor are equal, then the capacitance of the glass is 6.5 times the capacitance of the air.
Next, let’s look at voltage distribution of two capacitors in series. In a series circuit of two capacitors, voltage divides inversely according to the value of the capacitance. Since the air is the smaller of the two capacitors, then more voltage will appear across the air than across the glass. If we assume the areas and distances of the two plates are equal for both the glass and the air, then about 87% of the voltage appears across the air, and 13% of the voltage appears across the glass.
This means that about 26 kilovolts is at the outer surface of the CRT faceplate!
Let’s examine whether it is reasonable to assume that the two capacitors have plates of equal area and equal distance.
We assumed the plates of the capacitors are of equal area. Since the anode area is that of the entire CRT faceplate, then the glass “phantom” terminal area is equal to the anode area. Therefore the “phantom” terminals of the air are equal to that of the “phantom” terminal of the glass (which is equal to the anode terminal area).
We assumed the plates of the glass and air capacitances are of equal distance apart. CRT faceplate glass is about 1 inch thick. If we bring a small grounded object to about 1 inch from the CRT faceplate, we will note a crackling sound. This indicates a transfer of charge, which means a change in the capacitance in air. So, 1 inch of air is in the ball park.
If we presume the distance in air is two inches, then the voltage distribution is 76% across the air, and 24% across the glass. In this case, we still have almost 23 kilovolts at the outer surface of the CRT faceplate!
Now, let’s examine what happens when someone touches a finger to the CRT faceplate.
In this system, the body resistance is very much less than the insulation resistances of the glass and air. So, we can treat the body as being a conductor of negligible resistance.
When a finger approaches the CRT faceplate, it displaces the air in the vicinity of the faceplate. This air, before the finger approaches the faceplate, is holding a charge. As the finger approaches the faceplate, the charge in the air being displaced by the finger AND the associated charge in the glass must be conducted to ground by the finger. Because the finger establishes the area of the plate of the capacitor, the charge conducted to ground by the body is very small and usually not detectable as an electric shock sensation.
In some cases, it is possible to detect a crackling sound as the finger approaches the CRT faceplate.
However, if a hand is placed on the CRT faceplate, the capacitance, being established by the area of the hand, is many times the area of the finger, and a very much larger charge must be transferred to ground. This larger charge often will result in the unpleasant sensation of electric shock.
If the body is insulated from ground when it touches the CRT faceplate, then the body replaces the air as the capacitor to ground. In this case, the body charges to some value of voltage. Usually, the body is near a ground such that an arc will occur from the body to the ground when the voltage is sufficient to overcome the breakdown voltage of air for that distance. This explains how an electric shock occurs from the ear to headphones or to a telephone.
Note, however, no matter what the shock mechanism, only one shock will occur. Once the air and glass are “discharged,” they can only re-charge by means of the insulation resistance of the glass. Since the insulation resistance of the glass is so very high, and the capacitance of the air so very low, the time to re-charge the capacitance to a significant voltage is of the order of tens of hours.
Let’s now look at what happens when the power is turned off.
If the charge at the CRT faceplate is left undisturbed while the high voltage was on, then, when the CRT anode voltage goes to zero, the charge at the CRT surface is drained through the power supply to ground.
However, if the charge at the CRT faceplate was drained while the CRT high voltage was on, then, when the CRT anode voltage changes from +30 kV to 0 volts, a negative charge is created at the surface of the CRT faceplate! Now, even though the power is off, if a hand is placed on the CRT faceplate, a shock may occur! This explains how an electric shock occurs when carrying a monitor.
Experiments and Measurements
To confirm these explanations, we set up some experiments and measurements.
We taped some aluminum foil to the surface of the CRT. The foil was about 10 x 20 centimeters (the IEC 1010 standard area for a hand). We used a 1,000-megohm, 3 picofarad high voltage probe and 100 megahertz scope to measure voltage and change in voltage.
When we turned on the monitor, the foil voltage went to 7 kilovolts, and then decayed to nearly zero, all in a period of 100 milliseconds. Clearly, the probe resistance discharged the capacitance.
We waited, and then again measured the voltage. If we waited a long time, tens of minutes, we measured voltages on the order of 600. At shorter times, we measured correspondingly lower voltages. Clearly, the charging time constant was very long.
Then, without the probe, we turned on the monitor, and brought the probe to the foil. We heard crackling, and had a corresponding display of transients, with the highest more than 8 kilovolts. From the shape of the discharge, we can surmise that the initial foil voltage was more than 16 kilovolts.
Again, without the probe, we turned on the monitor. This time, we brought a grounded probe to the foil, and incurred a 1-centimeter arc. Then we grounded the foil, removed the probe, and turned the monitor off. Again, we brought the probe to the foil and again we incurred a 1-centimeter arc.
We repeated the grounded probe experiments with several 1.5-square centimeter foils taped onto several positions on the screen (to simulate a finger-tip). As we brought the probe to each foil, we could hear crackling, but did not observe any arc.
Each foil behaved independent of the others. That is, if we discharged one foil, the others remained charged. This means that the surface insulates each incremental area from the adjacent area.
From the measurements and other data, we calculated the value of capacitance involved. With the known resistance and capacitance of the high voltage probe, and with the peak voltage and decay time, we determined the total of the air and glass capacitances for a 10 x 20 centimeter foil at about 10 picofarads.
Using the published dielectric constant for glass, 6.5, and assuming a glass thickness of 1 inch, and using the 10 x 20 centimeter foil as the area, we calculated the capacitance of the glass at about 4 picofarads.
For such gross measurements and assumptions, we have reasonable agreement.
Is the CRT capacitive discharge injurious?
The shock is due to the high voltage and the stored charge. According to IEC 1010, for voltages above 15 kilovolts, the stored energy must not exceed 350 millijoules. If we assume 30 kilovolts, then the available energy from 10 picofarads (10 x 20 centimeter area) is 4.5 millijoules.
If we can believe that a 350 millijoule discharge is not injurious, then a 4.5 millijoule discharge is comfortably below the limit and should be non-injurious.
Nevertheless, the discharge can be felt, and is, to me, sufficiently uncomfortable that I prefer to avoid touching the CRT faceplate. If I do touch the CRT faceplate, I use my finger and approach the CRT faceplate slowly. (This bleeds the charge slowly and thus keeps the current below my level of sensation.)
Monochrome versus Color CRT
One major difference between monochrome and color CRT is the anode voltage. Typical anode voltages for monochrome CRT are about 15 kilovolts. Typical anode voltages for color CRT are about 30 kilovolts.
The originator of the insulation model used here is Ray Corson of Hewlett Packard’s Loveland Instrument Division. The testing reported here was by Kevin Cyrus of Hewlett Packard’s Vancouver Division.
Richard Nute is a product safety consultant engaged in safety design, safety manufacturing, safety certification, safety standards, and forensic investigations.