what is the joule rating on surge protectors?

If it gets a really big surge it'll 'burn out'. For obvious reasons ( heating ).

I've never seen it happen myself. What's your reason for asking ?

Graham

You may think of the joule rating as similar to a gas bottle. It is rated in how many PSI of gas it can hold. Once you go past this (not allowing for any safety factory and such) , the bottle blows apart. The surge protector will absorbe so much power and then it will start to over heat. If you are lucky it will do like some did to me and short out and blow a fuse before the protected circuit gets overloaded. If hit with a very large surge, it will blow apart and open up and not provide any more protection.

A surge protector absorbs energy from the surge and turns it into heat. It is assumed that this process takes place so fast, that none of the heat has time to escape from the component absorbing the energy, so all of that energy goes into raising the temperature of the component (called an MOV for metal oxide varistor). The joule rating is based on the mass of the MOV, its specific heat (how much the temperature of a gram of that material rises per joule of energy added) and its peak temperature rating. Higher joule rating devices are physically larger.

Joule rating is similar to a life expectancy number. However little relationship exists between MOVs joule rating and how many joules may actually be shunted. For example, an MOV with higher joules rating will 'absorb' less joules over a lifetime that shunts many times more joules; when compared to an MOV that is smaller.

MOVs that vaporize or disintegrate were used beyond what the manufacturer intended. An unacceptable and catastrophic MOV failure is not what joules are measuring. MOVs fail by degrading. A degraded MOV has no physical burning, but has a voltage change of typically 10%. But this does not promote protector sales. Many protectors that are undersized will then promote more such sales.

To determine how long an MOV may last, consult charts provided by manufacturers. For example, a 180 joule MOV may shunt 1 million 20 amps transients, 1,000 300 amp transients, 10 600 amp transients, or 1

4000 amp transient. In each case, the MOV has no observable damage. Any burn marks on MOV is totally unacceptable operation.

As MOV joules increase, joules that it may shunt to earth increases exponentially; increased life expectancy. Those joules shunted to earth are vastly more than the MOV joule rating because the function of an MOV is not to absorb energy. Its function is to shunt as much energy as possible to earth.

As MOV degrades, its varistor voltage changes. MOVs degrade; must not blow out or open circuit. Some mistake that open circuit MOV as normal operation. That open MOV as demonstrated in these scary pictures is an unacceptable condition often found in plug-in protectors that may have been grossly undersized and are a fire threat:

Above third URL dem> I have a question about the joule rating on surge protectors. From some

Then there must be two "joule" ratings: how much energy it can absorb in any one shot, and how much it can gobble over its working lifetime.

Do the datasheets make this clear?

John

Its more like a burned out semiconductor. I read somewhere that the protector can be looked at as a random clump of p/n diodes, and each surge damages a number of them, until they are all short circuit. At that stage it will also short your power and blow the fuses.

MOVs absorb energy like wires absorb energy. Neither is intended to absorb energy. Both are intended to deliver energy elsewhere. An energy rating is provided for MOVs just like a current rating for wires. But MOV parameters are more complex (non-linear). Therefore we don't only rate MOVs in amps. We rate MOVs in joules. Don't confuse MOV joules with energy absorption. In simplest terms, joules are a parameter to estimate protection.

For residential AC electric, the 'whole house' protector should be at least 1000 joules (which is maybe 3000 joules equivalent for a plug-in protector) and 50,000 amps. As MOV joules increase, then protector life expectancy increases exponentially.

If an MOV protector blows the fuse, well, that fuse exists to keep a grossly undersized protector from burning down the house. Effective MOV protectors do not fail shorted. They fail by degrading. Vaporization or shorting out is unacceptable and is not effective protection. But when promoting plug-in protectors on myths, then an undersized protector causes a naive consumer to declare, "My protector sacrificed itself to protect my computer". Wrong. A catastrophically failed protector left a computer's internal protection to protect itself.

A surge too small to overwhelm protection already inside a computer, instead, vaporized, burned, or shorted a grossly undersized and ineffective protector. This promotes sales based on myths.

The effective protector earths surges and remains functional. The effective protector performs without the human knowledge of that surge. But that effective silence does not promote grossly overpriced plug-in protectors. And so we also have this myth that MOVs burn or short out to provide protection.

As defined by manufacturer datasheets, the MOV is properly sized so that it only degrades. Burning or 'shorting out' is totally unacceptable operation and is not effective protection. But it does promote more sales myths.

Meanwhile, a different protector techology is avalanche diodes.

No, this is wrong. The wire is to deliver energy, but the MOV is to divert excessive energy (from a load dump, say, or a lightning strike), popularly called an "energy surge" but its purpose is to turn that energy into heat, such that it (the surge) doesn't damage the equipment.

In the process, it is possible for the MOV to be destroyed, in which respect, they're more like a fuse, but protecting against overvoltage rather than overcurrent.

Hope This Helps! Rich

An

An "effective" MOV wouldn't fail at all. ;-)

Cheers! Rich

A MOV across the AC line doesn't "deliver energy elsewhere", it absorbs and dissipates transient energy. You can't "deliver energy" into ground. (Hint: P = E*I, so if E=0 then P=0)

But my question was whether there are distinct per-event and lifetime joule ratings.

Oh, there's no confusion; they are precisely identical.

Really, I'm surprised I have to explain basic stuff like this.

John

And since no decent computer needs external MOV protection, the most effective MOV is no MOV at all.

John

Both John Larkin and Rich Grise are promoting MOV myths. They provide neither numbers nor citations; just claims that sound more like 'trust me'. Somehow they know MOV absorbs all energy of a direct lightning strike. Their proof?

Well Rich, which is it? Do they divert surges to earth (a shunt mode protector) OR do they absorb the entire energy (a series mode protector)?

First, industry professionals contradict both John and Rich:

From Dr A J Surtees in Power Quality Magazine entitled "Principles of Lightning Protection for Telecommunications Facilities:

Second, consider a predecessor to MOVs - the GDT. If MOVs absorb all energy (as a series mode protector), then so does a Gas Discharge Tube. However GDT changes into plasma; becomes conductive. So where is that massive energy radiating from 'absorbed' (dissipated) energy in a GDT? Like MOVs, the GDT also does not absorb energy. Energy is dissipated in earth. John somehow believes earth to be a perfect conductor - dissipates no energy. Then where is that energy dissipated? In the GDT? Get real. That glow in a GDT is thousands of joules? 1000 joules dissipated in a GDT during the transient is 33 million watts radiated. Where are those million watts of radiated energy if GDTs (like MOVs) dissipate thousand of joules during a 4500 amp transient? Or maybe the energy is dissipated, instead, in earth?

Consider a Franklin lightning rod. If energy is not dissipated in earth, then what dissipates thousands of joules during a direct lightning strike? The lightning rod? Must be because John claims

Or maybe John is wrong and above industry professionals are correct?

In GDT and lightning rods, energy dissipation occurs in earth. Lightning rods and shunt mode protectors (GDTs, MOVs, etc) shunt (conduct, connect, divert, clamp) surge current into earth where energy is dissipated.

Third, review MOV manufacturer numbers. For example, from

is a 2V130 MOV rated at 39 joules. Energy from a 4500 amp surge is dissipated in an MOV rated at only 39 joules? Baloney. 4500 amps through miles of earth does not dissipate energy? Both John and Rich claim all energy in that 4500 amp surge is absorbed in a 39 joule MOV. Assuming a typical earth resistance of 15 ohms times 4500 amps through a 39 joule MOV is maybe 4500 joules dissipated in earth. But somehow earth does not dissipate energy? Somehow a 39 joule MOV dissipated thousands of joules?

Again review other numbers for that 2V130 MOV. 0.6 watts. MOV that absorbs all energy in a 4500 amp surge is only rated at 0.6 watts dissipation? That is what John and Rich tell us. When does a 4500 amp surge contain so little energy as to be dissipated as 0.6 watts? Notice why both John and Rich don't provide numbers, don't quote industry professionals, and only make blanket (unsupported) claims.

39 joule MOV shunts thousands of joules to be dissipated in earth. MOV delivers energy just like a wire. Shunt mode protectors are effective when shunting (not dissipating or absorbing) energy into earth. MOVs degrade when voltage changes by 10%. A vaporized MOV means a grossly undersized (ineffective and dangerous) protector. Vaporized or burned MOVs occur when manufacturer's maximum spec numbers are so grossly exceeded. Consider these pictures when John and Rich insist that MOVs absorb and dissipate all the energy of a surge:

Tell me exactly where I said "An MOV absorbs all energy of a direct ligntning strike".

In other words, be kind enough to refrain from attributing words to me that I didn't say.

Nothing can stop a lightning strike, but if there's one nearby, it can cause a surge that MOVs will sacrifice themselves trying to protect your equipment from.

I once worked at a place that used MOVs all over the place - they made an analytical instrument with a vacuum chamber and an 8' console of electronics, which had three or four high-voltage power supplies.

MOVs wear out. MOVs blow up. As soon as one of the engineers discovered Transzorbs, we ECO'd everything to change to Transzorbs, and tossed the MOVs.

And they don't protect like a fuse. A fuse goes in series, an MOV (or Transzorb) goes in parallel.

Thanks, Rich

Where>

I *asked a question* that nobody has answered, including yourself.

That's just wrong. This guy is confusing current with energy.

Spark gaps, like a GDT, have a negative impedance; once they break down, the arc voltage drops to a small fraction of the striking voltage. That lowered voltage drop reduces the power dissipation, allowing them to conduct a lot of peak current. MOVs have a nonlinear but always positive I:V curve, like a sloppy zener, so dissipate more power than a GDT. It's pretty simple math.

A lightning bolt is a near-perfect (if very fast) constant-current source. After all, the height of a church makes up a tiny share of the length of the bolt. Power is I^2 * R, so if we reduce the R of the church (the grounded lightning rod has a lot less R than wood and hymnals) then much less power (and energy, which is just the time integral of power) is deposited into the sanctuary.

Ditto an MOV on the power line. If we get whacked by a high-impedance current zot, the lower an impedance we present, the less power is dissipated by us. If the ground is "soft", some power will indeed be dissipated in the earth nearby, but the source impedance os the transient is usually high, so by lowering our impedance, more is dissipated in the source path itself, literally reflected back into the source.

An MOV does dissipate power, E*I, just like anything else. That's why they make big ones and little ones.

Well, I'm an engineer, and they're selling MOVs. Who you gonna trust?

Baloney, I said nothing of the sort. I do claim that a MOV can't evade tha laws of physics.

15 ohms is a pretty ratty earth ground, but OK. The real issue is what is the source impedance of the transient? I suggest that a lightning bolt or a line-conducted transient has a rather high impedance (the math isn't hard here, either) so is essentially a current source. And we don't have to dissipate the 4500 joules, we merely have to get out of its way.

Well, we don't commonly use wires at 300 volts drop.

And nobody's answered my question.

John

John, you are wasting your time with this ignorant troll. All you have to do is mention MOV or lighting protection in a long list of newsgroups and he shows up with his cut and paste bullshit. He has his own agenda that doesn't agree with industry experts, so he claims they are all wrong while he has absolutely nothing of any substance to back up his position. He doesn't understand anything, and refuses to listen.

Hundreds of people have argued with him, yet he still claims that everyone else is wrong. He cites small snips of a document out of context, then complains when people point out what it really says. There are times when he makes Rich Grise look like Albert Einstein.

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

(snip)

I think that joule rating is for the largest pulse that doesn't degrade the part outside its tolerance. The capability to absorb energy goes up dramatically, if the absorbed energy packets are much smaller and the device has time to regain ambient temperature before the next one (in other words, is related to the peak temperature the material reaches).

Here is a data sheet from Epcos:

They provide a joule rating for a 2 millisecond duration (essentially no energy loss from the surface of the device during the pulse, so all the energy goes into temperature rise). The total energy (throughout its life) the device can absorb goes up dramatically, as the peak goes down, if the device returns to ambient before the next pulse. The wear out mechanism is evidently, closely tied to peak temperature.

Lots to read on this subject at:

I suppose so. He can't answer my question.

Funny how people get fixated on things like this. I'm sure he's confident his MOV science works, just like people are confident that wearing a garlic necklace has kept the vampires away.

Personally, I don't design MOVs into my equipment or use them externally... just makes them more expensive and less reliable. I figure that if it passes UL and CE, I can sell it.

Now, be nice.

John

The best guide I have seen on surge protection is at

- it was published by the IEEE in 2005

A second good guide is

- it is published by the National Institute of Standards and Technology, in 2001

MOVs, as some others have stated, is a device that clamps the voltage across it, much like a bidirectional Zenier diode. The advantage of a MOV is that the voltage is dropped across grain boundaries throughout the bulk of the device, not at a single junction as in a semiconductor. That gives a MOV an advantage of larger power rating for a given volume.

With a large surge current through them, and a clamp voltage (maybe 300V for 120V devices), MOVs obviously dissipate power. A surge at well under a millisecond makes their use practical since energy deposition is relatively low. For an "overvoltage", significantly longer, a MOV will rapidly fail, and one authority has said this is the most frequent cause of failure of surge protectors.

I agree with John? that the power/Joule rating for a single device is that which will produce a 10% change (lower) in the voltage for a current of 1 mA. If they take a large energy hit, grain boundries fail resulting in a lower clamp voltage with eventual total failure. I agree with w_tom that you size the protector energy rating much higher that the expected single hit expected to give a practical long life.

Transzorbs, avalance diodes, gas discharge tubes all are voltage clamps that absorb energy and will fail if absorbing too much. Spark gaps also absorb energy, and can have a problem with continuing to arc at the restoration of normal voltage (power follow). Spark gaps and gas discharge tubes take time to operate, which can be a problem since surges have maybe a 10 microsecond rise time.

A MOV surge suppressor on an electric service entrance will clamp the voltage, typically hot to neutral, where neutral in the US is connected to earth. Some energy is absorbed in the MOV, some in the earth resistance. If a 10,000A lightning induced surge (reasonable) flows through a 15 ohm neutral-to-earth resistance (reasonable) the result may be a voltage drop of 150,000V from neutral to 'absolute' earth potential. The H-H, H-N voltages will be clamped by the surge protector at a sustainable voltage for equipment, riding on this drop to 'absolute' earth.

With a surge like this, phone and CATV wires can wind up thousands of volts different from power, with disasterous effect on equipment with power and phone/CATV connection. Protection is effected by connecting the phone NID and CATV ground block via short wire to the power-earth common point, which is the conductor to the grounding electrode at the service panel. This is a single point ground, and is very important whether or not the electric service has a surge protector.

In a plug-in surge suppressor, all wires are clamped to the common ground at the surge suppressor. If there are phone/CATV/LAN connections they also should go through the surge suppressor. All wires are clamped to a voltage level that the equipment can sustain. This is clearly described in the IEEE guide as a Surge Reference Equalizer (SRE).

The IEEE guide advises against comparing surge suppressors based on energy/Joule rating because there is no standard for how it is measured.

bud--