On Mon, 26 Oct 2020 23:03:09 +1100, Chris Jones
wrote:

snip

I didn't know if the buzzer trigger could / should be done by a self
latching circuit of some sort or if one should use a zener to manage
the voltage allowing a better release action?

Do you even need any trigger/latching circuit?

Possibly not, with the right solution. ;-)

Often doorbells only make
a sound when the button is pressed,

This one is just a solenoid hitting one chime (Ding) and then dropping
back to the other (dong) when the button is released.

so the visitor should not be
bothered by that, and it would not require such a trigger circuit.

Ok ...


Why not do pretty much exactly what you already suggested:

Get a big capacitor (perhaps a couple of 1 Farad 5.5V supercapacitors in
series, they are quite small and should withstand 6V even without
equalising resistors), charge the capacitor from the existing doorbell
wiring via a 1N4004...1N4007 diode,

Ok, so when I first hooked it up it will pull the bell in 'Ding' and
hold it in till the charging current dropped back enough to allow the
solenoid to release (which might be slow so we might not ever hear a
clear 'dong')?

and put a piezo or mechanical buzzer
connected in parallel with the diode in the direction with the positive
of the buzzer connected to the cathode of the diode. During charging,
the buzzer will experience 0.7V of reverse bias which probably won't
bother it, and when the button is pressed, the diode is reverse biased
and the buzzer receives whatever voltage the capacitor was charged to.

Ok. I follow that.

For simplicity, don't bother with any extra series resistor to limit the
charging current, as that would expose the buzzer to more reverse bias,

What if the buzzer was (across your diode) in series ... with another
reversed diode and resistor in parallel?

eg, charging current passes though original diode / buzzer, combo,
through the resistor and into the cap. Then when the button is pressed
the cap charge passes through the second diode, though the buzzer and
to the switch (and back to the -Ve of the cap etc)?

and it is probably unnecessary after the first charging cycle. The
capacitor won't have time to discharge much when the button is pressed,
which means it won't take long to recharge either.

What sort of leakage do these caps have Chris?

When the circuit is
initially installed, it might take a while to charge the capacitor,
perhaps a minute or so.

Understood, or I could pre-charge it? ;-)

So you need capacitor, diode, buzzer.

plus another diode and suitable resistor ...


If the existing bell has a lot of inductance, when the button is
released there might be a large voltage kick across the switch.

Well it is a solenoid so it might have a bit.

The
diode ought to catch this so I would not expect it to fry the new piezo
buzzer,

I was hoping to go for a electro mechanical buzzer so that it actually
resonated the door frame a bit, rather than just be a beep in parallel
with the existing doorbell.

but it might alter the sound of the existing doorbell when it
releases.

Not if I get the charging resistor right?

Basically, I don't think I need a big cap, just one big enough to
allow the buzzer to sound briefly, just to provide feedback to anyone
outside that the button did work (and ideally I wouldn't want it to
run much longer in case they did hold their finger on the button).

I'll have to rig something up and see how it goes (and it's easier to
find 6V buzzers). ;-)

Cheers, T i m