Hello!
I plan to make notifications when incoming call (when ringing). My home telephone / intercom is Golmar T-5822VD with 2 wires only (BUS) where I measured 15,8—16V DC.
I don’t plan to control / trigger opening gate since I don’t want to solder wires directly to the board (button P1 in top-right corner).
I wrote simple JS script which is running on Shelly device and monitors voltage levels (from analog input - as voltmeter) every 2 seconds. When voltage changes more tan 2V it sends me a notification (HTTP GET request) directly from Shelly to my Gotify docker (https://gotify.net/) running on Raspberry Pi 4.
Warning
Powering Shelly device from BUS (where is 15-16V) is not a good idea because during ongoing call it makes some interference and nobody can hear anything on both sides.
Since I have a light switch 3cm above my telephone where is 24V DC, I will use that 24V to power Shelly Uni. And my project will be solved.
Hi, I like this setup too. Hopefully, someone can answer your question if this would work or not. I have bought a Shelly P1 and Shelly Plus Uni for the same project and had the same thoughts as you have. I also need to wire a door sensor contact. So that was the reason I bought both of the sensors at once. I don’t think I can get all to work with one of both sensors.
I don’t think this is possible it seems that the IN 1 and IN 2 need to be fed with a stabilized 5VDC input power. Hopefully, someone with a more electrical engineer background can confirm or explain this.
Hi @mr_density , could you please elaborate a little about how the shellies script is supposed to work here? I removed the Shelly integrations, installed MQTT broker, MQTT integration, python_scripts integration and the shellies discovery script.
I’d assume this would lead to discovered entitites under the MQTT integration, but nothing happens.
Have done several restarts of HA as well. Seems like I’m missing a step.
I have the Shelly 1 plus and I have setup everything and works like a charm. I have a notification when the button is pressed.
I have one more thing which decreases the WAF.
Shelly was configured as detached and I had HA press the doorbell for 1 second. But that is a bit long. Normal people just press very short and the minimum I can set is 1 sec.
Therefore I have set shelly to edge. Doorbell press is directly forwarded to the doorbell sound and I receive a notification via HA.
But I want to stop the sound around 20:00. Does anybody managed to switch shelly from Edge to Detached via a PUT request?
I have a two-wire Golmar Tekna V2 Plus, and with the inspiration from this thread and many others was able to connect it to HA. To help other folks with a similar model, here’s what I did:
Goals
Notify me when someone is buzzing the building door
Allow me to trigger the open button remotely
Don’t interfere with the existing functionality (hardware features should still work)
The Shelly Plus Uni is the star here, as it has a voltometer that allows you to detect the buzzer being pressed, and can be powered through 5V, so I currently have it connected to a USB power pack. In a future iteration I believe we can power it through the 24V bus with an appropriate step-down converter
Wiring Buzzer
The Tekna has two unused pins marked BUS OUT for connecting another unit or accessory. Normally this operates at 24v, but drops to 18v when the video screen is on (buzzer pressed). We connect these two pins to ANALOG IN & GND in the Shelly. When the buzzer is pressed, voltage drops and the Shelly sends a webhook to HA. The BUS IN pins operate the same way, so I suspect if you have a Golmar model without the BUS OUT you can use those pins instead.
Door Button
Pretty simple conceptually, just connect them to the button contacts on one end and the Shelly OUT 1 on the other. This does require you to disassemble the phone, which only requires a screwdriver and some flat tools for prying. To connect to the buttons I ended up twisting a wire around the contacts and dabbing it with a bit of glue to keep them in place. One of the wires didn’t have a good signal so I put a dab of solder on it as well. Also shrink tubing to keep the exposed parts of the wires from touching, but not necessary if you strip to the exact length in the first place.
Power
As I said, Shelly takes 5VDC so you can just get a USB power cable and connect it to the +5VDC pin on Shelly.
Configuration
Pretty straightforward, a few gotchas/tips:
Voltage will be “undefined” unless you switch the Voltometer Measurement Range from 0-15 to 0-30
To prevent the shelly from alerting when wires were disconnected or not powered (0VDC) I created a “Custom Voltage Expression” of Math.abs(x-18) and alerted on that custom unit instead
To have the button actually operate like a button (On for a half second or so), you have to set it up first in Input/Output settings (Button, Momentary) and then in Automations → Timers (Auto OFF)
I have had this working for some time now like the original post. But now i want to change my Doorbell button to something more fancy. Does anyone have an idea on how I would need to wire this:
This is the page one AliExpress. It is an No touch button, so people can hold their hand in front of the button and the bell rings. When their hand is removed, it will stop (at least that is my idea.
I’ve got a Shelly1 gen3 and a 230VAC Schneider brand doorbell.
The doorbell is super simple. Normally there is no power going to the doorbell. Pressing the button completes a circuit and sends 230VAC to trigger the bell. No additional wires for a push-button, just 230VAC L and N in, as you can see below:
The L & N going to the bell - are they energized at the bell location, when nobody is touching the button? Can you use some multimeter to test the voltage at the bell location?
Is the bell itself one of those electrical buzzer style, or is there a strike plate (or plates) and a hammer that would physically move via magnetoelectrical means, and “ding” the strike plate(s)…?
There is a variable 5VAC on those wires at all times. I attribute that to some sort of leak in my household wiring system. I’m pretty sure it’s not by design.
Yes, those are bells on either side of the electromagnetic relay.
Are the 2 lines coming out of the upper left corner energized at all time? 5VAC? 240VAC? Is it stable?
I mean, if the L and N screws are at 5V and only going to 240V when someone presses the doorbell… is the other blue & black pair behaving the same??
I am asking because if you want to do Shelly at all, you will need a way to power the Shelly (or anything wifi or ESP based) at all times. And that was exactly the reason why the very 1st post of this thread is talking about a 12VDC power source.
= = =
Otherwise, you will want to look into other devices that are battery powered. 2 ideas:
There are people repurposing a window/door contact sensor. With clever placements, you do not need the megnet half of the door sensor, and supposed the magnet field changes from the ding-dong strike plate mechanism would be enough to trigger the window/door sensor. example 1, or 2, or 3, or 4
There is also a thing that is called “Sage Doorbell Sensor”, that was zigbee and was discontinued maybe a decade ago… but still can be found on eBay from time to time. You will need to double check whether that would work with 240VAC wiring, though because that device was designed for the US market I believe. Here is a better read.
Sorry, I think my photo was not clear. There is only a single pair of wires here, they are twisted and double back around so it does look like there is a second pair.
The idle current on the wires is not stable, my multimeter shows it fluctuating from 3~6VAC. I have noticed this elsewhere in the house so I suspect it is just some ghost power in the circuit. According to the bell’s spec-sheet it doesn’t require any current when idle, only 230V to activate the relay.
I did try the magnetic door sensor trick but it was very fussy to actually activate it, a short press on the button just didn’t work, it needed a longer press, and when it worked there was a good 5+ second delay before I would get a notification from HA.
As I mentioned earlier, I can easily bring a new 230VAC wire down into the box to power the Shelly. For that matter I could also bring in 12VDC if that would be better. So powering the Shelly is not a concern.
I’m just not familiar enough with Shelly devices, so my main question here is whether running 230VAC into the SW when the bell is rung is the correct way to trigger the Shelly. I don’t want it to explode
The fact that you have to wait 5 seconds for any window/door contact does not sound right. What are the make/model of the door sensors you have tried?
Anyways, I have z-wave, and (at least in my case) my tests with my Ecolink door sensor is both consistent and immediate.
===
on the other hand, what you really need, conceptually, is a sensor that can tell you whenever the sensor sees 230VAC betwen L and N. So kind like a “power outage sensor”, but in reverse logic.
And to be clear, you do not need “power monitoring sensor” that tells you Voltage/Amp/Watt, etc., but just a “power outage” that tells you whether 230VAC being on or off. I have not seen any off-the-shelf sensor devices that you can buy and it would just work, so most if not all of the examples are DIY, and I can tell you wouldn’t mind DIY:
===
Do you have any link to the manual of the Schneider brand doorbell?
I am curious because is that really 230VAC both at the bell location and (gasp!) at the doorbell button location, that is outdoor and for people to press daily…? Is there a way to use multimeter to test the button outdoor, and see whether the bell button is indeed switching 230VAC…?
I mean, if so, I’d be interested to see how Schneider addresses safety concerns.
I got similar door bell with 230v AC which triggers the electro magnetic field and the bell rings. I have posted similar requirement. Check Here for the proposed solution. If you got same model of Shelly you may try that approach.
Edit: I don’t have any Shelly device so I could not confirm but as per the laws of electrical it should work.
