How fast is an MOV? Really?

This sounds like something John Larkin would do...

Do MOVs respond instantaneously (i.e. as soon as the electric field appears across the faces of the metal-oxide stuff) or do they have some fundamental response time which is slower than the RLC of the geometry? Could you, for example, build a waveguide (matched to the MOV's dielectric, obviously) which is able to mostly block radar pulses for the radar reciever? (I know, they use different approaches to solve this problem, but just for an example anyway...)

For that matter, avalanche diodes can, well, avalanche under peaky conditions (making some excellent steps, at that). Do MOVs have any negative resistance characteristic, or are they only plus-resistive?

Tim

Reply to
Tim Williams
Loading thread data ...

Who, me?

MOVs tend to have a lot of capacitance, and maybe lead inductance, so teasing out the actual conductive response might be a chore.

formatting link

"The response time of the MOV is largely ambiguous, as no standard has been officially defined. The sub-nanosecond MOV response claim is based on the material's intrinsic response time, but will be slowed down by other factors such as the inductance of component leads and the mounting method. That response time is also qualified as insignificant when compared to a transient having an 8 microsecond rise-time, thereby allowing ample time for the device to slowly turn-on. When subjected to a very fast,

Reply to
John Larkin

Use a "Transzorb" instead.

Reply to
Archimedes' Lever

Archimedes' Lever > =A0Use a "Transzorb" instead.

What's the response time, cost and lifespan?

Googling transzorb was somewhat unsatisfactory.

Reply to
Greegor

We usually do. A good DC power-input front end is a polyfuse, a unidirectional transzorb, and a couple of ceramic caps. That takes care of transients, RFI, downstream shorts, reasonable over-voltages, and polarity reversals.

AC line inputs can be designed to withstand any likely line insults without either. That's better, since either a MOV or a transzorb will explode if asked to absorb a really nasty event.

John

Reply to
John Larkin

I usually add a series inductor/ferrite bead. Also sometimes substitute a bidirectional TVS so as to prevent high currents if polarity reversed, with a series diode.

-[polyfuse]-------[FB]-----|>|-------- | | TVS --- | --- | | --------------------------------------

The TVS limits an ESD event to a few tens of volts. The FB/cap flattens out the remainder.

With a "mains filter" configuration of chokes and capacitors?

John Devereux

Reply to
John Devereux

We used to put filter/trap beads right on the legs of the devices on some of our designs

Reply to
Archimedes' Lever

Measuring response time is a problem. A technical paper at:

may be of interest.

A major theme is that the current through the protective device will induce a current in adjacent circuit loops. One loop is even a small loop from unshielded scope leads to the device. Rather extreme measures to eliminate the loop were used, and the resultant scope trace shows no reaction time visible for a MOV with a trace of 2 us/div.

The authors are interested largely in power circuits and extremely fast rise times are reduced by source wiring impedance. Pulses "have to travel only a few meters away from their origin to have their rise time and duration stretched into tens of nanoseconds or more". (But that may not help with nuclear EMP.)

The same induction into a loop adjacent to the protective device would transfer energy into a downstream circuit unless departing leads were twisted.

The paper also looks at lead length for power panel suppressors - lead length (primarily inductance) increases the effective clamp voltage.

--
bud--
Reply to
bud--

formatting link

Here is one of the MOV versus TVS shootouts:

formatting link

One must always keep in mind that MOV are like bank accounts. They provide a certain amount of cumulative protection per Dollar. Every surge where they have to come on draws from that account. Until one fine day ... *PHHOOOMP*

The latter event is occasionally accompanied by blaring sirens, flashing lights, fire hoses and lots of water.

--
Regards, Joerg

http://www.analogconsultants.com/
 Click to see the full signature
Reply to
Joerg

Thanks for posting this link, Joerg. It's the first writeup which appears to confirm something that I've been told in the past: that one of the ways in which an MOV can degrade over its lifetime, results in a decrease in its threshold voltage, eventually resulting in a shorted MOV after it takes a number of hits.

Yup. As I understand it, one should always place an MOV *after* a properly-sized series fuse or circuit breaker, to reduce the chance that it'll fail near-shorted and burn up.

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
 Click to see the full signature
Reply to
Dave Platt

Exactly. It is meant to 'fail' or 'short' enough to open the fuse.

It isn't really meant to shunt an overvoltage. It is meant to cause a breaker to open because of its reaction to an overvoltage. In that sense, power sources feeding equipment failures from open fuses should be investigated where MOVs are used in the front ends of such devices. They are like alarms for your power condition.

Reply to
Archimedes' Lever
[...]

Well, you'd think that a 15A breaker ought to do that job for a surge-protected power strip but:

formatting link

--
Regards, Joerg

http://www.analogconsultants.com/
 Click to see the full signature
Reply to
Joerg

Yup, I've had a similar incident. Came home and noticed that unmistakable odor of crisped electronics. Fortunately, the MOV in question was in a metal-cased power strip and the breaker tripped before anything nastier happened.

Must have scared the cats out of a life or two, though ...

--
Rich Webb     Norfolk, VA
Reply to
Rich Webb

formatting link

MOVs do not have cumulative damage from absorbing energy, at least nowhere near linearly cumulative. An MOV rated xx joules can take xx joules in an event short enough to give it little chance to cool before it is done absorbing energy. That same MOV can take a lot more than xx joules in smaller events spaced far enough apart in time so as to not experience much temperature rise. It appears to me that MOVs fail from overheating, and that small events do not cause significant damage.

Of course, MOVs should be deployed in a manner that allows them to blow up without starting a fire. A lightning strike or a transformer failure can deliver a lot of energy at voltage at which the MOV conducts.

I have also seen signs of massive line voltage surges a couple times with lots of energy from unknown cause at one location a few blocks from Philadelphia's very major 30th Street rail station (not known to be related to thye blowout events) - most of the fluorescent lamps (F40T12 with 2-lamp rapid start ballasts) were blown by these events. Apparently the voltage was high enough long enough for arcs to form across the filaments of the lamps (takes about 12 volts to do that, normally present with F40 rapid start ballast is 8.5 volts). These blowout events had excessive voltage for enough time and with enough current to blow at least one filament in a couple dozen F40 lamps.

An MOV should be deployed in a suitable metal enclosure, with a fuse or circuit breaker upstream of the MOV, maybe also a thermal fuse activated by the MOV overheating.

If the MOV is fused independently of what it protects, then it can be a good idea to put an indicator lamp in parallel with the MOV to indicate that it is connected.

- Don Klipstein ( snipped-for-privacy@misty.com)

Reply to
Don Klipstein

No, it is meant to shunt an overvoltage. MOVs must conduct current in microseconds. Circuit breakers and fuses take milliseconds to trip

- one thousand times too long.

A 15 amp circuit breaker must never trip due to MOV failure. MOVs must never fail by burning or shorting as some have described. MOVs threshold voltage must change no more than 5% and have no visual appearance changes. MOVs must never spit sparks and smoke.

What happens when a grossly undersized (ineffective) protector fails catastrophically (in violation of MOV manufacturer Absolute Maximum Parameters)? It can only blow a tiny thermal fuse. That fuse disconnects the MOV and leaves appliances connected to the surge. However even that tiny fuse is sometimes insufficient to avoid a fire threat seen often by fire companies and as described by a NC fire marshal in:

formatting link
entitled "Surge Suppressor Fires".

An effective surge protector must even shunt a direct lightning strike (to earth) and remain functional. But no failure means the naive may not recommend it. So many are built to maximize profits. Use too few joules in and install that fuse to blow faster. That gets many to promote that protector when it really provided no protection (the same protection also listed in its numeric specs).

An effective protector shunts current to earth, remains functional, and people never even knew the surge existed. A protector is to shunt surge current so short to earth that voltage is minimal. Lower voltage (due to more joules in a protector) means less energy is dissipated inside the protector or house AND more energy is safety dissipated in earth.

A protector's purpose is not to stop or absorb surges. It is a shunt mode device. It connects the maximum transient current to earth with minimal energy dissipated in the protector. And so fast that no circuit breaker opens. If the MOV burns or vaporizes, then it has completely violates MOV manufacturer specs AND has created a dangerous and unnecessary human threat. Worse, many do so adjacent to a desktop full of papers or on a rug behind furniture.

All reasons why effective MOVs are located in safer locations that are also closer to earth (low impedance connection) such as attached to a breaker box.

Reply to
westom

I took a look now.

This does confirm my previous response that damage is not linearly accumulated by absorption of energy, but does still state that degradation occurs. It states that a 20 mm MOV can be expected to fail from 1,000

40-amp "10/1000 us pulses" (risetime 10 us from 10%-peak, falltime 1 millisecond from peak to 50%), amounting to cumulative conduction of mostly somewhere between 20 and 40 amps for a second.

- Don Klipstein ( snipped-for-privacy@misty.com)

Reply to
Don Klipstein

... and the problem is that the user may not know that such pulses occur and certainly not when the 1000th pulse has be stomached.

--
Regards, Joerg

http://www.analogconsultants.com/
 Click to see the full signature
Reply to
Joerg

Joerg,

You can tell the quality of an engineer by his sound effects.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Every "ka-blooie" is a learning experience ;-)

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Most of you sound like a female fart.

Reply to
Archimedes' Lever

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.