UART connection between ATSAMD20 and ATtiny4313

Yes, we are thinking about this solutions too.

No problem with capacitive loads.

Ceramic resonator has three pins, two that goes to the oscillator pins of the MCU and one is GND.

The program seems stopped. Sincerely I didn't measure the oscilator frequency after the noise, so I don't know if it stopped completely and/or it is disturbed only during the noise.

Anyway, the total effect seems a stopped program. I noticed that in some cases, this noise generates strange behaviours: GPIO connected to LED toggle without a reason and so on.

Reply to
pozz
Loading thread data ...

The current used for door bells - even modern monstrosities with cameras

- is tiny. I really think something must be wrong with the card if it is susceptible to interference from such cables. Oscillators - crystal or ceramic - usually need load capacitors and a damping resistor. These are small loads (perhaps 15-20 pF and 10 MΩ), and are sometimes built into the microcontroller. But if these are missing, your oscillator could be unstable and easily affected by outside influence.

Reply to
David Brown

Il 05/05/2023 12:59, David Brown ha scritto: [...]

In my installations they don't use that kind of door bells that are digital circuits with digital audio output. I don't know what's their exact name, similar to high power buzzers.

Reply to
pozz

If it is some solenoid driven thing with a mechanical switch it may cause all sorts of interference, depending also on luck - coincidence between button being released, solenoid switch, sine wave peak etc. Try to determine whether the interference which gets you is inductive or capacitive, or if your board is powered off the same lines. Things like that can be tricky to fix, you need to understand the nature of what gets you first.

Reply to
Dimiter_Popoff

You are talking about exactly like a buzzer, but with a clapper for the bell. A mechanical switch interrupts the current to the coil and very likely arcs on opening and closing, generating significant EMI. Perhaps you could convince them to add a capacitor like they used in automotive spark systems? The spark should be in the right place.

Reply to
Rick C

People use 230VAC for _doorbells_?!?! Are they trying to wake the dead?

Here in the US doorbells are usually 16VAC, though 24VAC is also used. I've also seen 10VAC and 12VAC mentioned, but I think that's uncommon.

Reply to
Grant Edwards

They are very common here (for example [1]). 230Vac is the common mains voltage so the doorbells are connected to mains, through a momentary switch out of the door.

[1] https://comenzielectrice.ro/pdf/Bticino%20-%20Matix%20Catalog.pdfSearch for AM5048
Reply to
pozz

I saw it... very well.

I will try. It's much more simple to avoid the problem at all disabling internal oscillator.

Reply to
pozz

Ok, thanks for suggestion.

Reply to
pozz

I don't know about where you live, but doorbells in the US are powered from low voltage, so they don't require all the safety precautions. A device on a high voltage circuit has to be mounted on a box or has to be a box rated for such wiring. A door bell uses much simpler wiring and needs no special boxes or precautions. I have seen no small number of doorbell buttons mounted directly in the door frame, having been drilled out to 1/2 inch for mounting the button.

Reply to
Rick C

Guessing from the URL, he could be in Romania, or a country near enough to use Romanian webshops.

I have no idea about the regulations here in Norway, but I know our doorbell has a transformer in the fuse cupboard, so it does not have high voltage mains out to the bell. Driving mains to the doorbell sounds like a safety nightmare. (Mind you, it's one way to deal with annoying kids that ring the bell and run away...)

Reply to
David Brown

Another thing to check is exactly where the noise gets into your system. Maybe it is actually coming in through wires (such as power supply wires) connected to your box, rather than directly from the bell wire to the PCB traces. Do you have any connecting wires that might be run parallel to the bell wire, acting as antenna? If so, try a capacitor on those lines as a simple EMC filter.

Reply to
David Brown

My board is not so sensitive. You need to put the cables near my plastic enclosure, exactly over the resonator. If you route the cables far from the resonator/oscillator pins, the problem disappears, even if I tie all the cables together (ring cables and 12Vdc power supply cables of my board).

Reply to
pozz

So you have a dedicated power supply for doorbells? Here 99% of house doorbells are 230Vac. The installation is much more easy, because you use directly the mains cables without additional hw.

The pushbutton that activates the doorbells are a normal button (but momentary) that could be used to activate lights in the room.

In the past, the doorbells were dedicated box mounted on the wall, near the door, inside the apartment.

Now they tend to install as a module inside push-button frames that are "inside" the wall ([1] just as an example).

Anyway my board is not near the buzzer, but could be near the pushbutton that activates the buzzer.

[1]
formatting link
Reply to
pozz

Similar here (Bulgaria). When I was a kid starting to try to put together this or that a "doorbell transformer" was a popular option to make some sort of power supply. I think they had a few outputs at the secondary, starting at 6V or something. But even so if the buzzer has nasty solenoids and switches it can cause plenty of interference, who knows.

Reply to
Dimiter_Popoff

Yes. A small transformer that costs a couple dollars. It allows use of far cheaper wiring and switches than would be required for 120V and avoids requirements for approved junction boxes, etc.

Reply to
Grant Edwards

LOL, when you say "dedicated power supply", you mean this.

formatting link
The high voltage side of the transformer is mounted on a proper high voltage box. The rest of the wiring is low voltage, so requires no particular safety precautions. The power is low, so even at the low voltage the wire is not heavy, typically 20 or 22 gauge.

Doesn't the button require a much more substantial mounting, as well as the bell itself? That's what the low voltage is avoiding. I think my transformer is mounted in the pantry ceiling not unlike a lightbulb (which is also in the pantry ceiling).

I don't know what that means "normal". Here, all wiring junctions or terminations must be done in an approved box.

formatting link
The wire junctions shown here are valid, but not common in the US. Mostly people are used to "wire nuts", which are like this, and work well.

formatting link
The mistake that is often made when using them, is to not twist the wires together before twisting the nut on. The nut is not intended to be a mechanical connection, just electrical. So the insulated portion of the wire should be twisted together first.

The bell itself is not an issue. The PITA is the button.

formatting link
These are much easier to mount where they are most convenient. See the small button with the screw terminals? Below is a crude video of a similar button showing the minimal requirements for installation. This guy is upgrading to a video doorbell.

formatting link
Much simpler and less expensive than running high voltage wiring to the button.

I'm wondering if your doorbell has the low voltage transformer inside and the wire to your doorbell is also low voltage? It just makes zero sense to run high voltage to the button.

Reply to
Rick C

I understand a small transformer is cheap, but anyway it's cheaper and simpler not having it at all.

Here cables for doorbells are routed together with other cables, for example for lightning, so they uses the same junction boxes. Moreover the push-button for the doorbell is identical to the one used to switch on a lamp. It's not uncommon to have a 230Vac frame with push-button for doorbell, push-button to activate outdoor lamps and a RFID reader to disable the anti-burglar system. All these three elements in a frame like this.

[1]
formatting link
Reply to
pozz

The price on this link is 23 euros, I think a 230Vac doorbell costs less.

Here they usually route the doorbell 230Vac cables together with other cable of electrical system (lightning...) using the same junction boxes.

See [1]. That is a 3-modules frame that can be installed in the wall. It's already compatible with 230Vac push-buttons. With one of this you can have the push-button for an outdoor lamp, a push-button for the doorbell and a keytag reader for anti-burglar system (for example).

You don't need an additional substantial mounting dedicated for doorbell pushbutton.

Yes, of course. The same box used for a switch that turns on a light. There's no need to add a dedicated box for doorbell.

If you have only a doorbell pushbutton outdoor, yes it could be simpler. However here usually installs a "normal" in-the-wall 3-modules box, so you can have multiples commodities: doorbell push-button, button for a light and so on. This box is already compatible for 230Vac signals.

[1]
formatting link
Reply to
pozz

LOL You don't seem to understand. You either have the transformer as a separate entity, or its functionality is incorporated in something more expensive, because it is rated for 240V.

I think you mean "lighting". But you can't lump the doorbell with anything else. At least not in existing construction. It is all irrelevant. Do what you wish. Spend the money on 240V wiring at a dollar a foot and high voltage boxes. That's your choice. But the fact is, the low voltage doorbell is much cheaper to install.

You put your outside lights on an outside switch? Doesn't happen here, at least not at the front door.

Your photo looks like an indoor switch. Outdoor switches are typically much more robust and sealed from the elements. Which one is the doorbell switch? Or is all this inside the front door and a visitor opens the door to ring the bell?

Reply to
Rick C

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.