If the wire gets hot enough to vaporize copper but not Iron, It is going to set the house on fire, anyway. There are also more corrosion problems with copper weld than with copper. I think I (in my inexperience) would go for a heavier copper wire rather than a thinner copper weld one. The copper needs more support, than the iron does, though, to keep it from work hardening and breaking from repetitive movement in the wind. But a rod grounding conductor should be well supported against the magnetic forces of a strike, anyway.
If distance to rooftop remains short, then RS-232 is sufficient assuming other external noise sources are not strong. Even with shielded and twisted pair wires only 200 feet and using a baud rate of 9600 baud, we suffered unacceptable noise problems.
Second, rooftop weather station and computer must be powered from a power source sharing a common safety ground. This also made necessary because RS-232 is common mode signalling - to avoid ground loops. Voltage differences between computer power source and weather station power source can create noise in RS-232 signalling. Again, this caution only so you avoid surprise problems. Simple ICs can interface computer serial port to RS-485 cable meaning your choice was not 'fixed' by the embedded computer.
Original problem is about protecting a computer from lightning. That means RS-232 cable cannot connect direct from computer to weather station. Cable from weather station to computer must first connect down to earth ground. Notice how CATV service or telephone connects to building. They do same so that lightning strikes are made irrelevant.
If the path down to single point ground increases that distance beyond the 100 feet, then consider RS-485. RS-485 only mentioned because that RS-232 cable may need be longer than you originally thought. Make plans so that RS-485 can be easily implemented later as you learn from experience (IOW run a 4 wire cable - not a 3 wire).
Important facts were provided by but a few others. For example, opto-isolation may only provide 2000 or 4000 volts isolation. Lightning will overwhelm that protection - again - if you don't provide lightning a better path to earth ground. Ben Franklin's 1752 demonstration is essential for understanding lightning protection. Many even hype surge protector and opto-isolators without first learning how damage happens.
John Popelish provided good solutions. No matter what you do, without that single point earth ground, then no protection exists. RS-232 cable first connects to single point earth ground before rising back up to connect to embedded computer. Overhead protection such as a lightning rod or catenary wire can also connect that direct strike to earth ground. Why? So that the direct lightning strike does not find earth ground destructively via the computer.
This is what you want to avoid. As Ben Frankl> I'm using an embedded serial driver, so I am limited by 50-100feet,
Every solution to his specific problem. But to make a general statement, I think there is an exception. A metallic enclosed system, with no signals related to Earth potential is quite immune to lightning. Even electronics inside an automobile can sometimes stand quite a strike (depending on the electrical continuity of the body).
I've used RS-232 over 1500 foot cables, so 100 feet should be no problem for you. You can trade off some of the drive by putting RC circuits on the ends of the cable to route spikes to ground through capacitors and not through the chips.
This varies from place to place and with the age of the house.
Some places there is a wire from the fuse box to the water pipe. Some have a ground rod directly below the power meter. Some places, there are two ground connections. The OP really should loook at what he actually has before going forward. Placing a grounded conductive object up on the roof is a good way to fool Mother Nature into thinking it is a lightning rod. Mother Nature doesn't like to be fooled.
Most places, they use an iron wire or a "copper weld" wire. Iron vapor is harder to make than copper vapor. The difference in resistance isn't super important in this case.
Many a wooden barn has survived a lightning strike like that. You have to press something bright red hot to a hunk of wood for a fair time to get a self sustaining fire going. If its raining and the wood it wet it takes even longer.
Lightning often hits trees and wooden structures, causes parts of them to explode and yet does not lead to a self sustaining fire.
I googled about on the subject and could only find one reference for the type of wire and that said "copper".
If EM field transients were so destructive, then every car nearby a direct lightning strike would suffer damaged radios. This is example 1. Antenna would concentrate lightning created, destructive fields on the transistor most susceptible to damage - the RF amplifier. But that transistor is not damaged.
Principles to make an electronic circuit safe from nearby strikes is simple and routine. Furthermore, circuits constructed to not suffer damage from direct strikes (and yes, even a direct strike to AC electric utility wires is a direct strike to the house) are so resilient as to make field induced damage irrelevant.
Why does a nearby strike cause damage? Understand the circuit - example 2. Lightning must connect from a cloud three miles up to earthborne charges four miles distant to the left. What is the shortest path? 5 miles through air to those charges? Of course not. Shortest path is from cloud, directly to earth somewhere to the right. That may become a direct strike to the human body. Charges enter earth, then travel four miles from right to left. In the process, electricity travels a shortest path which is up the right leg and back down the left leg. This second example is really a direct strike to human (and why four legged animals are at even greater risk).
Demonstrated by that human body is also why electronics are sometimes damaged. But many instead speculate damage must have been fields from the nearby strike. Again, if that speculation were true, then (example 1) every nearby portable car radio, cell phone, TV connected to household antenna, etc (the first example) must be damaged. Reality - surge damage because current traveled into and out of electronics - just like a human who was standing with legs apart when lightning struck nearby.
To make claims of EM fields causing damage, one must provide numbers. Such claims are never provided with numbers. IOW damage by fields from a nearby strike are based totally on speculation: its a power field; therefore damage must occur.
How little are those lightning generated currents? Example
3: a long wire (100+ feet) antenna is nearby a direct lightning strike. Currents from those fields are so large that ... they light an NE-2 neon lamp. Milliamps current flow through that tiny neon glow lamp without causing lamp damage. EM field generated currents from a nearby strike are so little that even an NE-2 glow lamp conducts them without damage.
Example three provides ballpark numbers. Those numbers - and all those undamaged portable and car radios in example 1 - demonstrate damage from nearby strike is a myth. But then I also demonstrated how nearby strikes do cause damage - example
2 - that others *assume* must be from EM fields. Example 2 is also why earthing for surge protection must be single point.
Mark Thomas has asked how to make his weather stati> Yes, but.
1) The strike current is rather high and can induce high voltages inside a car (seen radars transmit thru a quonset hut wall and the echo off of birds come back).
2) A "fork" can go thru a window and bounce all over the place, creating damage (seen results of that).
People presumably ask for advice, but what they really want is to be told that whatever they are doing is the right thing. Ego Massage.
If someone *want* to do something, they should just *do it* - and be "men" about the *consequences*!!! Instead of wasting time and presenting like pathetic whingers and worriers.
I have seen "managers" ask more than 10 people for "advice" until, finally, some nitwit (or accomplished office politician) does a replay of the managers proposal - which will then be Good Advice.
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.