Data cable surge protection

Status
Not open for further replies.
T

T.M.Haja Sahib

Guest
The data cable entering from outside any building, even if provided with optical isolation at the computer end, still requires surge protection. I do not know the reason behind it. I request clarification. Thanks.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
120822-0647 EDT

T.M.

You need isolation at both ends.

If you want protection from a direct lightning strike to the communication cable, then use a fiber optic communication cable.

If several thousand volt optical couplers are used, then no clamping devices are really needed. But it may be useful to tie the common reference of the communication cable to ground at a single point (earth or an EGC, probably does not make much difference where). Two or more point grounding would probably cause some unwanted current in the cable.

.
 
T

T.M.Haja Sahib

Guest
gar
Thanks for your reply which stimulated me to get more clarification by myself.
An optical isolator will prevent ground loop current and voltages from causing damages to the system but not from being damaged itself by transient surge voltages beyond its rating.So a surge arrester of suitable rating for the data cable at entry point to the building may be necessary to divert most of the surge energy away from the optical isolator so that it is not damaged.
I found the above information in the following web link subsequently.
http://www.patton.com/technotes/about_optical_isolation.pdf
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
120822-1332 EDT

T.M.

If you do not use an isolator at both ends, then the low voltage semiconductor components at the non-isolated end can be damaged by a high voltage, may be in the 20 to 30 V range, being applied to a wire in the communication cable.

You have to be concerned with more problems than just ground currents.

Isolators should be very close to the electronic circuits (computer or whatever), like within 2 ft.

If the largest short time voltage transients that may be imparted to the communication cable are under +/-1500 V peak relative to ground, then ordinary optical couplers should be satisfactory.

.
 
T

T.M.Haja Sahib

Guest
What good would be the rest of the device if your opto-isolator gate is destroyed?!
The optical isolator has a transient surge voltage withstand rating. In some cases it would be sufficient. In such cases provision of surge arrester to the data cable may be dispensed with.
 

hurk27

Senior Member
Remember since a lightning event is a high frequency event it will also induce current/voltage into any loop circuit/systems, it is important to use some kind of isolation in any loops such as networks, audio (entertainment systems) and even phone and cable, a lightning attachment point even at a distance of a few hundred feet from the loop can produce enough voltage in a loop to cause damage, this LEMP (lightning electro magnetic pulse) has be in the 10's to hundreds of Mhz and can circulate current in very small loops of conductors.

Also since this is kind of a transformer effect this voltage is isolating and does not seek Earth as a return path rendering most Earth grounding and surge suppressor systems useless, this latter part is because of the way the loop is formed, take a computer, it might have two or more connections to it that when looked at as a whole will form a loop, if this loop is at the right impedance (think resonate frequency) a voltage will develop within this loop, and if the computer is part of that loop (as most will be) you will receive damage to the component of the computer that was part of the loop, take the network side, you have a network that connects to the router/hub that is connected to the power supply and EGC, the computer also has a connection to this router/hub and power supply EGC, the building wiring will complete the loop between the two devices (which can be several) now here is why most SPD's don't protect for this kind of event:

The lightning pulse (LEMP) induces the current on all the conductors in the loop as if they are one conductor, take a single Cat5 cable, all eight conductors are at one potential, same as the power cord supplying power to the computer, (in the US) the hot, EGC, and neutral are all one conductor in this type of event, kind of like having a circuit where the power cord is one conductor and the network cable is the other conductor, and guess what is inbetween, yes the network card.

since the SPD's only protect between each conductor in a single cable it will not protect between the power cord and the cat-5 where the damaging voltage will be.

Understanding this event takes a little out of the box thinking, and understanding how the wiring in a building can form loops along with other cables that will put many of our electronic equipment in these loops if we don't isolate to break the loop at a high enough voltage rating to prevent this type of damage, another way to prevent this type of damage takes no protection devices at all, just by wiring in what I call a spider fashion where you run your wiring in a method that does not create loops can go a long way to stop LEMP, one method I use is to run all conductors in a single path down the middle of a house then branch out to each area, keeping conductors close enough that the LEMP will treat them as just one conductor with no return path, but this would be impossible or not cost effective to do in existing installations, so isolation is the next best thing we have that can be easily done, fiber for network, wireless where possible, and other methods to achieve a high voltage break in any loops, it only has to be done in one point of each loop to be effective.

Do search on LEMP or lightning electro magnetic pulse, and it might open up a few eyes as to why SPD's are not protecting our equipment when a lightning strike hits a tree out side of our house.
 
Last edited:
T

T.M.Haja Sahib

Guest
Wayne:

Thanks for your reply.

However it appeared to have missed one point regarding applicability of SPD's.

Here it is.

If a transient over voltage is induced in a system of conductors, voltage will be present across any two conductors. Each conductor is usually provided with a SPD. These SPD's are interconnected through ground connection and the SPD's become shorted together when transient over voltages are present in the power lines connecting them so that those over voltages can not pass downstream to cause damages.

So the SPD's can perform protective function even in the case you stated.
 

hurk27

Senior Member
Wayne:

Thanks for your reply.

However it appeared to have missed one point regarding applicability of SPD's.

Here it is.

If a transient over voltage is induced in a system of conductors, voltage will be present across any two conductors. Each conductor is usually provided with a SPD. These SPD's are interconnected through ground connection and the SPD's become shorted together when transient over voltages are present in the power lines connecting them so that those over voltages can not pass downstream to cause damages.
So the SPD's can perform protective function even in the case you stated.

The above in red would work if we were talking about a voltage/current between the conductors in each cable but the voltage that is being induced is upon all the conductors in each cable as if they are all one big parallel conductor, in a high frequency event the impedance of wire goes through the roof as the frequency increases, so to a high frequency event these cables (network/power) become as if they are a winding in a transformer and the output is connected between the network cable (all 8 wires in a cat-5 cable) and the power cable (hot,neutral and ground) now you can see that the voltage developed between these two cables has the network card in a computer between them which is why it takes the most damage in most near field lightning events.

Here is a quick drawing of what I am talking about:

Untitled.jpg


blue lines are the network path, red lines are the power path, if we start from the panel to the first computer then to the router, then to the second computer then back to the panel we create a complete circle, the current induced into this circle will only follow these paths and not have a reference to any other paths, just like in a transformer secondaries, we leave the transformer to the load through the load then back to the transformer ( a circle) but at 50 or 60 hz the circle between the computers would be a short, but to higher frequencies it will be an open just like a transformer secondaries are a short to DC but acts like an open to AC, so this is why the SPD's will have no effect because they are not in the circuit but just part of the circuit, kind of like if you had the SPD connected between the strands on one cable and even though you had thousands of volts passing through that cable the SPD couldn't do anything since it was at the same potential on both leads.

for the SPD to protect in graphic above, one lead of the SPD would have to be connected to the network cable (all 8 wires) and the other lead would have to be connected to the power cord (all three wires) on each side of the computer it is protecting. Kind of impossible in my book, this is why isolation is the only answer if the routing of cables can not be done to prevent loops, there is also a smaller loop in my drawing between computer one and the router because of both of them being powered from the same circuit, this can also have current induced into it by lighting unless you keep these two paths close to each other so they will act as one conductor in the magnetic field.

Many do not relate the high frequency nature of a lightning strike, and how it can induce currents into loops of conductors in the near field to the event, many make the same mistake you did with the SPD's and try to apply low frequency (50 or 60hz-DC) theory into the circuit when trying to understand how it works and this is why many come up empty handed at the reason our equipment is still being damaged even though we have all kinds of protection on the equipment, we must re think the circuits applying a much higher frequency theory to understand how lighting works.

Edited to add: Well the drawing didn't come out to good, as it is too small, just draw a service panel two computers and a network router in a circle, supply computer 1 and the router from the service panel and computer 2 from the service panel, then draw the cat-5 5 from each computer to the network router and you will see the circles of a complete loop that can have a current induced by a LEMP event.
 
Last edited:
T

T.M.Haja Sahib

Guest
Wayne:
Let me ask you one question for more clarification on the subject.
If lightning strikes ground, ground potential rises. If any building service equipment is nearby, the lightning current can pass through the EGC via its connection with the GEC, seeking 'better' ground. Under this condition, do the SPD's connected across all the power lines of the building offer protection or not?
 
T

T.M.Haja Sahib

Guest
Wayne:

The answer to my question was by Mr.Ron hotchkiss as below.

Yes. However, it is important to understand that SPDs are not just intended to flow surge current. The more important function of the SPD is to limit (or equalize) the difference of potential (voltage) across the load or mode that is intended to be protected. Without a difference in potential, current cannot flow. So, in this case, regardless of the direction of current flow (ground to phase instead of phase to ground), the SPD is intended to limit the voltage across the load. Further, in the scenario of ground (earth) potential rise, it is important to have SPDs installed downstream of the service equipment at other susceptible loads to limit the difference of potential at those locations as well.
IEEE Std 1100-2005 has an extensive discussion of GPR.


So my point is that between any two incoming lines, whether power or communication, SPD's would be connected back to back across the lines via the common ground connection. If there arises any excessive over voltages between the lines, The SPD's would conduct, minimizing the over voltage to safe level across the equipment they were designed to protect.
 
T

T.M.Haja Sahib

Guest
Laszlo:
I do not think high protection class of external surge arrester is of relevance as applied to data cables which are of low voltage class and their surge withstand voltage may be from 3 to 4 kV. So the protective surge arrester for the data cable should have a transient voltage rating below that level, whereas some optical isolators have surge withstand voltage of 4 to 5kV.
 
Status
Not open for further replies.
Top