Surge Protection information - true or false?

Status
Not open for further replies.
Pierre C Belarge

Pierre C Belarge

Senior Member
Location
Westchester County, New York
Because there is so much false information circulating out there, it is sometimes difficult to determine what exactly is fact.

Here is a link that seems pretty interesting. How much is fact or fiction? I am hoping Dereck will see this and chime in after reading it. It does make sense to me, but I am not an engineer...

clickhere
 
Last edited:
Pierre I will be brief tonight, have other obligations. Short answer is there is some truth and some hype to what they are claiming.

There are two types of Surge Arrestors:

Parallel Mode (aka shunt mode). These are your Gas Tubes, MOV?s and SAD?s. They go across two points in any mode, and only operate when the clamping voltage is reached. There big advantage is brute force able to dissipate very large amounts of energy.

Series Mode: These are simple Low Pass Filters constructed of resistors, inductors and capacitors. Or simply stated RLC circuits. A transient is a fast or high frequency event which the RLC circuit attenuates. The big disadvantage of these types is they have to be able to handle the rated circuit capacity. For example: If used at the Service Entrance they would have to be rated and listed for such purposes.

I will try to come back later and expand on the advantages and disadvantages of both when time permits.
 
Dereck
Thank you for the answer. As you can tell from some of my other posts, I am trying to get as much information and education on this topic as I can. I am curious as to what other info you may provide. I completely understand the time factor and other issues that consume you and other individuals here...I have a ton of patience...I should have been a doctor ;)

If you find the time, if there is other areas of training or reading that you could provide, I would appreciate it.

BTW: if anyone else has good info or links, please chime in and let us know.
 
Pierre: Mike Holt has said and I agree- Concerning TVSS
You get what you pay for
There are a lot of snake oil salesmen selling TVSS with no value
More is better.

I would add the surge does not go to ground.
 
Pierre C Belarge said:
Once the transient is clamped, does it not travel to ground via the neutral connection to the GEC at the service?
Click to expand...
If the source of the transient is the ground, then yes. . .

Thinking about the customer side of a PoCo transformer, excluding a direct lightning strike, the more likely "supply" of the transient will be somewhere other than ground. The energy clipped by the TVSS will complete a circuit through the ground (EGC) to the main bonding jumper, out to the transformer and back in again on the hot conductor(s).
 
al hildenbrand said:
If the source of the transient is the ground, then yes. . .

Thinking about the customer side of a PoCo transformer, excluding a direct lightning strike, the more likely "supply" of the transient will be somewhere other than ground. The energy clipped by the TVSS will complete a circuit through the ground (EGC) to the main bonding jumper, out to the transformer and back in again on the hot conductor(s).
Click to expand...


Al
That sounds like a ground fault?



My understanding of how a transient is captured, is like this:

The transient occurs- very fast event,lets say 3000 volts and 500 amps (I made that up). The TVSS equipment clamps the transient, sends it either via the neutral or the EG conductor (depending on the "mode" connection) and the transient event travels back to the service enclosure (or where the neutral to GEC connection is made) and then is sent to earth to dissipate.

My understanding is that it does not travel the path you have posted.
 
Pierre C Belarge said:
The transient occurs- very fast event,lets say 3000 volts and 500 amps (I made that up).
Click to expand...
In this hypothetical, what is the source of the energy in the transient?

Is it an electromagnetic pulse inducing the energy onto conductors of certain orientations, . . .is it a direct contact of say a PoCo primary conductor to the secondary, . . .is it a discharge of a cloud to Earth accumulation of opposite charges?

Or maybe the transient is the result of the magnetic field collapse of an inductive load as contacts chatter?

My point is that, although very fast, the transient, in any instant, is acting in a circuit, and for the transient energy to flow, it must complete a circuit.
 
al hildenbrand said:
In this hypothetical, what is the source of the energy in the transient?

Is it an electromagnetic pulse inducing the energy onto conductors of certain orientations, . . .is it a direct contact of say a PoCo primary conductor to the secondary, . . .is it a discharge of a cloud to Earth accumulation of opposite charges?

Or maybe the transient is the result of the magnetic field collapse of an inductive load as contacts chatter?

My point is that, although very fast, the transient, in any instant, is acting in a circuit, and for the transient energy to flow, it must complete a circuit.
Click to expand...


Al
I know that for current to flow, there has to be a complete circuit. For the instance of what I am trying to understand, I am not discussing lightning events, or large utility instigated events...I am thinking more on the line of "in building" induced "Transient" (such as - AC, motors, laser printers, PLCs, elevators,etc...). The complete current path for these transients as I understand it, is to travel across the TVSS, get shunted, and then travel to the service and at the "single point ground" connection, get passed down to earth via the GEC. This would be for a parallel type of TVSS.


I am not saying I am positive about this path, this is what I am trying to understand if it is a fact. Or, if there is a different path.
 
Pierre,

The document at the link you posted above defines their TVSS as a series mode device. They define series mode as
Series Mode; a brand of surge protector that uses high-voltage filtering techniques to block and contain surge energy and filter transients without dumping energy on the neutral or ground wires.
Click to expand...
Thinking of your hypothetical 3000 volt 500 amp transient, I have a lot of trouble imagining the "container" that this manufacturer's TVSS has, inside its enclosure, that will "contain" the transient.

I could believe that the TVSS appears as a very high impedance to "fast" changes, and a very low impedance to "slow" changes, in voltage.

One thing I learned while working for a small PoCo about transients is that they tend to reflect off impedance changes. A lightning surge travelling down a overhead conductor, when coming to the end of the overhead where the overhead goes down a pole and continues in direct buried cable, . . .the lightning surge will double as the surge reflects back on itself. Insulator damage will occur at this end-of-overhead-pole that won't occur elsewhere along the line.

Seems to me that the series TVSS will have some of the same issues.
 
Pierre C Belarge said:
I am thinking more on the line of "in building" induced "Transient" (such as - AC, motors, laser printers, PLCs, elevators,etc...). The complete current path for these transients as I understand it, is to travel across the TVSS, get shunted, and then travel to the service and at the "single point ground" connection, get passed down to earth via the GEC.
Click to expand...
The "in building" transient is the interesting one, to me.

It helps me to think of the high speed transient not as a very fast thing, but, rather to expand time so that I am thinking of a very short little slice of the transient. So short of a slice as to be able to think of the voltage and current flow as DC.

In this instant of time, if the current path is as you describe, then the source of the transient becomes a monopolar phenomena. The current never returns to the source.

To be clear, there is a path from the GEC, through the earth, and back into the building conductors at the supply transformer's connection to earth, but the earth impedance will be an order of magnitude, or better, greater than the impedance in the service grounded conductor. The transient current will return to the "low side" of the point or area of "induced in building transient" to complete the circuit.
 
Have had several machine tool failures due to

Lightning, which in my experience have caused very fast impulses, blasting open MOVs,
Forklifts touching serial cables, which cause RC networks to overheat,
Handling damage to memory &/or MOSFETS, which cause microscopic damage, frequently with an intermittent failure.

The people who make low voltage power supplies for PCs are a strange lot as far as this goes. Some of them can add new protection to their product in a matter of days, depending on failures. Remarkable ability to adapt.

In any case, found a very organized and informative basic discussion in a Cahier Technique. It was interesting to review the IEC standards for lightning protection, which are summarized. They appear to be harmonized with UL 1449.

edit in the link I mentioned:
http://www.electrical-installation.schneider-electric.com/ei-guide/pdf_files/EIG-J-protection-voltage-surges.pdf
 
Surge protection using series inductors works as it is a low pass filter. 60Hz passes through unattenuated, but a waveform with a very fast risetime, which is the hallmark of a transient, will be seriously attenuated. The absorbed energy comes out as inductor heat, which is why it is a good job transients are very shorted lived, as there are large amounts of energy involved.

The best surge arrestors use combinations of over-voltage devices and inductive low pass filters. MOVs and gas filled tubes do make a mess of your ground potential if the return path isn't of sufficiently low impedance. Sharp bends in cable path to ground is enough to make a surge protector a waste of money for common mode transients. The ground may measure fine at DC and 60Hz, but with an equivalent frequency of megahertz it's anything but a low impedance path.
 
Pierre I have a little more time for this now.

There are only two types of transient events that have already been discussed; Normal mode (aka differential modes), and common mode. There are no others.

According to IEEE and the NLSI 95% of all transients experienced in home and businesses is ?Normal Mode? because they happen on the primary side of the utility transformer. So for a service entrance device all you need is protectors from L-L and L-N. Even if there were some way (like lightning striking your overhead service conductor) for a common mode event to happen, it is a waste of material to have common mode protection at a service entrance from L-G and N-G because of what requirement in the NEC? 250.24 maybe?

So here is what manufactures wrestle with when designing series rated equipment, it?s the same dilemma I wrestled with when collaborating designs with ACT (now known as GE). A ?series TVSS? used at the service entrance would have to be designed and tested as service entrance rated equipment fully capable of withstanding full load and fault currents. I hope you see the problem with that approach. Can you imagine all the configurations a manufacture would have to come up with, and how much money would be involved in listing each and every configuration? Well the simple answer is it is impractical.

The next best approach is what a lot of manufactures now offer is standard ?parallel protection built into the switch gear or main panels. The one I really like for 200-amp or less services is the meter collar type.

So when a parallel unit operates at the service entrance 1 of 3 things is noticed beyond that point:

1. Nothing

2. Brown-out because the event was large enough for the SAD?s, MOV?s, Tubes or whatever clamped long enough with a low enough impedance it briefly short out the service lowering the voltage temporarily before returning to normal operation.

3. Black out because the TVSS dropped the transformer service fuse or some other OCPD upstream.

Now for common mode, parallel units work the exact same way by clamping or lowering the internal impedance. So is it fair to say it dumps the event to ground? The answer is both yes and no. Yes from the perspective a short appears between L-G or N-G, so there is fault current flowing, but what is the impedance of the ground circuit at the frequency of the event like lightning? Well I am not going to cover the math again, but needless to say the ground circuit impedance at lightning frequencies is enormous.

OK, one example: The impedance of a single 10-foot long 750 MCM GEC is more than 3K-ohms. So what is the impedance of an EGC in series with the GEC? I don?t know or care, it is useless in this event.

Now that doesn?t mean we can?t have something useful at the point of use, we just have to design something that work using parallel devices and or series devices. So what is it? What I call a Surge Reference Equalizer, you know what they are, the power strips you use at your computer to plug in all the components for power including I/O ports for all the gizmos.

Inside the black box (Surge Reference Equalizer) is a ground buss and L-N buss. MOV?s are installed in all modes (L-N, L-G, and N-G) on the AC, plus protectors on each I/O port from L-L, and the magic L-G mode. So if an event occurs we now have a faux single point ground formed by the protectors, thereby limiting voltages to acceptable limits in all modes across all devices.

Now we can take that one step further by installing an isolation transformer inside the black box eliminating common mode occurrence on the output side of the transformer. Ever seen one? Lots of them on the market for home entertainment centers. The ultimate device is a true dual conversion UPS.

OK, enough for now.
 
Status
Not open for further replies.
Top