Smoke alarm retrofit - concept check

Other Hardware General Electronics
1

Hi all,

I want to make smart a smoke alarm (of course I will have non-modified ones just in case). I have read the datasheet, I just wanted someone more experienced to cast their eye over it.

Does this look like it will work hardware wise?

Here is the circuit layout I am proposing, a basic voltage divider to sense the state of pin 7. With my voltmeter it reads 8.6V when sounding, and 0V normally. As the datasheet state there is a built in 280K resistor I am thinking I will be fine with a voltage divider in the 160K/100K ratio.

I chose D5/GPIO14 as it does not get pulled low or high at boot, so will not trigger an alarm on the bi-directional pin 7 IO.

Overview proposed (small).jpg
Overview proposed (small).jpg2016×1512 878 KB

pcb existing (small).jpg
pcb existing (small).jpg1489×1280 658 KB

The datasheet say this about the IO pin 7:

3.5 Interconnect
The bidirectional I/O pin allows the interconnection of
multiple detectors. In a local alarm condition, this pin is
driven high immediately through a constant current
source. Shorting this output to ground will not cause
excessive current. The I/O is ignored as an input during
a local alarm.
The I/O pin has a 280k nominal pull-down resistor, so
the pin may be left unconnected.
The I/O pin also has an NMOS discharge device that is
active for 1 second after the conclusion of any type of
local alarm. This device helps to quickly discharge any
capacitance associated with the interconnect line.
If a remote active high signal is detected, the device
goes into remote alarm and the horn will be active.
Internal protection circuitry allows for the signaling unit
to have a higher supply voltage than the signaled unit,
without excessive current draw.
The interconnect input has a digital filter that ensures
filtering out pulses of up to 300 ms. Filter pulses will be
ignored and not affect internal timing of the part. This
allows for interconnection to other types of alarms
(carbon monoxide for example) that may have a pulsed
interconnect signal.
The remote alarm delay (370 ms to 1.27s) specifies the
time from the interconnect going active to sounding the
piezo horn alarm.

Thanks for any advice, and for reading!

Overview proposed (small).jpg
Overview proposed (small).jpg2016×1512 878 KB

pcb existing (small).jpg
pcb existing (small).jpg1489×1280 658 KB

2

I don't see anything obviously wrong with that (although experience tells me someone else will be along to point out what I missed) but why don't you just try it?

3

I’ll try it this week thanks

4

I take the view that this hobby is about trying stuff out and learning from what happens, what you are doing seems to me to be an excellent opportunity to try something and, err, see what happens!

Do please report what you find.

5

You are using pin 7 to detect an alarm only. That is, you are not attempting to trigger an alarm. Is that correct?
It looks like your smoke detector will then have 2 power sources, the existing 9v battery and a usb power supply. I suppose you could try also to use a boost converter to replace the battery (if the smoke detector can tolerate a ‘noisy’ power source).

6

PerryBebbington:
I take the view that this hobby is about trying stuff out and learning from what happens, what you are doing seems to me to be an excellent opportunity to try something and, err, see what happens!

Do please report what you find.

Hi again,

Let me know what you think, all working but could be tidier perhaps.

7

Thanks for the update.
The photos are difficult to view as they are zoomed in too much. If there's a way to zoom out I am not aware of it.

Does it do what you want? If so then good :slight_smile:

8

Yes they are way too big, github does not seem to resize them. I'll take a look at maybe adding some smaller ones embedded in the description.

Yes it does work at expected, thanks for the support! :wink:

9

Yes it does work at expected, thanks for the support!

Good!
And you're welcome :slight_smile: