If you have used "ice cube" type PCB mount power relays before, you're probably familiar with the de facto standard pinout for them:
Why is this such a standard pinout though? It's clearly not an optimal pinout — positioning the common terminal between the coil terminals forces you to put an isolation slot in the PCB in order to obtain reasonable creepage distances, and severely constrains the clearance distance available as well. It's also not something unique to cheap Cheese-shop-specials either: the Omron G5LE series uses this pinout, and so do equivalent relays from TE/P&B (ORWH) and Panasonic (JS1). You need to go to much costlier parts such as a Tyco PCH or RZ or Panasonic JW1 in order to get something that puts the common pin on the same side of the relay as the other contact pins.
Is there some sort of internal construction detailing that makes this type of relay unsuitable for mains isolation to begin with? Or why can't the relay manufacturers bring the common terminal out on the "correct" side of the relay to allow an isolation barrier to be established?
Best Answer
It's only to make the relay cheap. It's sub-optimal internally (electrically) as well as externally (pinout). The coil-to-contact breakdown voltage and capacitance is inferior to that of better relays. On the other hand it's cheap and the relay can be made reasonably sensitive (360mW typically for that construction).
One of the disadvantages of this construction is that the contact current always flows through the flexure, so a large surge (say to blow a fuse or circuit breaker) can anneal the spring and affect the operation permanently.
Below are a couple of photos of one I did a tear-down analysis on a few years back.
The common pin is naturally located at the other end from the contacts. They could have moved the coil pins to the front or back but either way they are close to the contact potentials unless the relay is made wider.