lightening damage and how to protect?

When a ground wire goes through a ferrous hole, like a metal electrical box, it acts like a choke to high frequency. Lightning surges are mostly  high frequency bursts, due to overshooting the electrical discharge, and then firing some electrons back again, repeatedly, in a ringing pattern.
 
For a decent ground try to go as straight, and large of a conductor as possible, without passing through any metal holes. Receptacles are not good grounds for this type of protection. Great for 60Hhz faults but high frequency lightning disturbances are different.
 
These are the components I used for my fiber system:
 
- Pre-terminated Indoor/Outdoor fiber assembly from Lanshack.   LC connectors both ends.   Multimode 50/125 OM3.  With pulling eye both ends.   Lanshack has some good videos on how to lubricate, pull, and cut the pulling eye off.
- Wall mount Termination box for both ends.   Got these from Lanshack too
- Fiber patch cables for both ends.   These are available from a variety of sources and are inexpensive.   
- You can get media converters from a variety of sources and they are inexpensive.   Just make sure that they are Multimode with LC connectors.   You may need a SFP transceiver if the media converter does not come with one.   
- You can also use a switch that has a SFP slot instead of the media converter.   Like the Unifi Switch POE8
- For the connection to my OmniPro II, I used a RS485 converter from Serialcomm on both ends.   The model number is SER-FIBER-MM    It works with RS232, RS485, and RS422.   It was plug and play.   No configuration required.
 
I have not had any problems with my system since I converted to fiber.
 
ano said:
Wireless is the way to go, at least for internet. Optoisolators seem like a good idea at first, and they work fine for isolating a few hundred volts, but these devices, which usually are maybe 1/2 inch in length, aren't going to do much from a lightning strike. And with wireless, make sure your using good voltage protection on the power supply, or the wireless part won't be of much help.  
I agree that wireless is best for isolation but would require more than the basic wifi equipment for my case.  For some nodes in the house that are connected to outside wires (sprinkler for instance) I will be moving to wireless.  For wired best is clamps for limiting the voltage and isolation.  Optos can easily do more than "a few hundred volts".  Our products are tested to 2500VDC routinely with surface mount parts.  The leaded DIP parts give more spacing and even come with a wider lead bend for more clearance for the pads on the PC board.  5000 VAC ratings are fairly common - nothing exotic.  Here's an example:
https://www.everlight.com/file/ProductFile/EL851.pdf
 
JimS said:
I agree that wireless is best for isolation but would require more than the basic wifi equipment for my case.  For some nodes in the house that are connected to outside wires (sprinkler for instance) I will be moving to wireless.  For wired best is clamps for limiting the voltage and isolation.  Optos can easily do more than "a few hundred volts".  Our products are tested to 2500VDC routinely with surface mount parts.  The leaded DIP parts give more spacing and even come with a wider lead bend for more clearance for the pads on the PC board.  5000 VAC ratings are fairly common - nothing exotic.  Here's an example:
https://www.everlight.com/file/ProductFile/EL851.pdf
Yup, and after the lightning burns out the LEDs or phototransistor and fills the IC with carbon, the insulation/isolation value is 0.
 
Had many of them burn through in electrical grid station SCADA systems. They are good for inhouse differential voltages, but not actual lightning disturbances.
 
LarrylLix said:
Yup, and after the lightning burns out the LEDs or phototransistor and fills the IC with carbon, the insulation/isolation value is 0.
 
Had many of them burn through in electrical grid station SCADA systems. They are good for inhouse differential voltages, but not actual lightning disturbances.
Interesting.  Substations see higher voltages from electrical storms than a typical residence I would think.  That's what all the literature says anyway.  6kV is considered the realistic max for residential but of course if you have a direct hit it's a gazillion.  Still there is lots of equipment that can't withstand 6kV (industry standard pulse waveshape/time).
 
JimS said:
Interesting.  Substations see higher voltages from electrical storms than a typical residence I would think.  That's what all the literature says anyway.  6kV is considered the realistic max for residential but of course if you have a direct hit it's a gazillion.  Still there is lots of equipment that can't withstand 6kV (industry standard pulse waveshape/time).
Yes, but it is about where there optoisolators can stand 2500v. Across the internal LED or transistor components they can only stand about 5 to 100 volts.
 
LarrylLix said:
LarrylLix, on 16 Aug 2020 - 11:24, said:
Yes, but it is about where there optoisolators can stand 2500v. Across the internal LED or transistor components they can only stand about 5 to 100 volts.
I don't see how the voltage limit of the LED or transistor is the real issue. How much voltage supply can the fiber optic transceivers handle? Probably not 100V. And there are ways to make the LED and transistor more immune by the circuity around them. The voltage withstand of the LED and transistor circuit are important but the bigger issue is the isolation voltage.

I regularly work on 3.3V microprocessors that are connected directly to 480V power without isolation. They work fine. Just don't connect to them without isolation.
 
JimS said:
I don't see how the voltage limit of the LED or transistor is the real issue. How much voltage supply can the fiber optic transceivers handle? Probably not 100V. And there are ways to make the LED and transistor more immune by the circuity around them. The voltage withstand of the LED and transistor circuit are important but the bigger issue is the isolation voltage.

I regularly work on 3.3V microprocessors that are connected directly to 480V power without isolation. They work fine. Just don't connect to them without isolation.
The LED section can handle up to about 2 Vdc and the phototransistor likely up to about 100-400 Vdc.
 
Lightning does not follow the design paths for normal voltage protections. With it's MegaHertz ringing bursts it will not even follow insulated conductors if they bend too sharply. If it did radio signals would not penetrate your walls.
 
Burn out the LED out inside an optoisolator and the carbon will not insulate the brag spec. of 2500V. Those insulation specs are only for healthy components.
 
IOW. Opto-isolators work as designed for  as designed. system and voltage applications.  Lightning doesn't usually fall into those categories.
 
Of course if you feel the first layer of protection against lightning and ground voltage gradients has been looked after the isolators may do the jobs just fine. Look for fully shielded independent transformer  power supplies with grounded winding shields to isolate your 120Vac common points also. Running both sides off the same power supply nullifies any benefit of an opto-isolator.
 
When I built my shop that is about 250 ft from the house I bought some fairly cheap bulk 6 fiber cable (50/125)  I found also on ebay some pre-polished LC connectors made by Leviton  49991 fastcam is the part.  only real specialized tool is a fiber cleaver.  Basically strip the fiber per the dimensions, cleave to correct length, insert carefully, press the cam release latch on the connector and remove.  If your conduit is big enough for pre-terminated that is easier, unfortunately the conduit that was buried many years ago to the old barn that was replaced with the new shop was only 1/2"  so had to terminate the cable in the field.  works great.  
 
My Caddx extension module has optical isolation built into the board.  So I am not contemplating if I should put surge protectors on both ends of the run between buildings.  Putting them on both ends will put more stress/voltage on the house (non-opto) end.  So the real question is if the optos are enough on the garage end.  I am going to look again at the circuit to see if there are any line to ground clamps on the extension board.  There were some large MOVs but I think they returned to the communication common terminal for the cable and not chassis ground.
 
JimS said:
My Caddx extension module has optical isolation built into the board.  So I am not contemplating if I should put surge protectors on both ends of the run between buildings.  Putting them on both ends will put more stress/voltage on the house (non-opto) end.  So the real question is if the optos are enough on the garage end.  I am going to look again at the circuit to see if there are any line to ground clamps on the extension board.  There were some large MOVs but I think they returned to the communication common terminal for the cable and not chassis ground.
The opto-isolators provide electrical isolation (up to a point) between the two systems so you won't need a common circuit ground between the two systems.  But that's not that same as surge suppression.
 
If you have opto-isolators only on the garage end, your panel back in the house is still wide open to a surge from the wires running outside to the garage.   The opto-isolators will protect the extension module in the garage up to the point that their voltage rating is exceeded.  Then, they will likely arc over to other components in the module.  All the energy from the surge has to go somewhere, and it will find a path through the rest of the circuitry, damaging it.
 
With surge protectors, you are clamping the surge voltage to a level that is below the damage limit of the expansion module components and giving the energy a path to ground through the protector.  That's something the opto-isolators don't do. Also, you will be intercepting the surge before it enters the building and has much of an opportunity to reach the expansion module.
 
Back
Top