Would T-junction help low current situation

Your second approach won't work either (you may have just drawn it incorrectly).

You need to only connect the relay coil across (in parallel with) the bell. The relay will get power when the bell does.

BELL |----| | |----------------------GND | |----RELAY------| |----| | |--------SWITCH-----VCC

Bob

How does a relay coil with ground on each end of it ever get energized? If the relay is rated for something close to

The problem is that I can't get to the Vcc wire between the switch and bell. It is inside the wall.

Well, then you'll have to go back to your original idea and either use a relay with a very low coil resistance or an opto-coupler. This way, you'll only lose a couple of volts across the bell.

One other idea comes to mind -- get a Sawz'all.

Bob

Try sensing the bell current with a transistor. See above:

You might need a resistor from B-E to shunt some of the bell current, but if you use something like an MJE170 or other medium size power transistor it should be able to handle about 1 amp. Don't forget to add a diode across the relay coil and also perhaps one across B-E. If you have an AC bell then a similar method will work, but it would need some extra components. You might also be able to use a SSR or optoisolator, with a bridge rectifier if AC, and an appropriate current shunt resistor.

Paul

I can

Or, you can wire the doorbell button straight to the relay coil, so that (when the button is pressed) it activates the relay and one pair of contacts rings the bell, while another pair of contacts informs your computer.

If you ground the 'common' terminal of an SPDT relay, the one branch can have a pullup resistor and your computer, the other can complete the bell circuit. With multipole relays, it's even easier.

This way, you're sensing the doorbell switch position, which is a simple task, rather than the high current of an operating doorbell.