I'm installing some surge protection for a radio installation, and I have a somewhat unusual requirement of installing the surge protection device (SPD) at an Earth ground where the antenna feedline enters, which is about 100' from the main panel. This is because the antenna by virtue of its design must be grounded, it's not feasible to run the feedline to the existing electrical service ground, and so effective surge protection requires limiting the surge voltage at another ground location, where the feedline enters the house.
Many of the SPD I find have three wires: one for neutral, and one for each phase. The presumption seems to be they are installed at the main panel where neutral and ground are bonded.
In a previous question I was told there are SPDs with four wires, separating neutral and ground. Thus, they can limit the voltage between ground and neutral without requiring a direct connection between them away from the main panel, which would be hazardous. However, I need to protect only one, 120V, single-phase circuit. Using one of these four-wire SPDs, one of the wires would go unused.
I'm wondering, is it safe and permissible to use a 3-wire SPD (like the SDSA1175 I have in my hand) in this situation? The neutral connection would go to ground, one line connection to line, and the other line connection to neutral? Despite the naming clash, I can't imagine this is electrically different than the four-wire case, essentially a MOV between ground and each protected wire. Neutral isn't directly connected to ground, but in the event of a surge the MOV would clamp it within home hundreds of volts relative to ground.
Best Answer
I've used that exact part exactly as you described before on a 30A commercial circuit without incident.
Now, do I know that it has protected equipment from surges? No. But No damage has occurred since its installation.
Do keep in mind, that one of the 'ready' lights won't come on, because each ready light depends on a separate hot having voltage potential between it and what it thinks is neutral. It's likewise not unrealistic to expect any faulty indicators to fail similarly. But the MOVs will do the same job regardless.