I am making a device that will have an external alarm in case of alarm condition or power outage. I was thinking of using a solid state relay with a NC connection, and applying current to hold it open while there is power, and if there is alarm or power outage, the current drops out and the relay will close, making the connection to sound the alarm.
Will the relay last for years if it is essentially held open all the time? I can only find data on number of operations, which is about 100,000 times.
Here is the relay I am looking at (I know it is 10 amps and I can use smaller, but I am using 4 other of these on the board now...i will likely change it later).
The device you linked to isn't a solid state relay, it's a mechanical relay. I'm not aware of any solid state relays which are NC when power is removed from the control circuit, although it would be fairly straightforward to make one. Will it be switching DC from a battery? If so, it would be very easy to do the switching with a mosfet.
I would expect a mechanical relay to be reliable if it is normally held open. Its coil power dissipation at nominal voltage is only 360mW, so keeping it cool should not be difficult.
I'd say a mechanical relay is the perfect solution!
Anyway, the only thing to worry about is heat, then? There is no other wear on the unit?
Mechanical relays tend to be very reliable. But they are mechanical devices... Corrosion can cause them to get "stuck on" (rare) or it can cause resistance in the contacts. "Exercising" the relay tends to "wipe" the contacts, and it might be a good idea to exercise (and test) your relays once a year or so. If there is enough voltage on the contacts, that can help to "break through" the corrosion too. (People sometimes have trouble with small-signals through high-power relays.)
Where I work, we have a product that has several small relays. We have a spec for contact resistance and if the resistance is slighly high, there is a utility in our test program to exercise the relays. That usually brings them into spec. Sometimes we get "stuck" relays during production & burn-in, but we rarely get a field-failure. If a unit is returned for repair, usually something else is wrong (not a failed relay). Sometimes the resistance is out-of-spec, but never enough to cause a problem for the customer. And I'd say most of the time, the exercise routine will bring it back into spec.
DVDdoug:
I'd say a mechanical relay is the perfect solution!
I'd say that if it's for switching DC, then a mosfet is a better and more reliable solution. Bear in mind that the relay will need a transistor to drive it. This makes the component count similar, i.e. 4 components for the mechanical relay (assuming no snubber is required), vs. 4 or 5 for the mosfet (depending on the voltage to be switched).