adding surge protectors in LB connector

[mikefamig]

Can you guys tell me what happens when a surge protector fails? Do they short your lines to earth or do they go open? Do they usually fail all at once?
 
I have read that they will reset automatically after a surge several times but eventually will fail. Is there any way to test a used unit?
 
Mike.

 

[RAL]

Most surge protectors use MOVs as the surge protection device (SPD).   Under normal conditions, the MOV looks like an open circuit and doesn't conduct any current.  But when a surge occurs, the MOV conducts, diverting the surge current to ground (or the other side of the line, e.g. hot to neutral).
 
I like to think of the MOV as a rope that frays with each surge.  At some point, when it has carried too much surge current, the rope breaks and the MOV is no longer capable of doing its job.  Manufacturers usually rate MOVs as being capable of a certain number of surges of so many Amps, such as 10 surges of 1000 Amps.  But if a single surge is greater than 1000 Amps, that uses up more of the MOV's capacity. 
 
What is supposed to happen with an MOV is that it conducts until it can handle no more, and then it turns into an open circuit. A large surge often causes the MOV to explode.  But this doesn't always happen, especially when the over voltage condition is small.  Then, the MOV may just remain as a short circuit, heating up and smoldering.
 
To protect against that case, a thermal cut out (TCO) can be added by the manufacturer of the surge protector.  There are also MOVs that have an internal TCO, and then it is called a TMOV.  If the temperature rises too much, then the TCO blows and the short circuit condition is removed.
 
So the bottom line is that the MOV may show obvious signs of damage due to a surge, but not always.  Some surge protection devices include "failure" LEDs, but depending on the design, these aren't always a reliable indicator either.  That makes knowing whether the SPD is still working or not a very frustrating exercise.

 

[Del91574]

Diteks work slightly different per se.
 
The quick answer is they flash over to ground and the unit no longer works properly.(hopefully)

We have a site that has a FA SLC circuit out to a guard shack (poor design choice) some distance away from their main building. They lived with a ground fault on their fire alarm panel for about 5 years because the other vendor couldn't (or didn't) troubleshoot the SLC and meter the unit to ground. The SLC still worked perfectly fine.

 

[pete_c]

Here my main surge protector has a little computer panel built in to it.  Its the basic model.  I did add another surge supressor outside and adjacent to the HVAC compressor.  It has little status LEDs on it that I will never look at.
 
I had one lightning surge in the 1990's which took out serial ports and a network hub (not the security panel at the time).  There I had a detached garage and had run gas lines, LV, HV to the garage.  I did have a sub panel there and never did have any issues. 
 
Another one in the 2000's which took out two irrigation controllers.  (I learned over time and took my chances on what I had done / assumed I did correctly).
 
I am guessing here Mike you have done the best you can do based on what you have learned right here on the Cocoontect forum.

 

[mikefamig]

I am guessing here Mike you have done the best you can do based on what you have learned right here on the Cocoontect forum.

 
I like to say "aim low, hit nothing" and I don't like leaving anything to chance if I can help it. It seems that when I do it always comes back to bite me.
 
Mike.

 

[pete_c]

I like to say "aim low, hit nothing" and I don't like leaving anything to chance

 
Yup; many folks today getting in to automation forget this stuff or make the assumption that with no infrastructure software or hardware based automation is magical.
 
Of course if you do not see what can ding you; then the assumption is that it will not ding you and folks prefer that these days.

 

[westom]

Most surge protectors use MOVs as the surge protection device (SPD).   Under normal conditions, the MOV looks like an open circuit and doesn't conduct any current.  But when a surge occurs, the MOV conducts, diverting the surge current to ground (or the other side of the line, e.g. hot to neutral).

Expanding on this.  Surges that do damage can be tens of thousands of amps.  Effective protector must not fail on any such surge.  Protection is always about where hundreds of thousands of joules harmlessly dissipate.  Read specification on many popular (and near zero) plug-in protectors.  What happens to its hundreds or thousand joules during a potentailly destructive surge?  Either the protector parts must disconnect as fast as possible (leavning that surge still connected to appliances).  Or that protector fails catastrophically (a potential house fire).
 
Protection is not provide by a protector.  Protection is always about where hundreds of thousands of joules harmlessly dissipate.  'Whole house' protectors protect even from direct lightning strikes due to a low impedance (ie 'less than 10 foot') connection to what does the protection: single point earth ground.  Plug-in protectors do not have that always required low impedance connection to earth; do not claim to protect from typically destructive surges.  Worse, plug-in protectors that fail get the naive to recommend them.
 
Since a destructive surge (ie lightning is one example) can be 20,000 amps, then a minimal 'whole house' protector is 50,000 amps.  Because protectors should provide protection for decades.
 
That indicator lamp only reports a potential catastrophic failure.  The light only says the protector was grossly undersized - should not have been there.  Normal failure mode for any protector is degradation.  That degradation must be so minimal that the protector is good for decades.  Again, protection is about where hundreds of thousands of joules are absorbed.  Only a properly earthed 'whole house' protector provides that protection.  And is necessary to even protect undersized power strip protectors.
 
Finally, the 'whole house' protector is only a 'secondary' protection layer.  Every layer of protection is defined by what harmlessly absorbs that energy.  Homeowners are strongly encouraged to inspect their 'primary' surge protection layer.  A picture demonstrates what to inspect; what defines protection:
 http://www.tvtower.com/fpl.html
 
Protectors are only simple science; connecting devices to what does protection.  The 'art' of protection is defined by the only component that must always exist in every protection system: single point earth ground.  Plug-in protectors do not have that essential earthing and do not claim to protect from the other and typically destructive surge.  Again, protection is always about where hundreds of thousands of joules harmlessly dissipate.

 

[LarrylLix]

Most surge  protectors are not surge protectors at all. Surge protector is a misnomer. Voltage spikes are not surges.
 
Using MOVs will clip and dissipate spikes on voltage waveforms but not handle  surges.  A surge is a series of sustained voltage waveforms that continually exceed the spec. maximum for many waveforms. The only real way to stop surges is using an isolated UPS or very complex and expensive tuned circuits, probably weighing hundred of pounds, depending on the current capacity rating.
 
For lightning noise a very educated power grid equipment company advised us years back to install disc capacitors in series with a bank of MOVs. The MOV's will conduct the high frequency noise from lightning hits and pass trough the capacitors like they were not there. When the 60 Hz power surge that follows the high frequency busts from the lightning hit comes the capacitors protect the MOVs by being high-impedance to the 60 Hz. After many repairs over ten years we never had another burn-out for my last 15 years there. This was in a 230kv hydro station environment so the faults were big and centred right on this equipment.
 
Most equipment can take power line surges no problem. A piece of equipment connected 120 vac that can't take a 200 vac surge for a second or so doesn't belong plugged in. However electronic equipment has a problem with 1-2 MHz bursts of 2-400 vac that lightning discharge and restrikes create.
 
This type of protection will do nothing for your X10 signal reliability, usually. A noise filter would be a better answer but they can get expensive and large with the tuned coils and capacitors that are required for larger current capacities.
 
In the end you take your best shot and pray.

 

[pete_c]

In the end you take your best shot and pray.

 
I have personally witnessed "willy nilly" DIY home electric done and what happens when a lightning bolt strikes. 
 
What I saw wasn't pleasant.
 
Most evident were light bulbs bursting; which I have never seen before or even historically mentioned to me.
 
Friend asked about using a surge protector in the panel. 
 
I said it would help a bit but if the willy nilly electric was not fixed it probably wouldn't work.

 

[mikefamig]

Most surge protectors use MOVs as the surge protection device (SPD).   Under normal conditions, the MOV looks like an open circuit and doesn't conduct any current.  But when a surge occurs, the MOV conducts, diverting the surge current to ground (or the other side of the line, e.g. hot to neutral).
 
I like to think of the MOV as a rope that frays with each surge.  At some point, when it has carried too much surge current, the rope breaks and the MOV is no longer capable of doing its job.  Manufacturers usually rate MOVs as being capable of a certain number of surges of so many Amps, such as 10 surges of 1000 Amps.  But if a single surge is greater than 1000 Amps, that uses up more of the MOV's capacity. 
 
What is supposed to happen with an MOV is that it conducts until it can handle no more, and then it turns into an open circuit. A large surge often causes the MOV to explode.  But this doesn't always happen, especially when the over voltage condition is small.  Then, the MOV may just remain as a short circuit, heating up and smoldering.
 
To protect against that case, a thermal cut out (TCO) can be added by the manufacturer of the surge protector.  There are also MOVs that have an internal TCO, and then it is called a TMOV.  If the temperature rises too much, then the TCO blows and the short circuit condition is removed.
 
So the bottom line is that the MOV may show obvious signs of damage due to a surge, but not always.  Some surge protection devices include "failure" LEDs, but depending on the design, these aren't always a reliable indicator either.  That makes knowing whether the SPD is still working or not a very frustrating exercise.

 
Then wouldn't it be good practice to replace these devices periodically assuming that their ability to protect the circuit is deteriorating over time due to occasional surges?
 
I had an interesting thing happen here. After installing the surge protector on my data bus the system returned to an occasional errant/false chime. I was very careful installing it and inspected all connections carefully and everything looked good so I replaced the surge with a new one and the system has been good for a week or so.
 
I bought these devices on Ebay for about $10 as NewOldStock against ~ $40 new so I guess that I got what I deserve. Maybe they were not so much NOS as just old and used.
 
Mike.

 

[westom]

Then wouldn't it be good practice to replace these devices periodically assuming that their ability to protect the circuit is deteriorating over time due to occasional surges?

 
Switches also degrade with use.  So replace all switches in all appliances annually.  Because best reasoning is subjective - ignores all numbers.
 
Those concerns about MOVs are promoted by manufacturers whose protectors are near zero.  Take a $3 power strip.  Add some ten cent protector parts.  Sell what is equivalent to your $10 protector for $25 or $90.  So yes, they did not lie.  They are preaching to consumers who only learn subjectively - do not obtain perspective - the numbers.
 
Anyone who believes a protector is protection is essentially scammed.  An effective protector is only doing what a hardwire does better:  make a low impedance (ie 'less than 10 foot') connection to single point earth ground.  Obviously that $10 protector has no earth ground, is too far away from earth ground, and read its specification numbers (it does not claim to protect from surges that typically do damage).
 
A properly earthed 'whole house' protector should be at least 50,000 amps.  Then a direct lightning strike (ie 20,000 amps) does not damage or degrade it.  Protectors that actually do protection should remain functional for decades - just like every switch on every appliance.

 

[mikefamig]

Switches also degrade with use.  So replace all switches in all appliances annually.  Because best reasoning is subjective - ignores all numbers.

 
Unlike a switch, a surge device might have no symptom when it fails.
 

Those concerns about MOVs are promoted by manufacturers whose protectors are near zero.  Take a $3 power strip.  Add some ten cent protector parts.  Sell what is equivalent to your $10 protector for $25 or $90.  So yes, they did not lie.  They are preaching to consumers who only learn subjectively - do not obtain perspective - the numbers.
 
Anyone who believes a protector is protection is essentially scammed.  An effective protector is only doing what a hardwire does better:  make a low impedance (ie 'less than 10 foot') connection to single point earth ground.  Obviously that $10 protector has no earth ground, is too far away from earth ground, and read its specification numbers (it does not claim to protect from surges that typically do damage).
 
A properly earthed 'whole house' protector should be at least 50,000 amps.  Then a direct lightning strike (ie 20,000 amps) does not damage or degrade it.  Protectors that actually do protection should remain functional for decades - just like every switch on every appliance.

This discussion is about a low voltage surge protector on an rs-485 data bus which is properly grounded and properly installed in proximity to earth and the device that it is protecting. I also think that RAL's remarks about MOV's was referring to surge protection devices in general.
 
I do have a whole house surge protector on my load center for the house power but that is another topic and isn't relative to this conversation.
 
Mike.

 

[pete_c]

What did you purchase Mike?  Just curious here.
 
I understand that you have protected the Elk RS-485 data bus from the home to the detached garage.  Thinking here you mentioned doing this externally eh?
 
Here and it's been many years ago I had a local lightning strike take out my two irrigation controllers which were inside of the house and connected to the outside solenoid / manifold cluster.  I did afterwards protect the RS-232 connections. 
 
I have to date though not protected the external LV sensors plugged in to the panel and wondering if you are planning to do this or is anyone else doing this with their panels or even if it is worth an effort to do? 
 
I do have a separate sub panel for power connected to these outdoor devices and direct connects to the alarm panel for NC/NO switch parts of them.

 

[westom]

I do have a whole house surge protector on my load center for the house power but that is another topic and isn't relative to this conversation.

 
RS485 (depending on its interface hardware) may already have protection even possible up to 2000 or 15,000 volts.  (RS-232 is harder to protect.)  So that existing RS485 protection is not overwhelmed, a 'whole house' protector must be properly earthed.
 
Nobody said anything about grounding RS485.  Ground a surge; not its victim.
 
Even the 'whole house' protector is not protection.  What the 'whole house' protector connects to defines the protection for RS485.  Is it connected low impedance (ie 'less than 10 feet, no sharp wire bends, etc) to single point earth ground?  That (and not a protector) defines low voltage (including RS485) protection.
 
 Protection is always about a surge current not inside the house.  Once that current is inside and hunting for earth, then nothing will avert its destructive hunt.  Protection of all low voltage devices is about earthing a surge BEFORE it can enter a building.  Then existing protection inside appliances (ie RS485 interfaces) is not overwhelmed.
 
How does one increase that protection?  Upgrade the single point earth ground.   BTW, how do telcos protect their $multi-million computers from surges on low voltage wiring?  Their COs suffer about 100 surges with each storm.  How often is your town without phone service for 4 days while they replace that computer?  Never?  Because they also use properly earthed 'whole house' protectors.  Increased distance between that protector and electronics also increases protection.
 
Protectors that must be replaced periodically are near zero (and many times more expensive) devices that do not claim to protect from another and typically destructive surge.

 

[mikefamig]

What did you purchase Mike?  Just curious here.

 
I used a Ditek LVLP-X for the rs-485 catx line and a  Ditek DTK-110C6A for the ethernet catx line. DEL recommended that he liked the Ditek brand and I chose these from their web site.
 

I understand that you have protected the Elk RS-485 data bus from the home to the detached garage.  Thinking here you mentioned doing this externally eh?

I ended up installing in a metal box inside the garage.
 

Here and it's been many years ago I had a local lightning strike take out my two irrigation controllers which were inside of the house and connected to the outside solenoid / manifold cluster.  I did afterwards protect the RS-232 connections. 
 
I have to date though not protected the external LV sensors plugged in to the panel and wondering if you are planning to do this or is anyone else doing this with their panels or even if it is worth an effort to do? 

 
I have no outside sensors or other devices but it is recommended that you protect all outlide wire with a device on each end.
 

I do have a separate sub panel for power connected to these outdoor devices and direct connects to the alarm panel for NC/NO switch parts of them.