Luteijn/Relay - Revision history

Luteijn: /* DONE: */

2018-01-25T08:36:02Z

DONE:

← Older revision		Revision as of 08:36, 25 January 2018
Line 257:		Line 257:
	* garage-door opener controller for NodeMCU as PoC - web frontend.		* garage-door opener controller for NodeMCU as PoC - web frontend.
	* garage-door opener controller for ATT45, serial 'front'end.		* garage-door opener controller for ATT45, serial 'front'end.
	* made a box for mains->USB power brick to connect as separate power supply for coils, but since the TCO of just using the ATT and/or USB-serial converter as the 'fuse' compared to running another powerbrick seems lower, just connected ~~everything~~ to the same Vcc from the USB-serial converter.		* made a box for mains->USB power brick to connect as separate power supply for coils, but since the TCO of just using the ATT and/or USB-serial converter as the 'fuse' compared to running another powerbrick seems lower, just connected the relay Vcc to the same Vcc from the USB-serial converter. NOTE: the enable pin feeds optocouplers, so the 'jumper' is open.

	====TODO:====		====TODO:====
	* invert logic on door open detection reed contact so open circuit when door closed?		* invert logic on door open detection reed contact so open circuit when door closed?
	* add more reed contacts at different voltages to detect e.g. fully opened door, or checkpoints along the way. maybe some kind of ultrasound distance meter to detect how far the door is opened?		* add more reed contacts at different voltages to detect e.g. fully opened door, or checkpoints along the way. maybe some kind of ultrasound distance meter to detect how far the door is opened?

Luteijn at 08:32, 25 January 2018

2018-01-25T08:32:08Z

← Older revision		Revision as of 08:32, 25 January 2018
Line 14:		Line 14:
	http://howtomechatronics.com/wp-content/uploads/2015/09/Relay-Module-Circuit-Schematics.png		http://howtomechatronics.com/wp-content/uploads/2015/09/Relay-Module-Circuit-Schematics.png

	I don't provide Vcc to the optocoupler, but use an 'enable' pin. Also, the idea is to not have a common ground. The relay board and a separate powersupply is of course way more expensive than the bare ATT45 I ended up using instead of an arduino, so if the relay borks and puts 30 VDC (or mains, if you're switching that) on the low voltage side, then it wouldn't matter to much that the Tiny is also fried. Still, especially during prototyping I'd rather keep things optically isolated. Now that it works, and I know the relay doesn't overload the Vcc, I've cheated and use one supply for both relay board and the Tiny.		I don't provide Vcc to the optocoupler, but use an 'enable' pin. Also, the idea is to not have a common ground. The relay board and a separate powersupply is of course way more expensive than the bare ATT45 I ended up using instead of an arduino, so if the relay borks and puts 30 VDC (or mains, if you're switching that) on the low voltage side, then it wouldn't matter too much that the Tiny is also fried. Still, especially during prototyping I'd rather keep things optically isolated. Now that it works, and I know the relay doesn't overload the Vcc, I've cheated and use one supply for both relay board and the Tiny.

	The relay-controller runs on an ATT45, which talks over (software) serial to a host computer (GLInet router in my case) and from there to the rest of the world. Access control is handled on the host computer, for flexibility and to save space on the ATT45. All output is also rather terse to save space on the ATT45. For the same reason, the code has its static strings broken up into smaller units and I use #defines instead of 'const' variables.		The relay-controller runs on an ATT45, which talks over (software) serial to a host computer (GLInet router in my case) and from there to the rest of the world. Access control is handled on the host computer, for flexibility and to save space on the ATT45. All output is also rather terse to save space on the ATT45. For the same reason, the code has its static strings broken up into smaller units and I use #defines instead of 'const' variables.

	As I was short on pins, I'm using the active low reset pin as an analogue input, making sure not to drop it too low to reset the Tiny. I found I needed a rather big Capacitor to decouple the reset line or when one of the relays switched on, the Tiny would reset due to what I suspect was (E)MI.		As I was short on pins, I'm using the active low reset pin as an analogue input, making sure not to drop it too low to reset the Tiny. I found I needed a rather big Capacitor to decouple the reset line. Otherwise when one of the relays switched on, the Tiny would reset due to what I suspect was (E)MI. It might have also helped not to run the switched 30V right next to the long leads connecting the reed relay to the reset pin...

	The reed contact that senses door state is normally kept closed by a small magnet. It connects the reset pin to ground via a pulldown resistor, there is also a pullup resistor, forming a voltage divider. Dimensioned so that Voltage goes from Vcc (5v from USB-serial dongle) when the reed switch (and the door) is open so just the pull-up is connected to about 4.2 volt when it is closed. Unfortunately it is a bit tricky to get the switch to be reliably open when the door is closed and closed when it is not. So there's always some current flowing here, but as the total R the Vcc is over is about 50kΩ, we're talking about 100μA, which I think we can afford, even if the thing would be battery powered, as the tiny is using 10 (50) times that running at 1 (8) MHz. Probably could use even higher value resistors if needed.		The reed contact that senses door state is normally kept closed by a small magnet. It connects the reset pin to ground via a pulldown resistor, there is also a pullup resistor, forming a voltage divider. Dimensioned so that Voltage goes from Vcc (5v from USB-serial dongle) when the reed switch (and the door) is open so just the pull-up is connected to about 4.2 volt when it is closed. Unfortunately it is a bit tricky to use a bias magnet to get the switch to be (reliably) open when the door is closed (and closed when it is not). So there's always some current flowing here, but as the total R the Vcc is over is about 50kΩ, we're talking about 100μA, which I think we can afford, even if the thing would be battery powered, as the tiny is using 10 (50) times that running at 1 (8) MHz. Probably could use even higher value resistors if needed to drop the sense current further.


Line 55:		Line 55:
	void setup() {		void setup() {
	pinMode(enablePin,OUTPUT);		pinMode(enablePin,OUTPUT);
	digitalWrite(enablePin, LOW); // make sure optocouplers are disabled asap.		digitalWrite(enablePin, LOW); // make sure optocouplers are/stay disabled asap.
	pinMode(doorPin,OUTPUT);		pinMode(doorPin,OUTPUT);
	digitalWrite(doorPin, HIGH); // active low, so enable current can be sunk, to light the led inside opto		digitalWrite(doorPin, HIGH); // active low, so enable current can be sunk, to light the led inside opto
Line 92:		Line 92:
	}		}
	void disable() { // pull enable pin down to de-arm the optocouplers		void disable() { // pull enable pin down to de-arm the optocouplers
	if (enabled) { // we call this ~~everytime~~ the watchdog times out, just to be sure, but no need to do feedback when not enabled anyway		if (enabled) { // we call this every time the watchdog times out, just to be sure, but no need to do feedback when not enabled anyway
	line();		line();
	Serial.write("D");		Serial.write("D");
Line 257:		Line 257:
	* garage-door opener controller for NodeMCU as PoC - web frontend.		* garage-door opener controller for NodeMCU as PoC - web frontend.
	* garage-door opener controller for ATT45, serial 'front'end.		* garage-door opener controller for ATT45, serial 'front'end.
			* made a box for mains->USB power brick to connect as separate power supply for coils, but since the TCO of just using the ATT and/or USB-serial converter as the 'fuse' compared to running another powerbrick seems lower, just connected everything to the same Vcc from the USB-serial converter.

	====TODO:====		====TODO:====
	* invert logic on door open detection reed contact so open circuit when door closed?		* invert logic on door open detection reed contact so open circuit when door closed?
	* print a box for mains->USB power brick and connect as separate power supply for coils?
	* add more reed contacts at different voltages to detect e.g. fully opened door, or checkpoints along the way. maybe some kind of ultrasound distance meter to detect how far the door is opened?		* add more reed contacts at different voltages to detect e.g. fully opened door, or checkpoints along the way. maybe some kind of ultrasound distance meter to detect how far the door is opened?

Luteijn: /* Project "Microcontrolled relay board": */

2018-01-19T19:14:21Z

Project "Microcontrolled relay board":

← Older revision		Revision as of 19:14, 19 January 2018
Line 19:		Line 19:

	As I was short on pins, I'm using the active low reset pin as an analogue input, making sure not to drop it too low to reset the Tiny. I found I needed a rather big Capacitor to decouple the reset line or when one of the relays switched on, the Tiny would reset due to what I suspect was (E)MI.		As I was short on pins, I'm using the active low reset pin as an analogue input, making sure not to drop it too low to reset the Tiny. I found I needed a rather big Capacitor to decouple the reset line or when one of the relays switched on, the Tiny would reset due to what I suspect was (E)MI.

			The reed contact that senses door state is normally kept closed by a small magnet. It connects the reset pin to ground via a pulldown resistor, there is also a pullup resistor, forming a voltage divider. Dimensioned so that Voltage goes from Vcc (5v from USB-serial dongle) when the reed switch (and the door) is open so just the pull-up is connected to about 4.2 volt when it is closed. Unfortunately it is a bit tricky to get the switch to be reliably open when the door is closed and closed when it is not. So there's always some current flowing here, but as the total R the Vcc is over is about 50kΩ, we're talking about 100μA, which I think we can afford, even if the thing would be battery powered, as the tiny is using 10 (50) times that running at 1 (8) MHz. Probably could use even higher value resistors if needed.


	<pre>		<pre>

Luteijn at 18:51, 19 January 2018

2018-01-19T18:51:51Z

Show changes

Luteijn: /* Project "Microcontrolled relay board": */

2018-01-19T16:56:18Z

Project "Microcontrolled relay board":

← Older revision		Revision as of 16:56, 19 January 2018
Line 7:		Line 7:
	}}		}}

			My garage-door-opener can be controlled by pulling a particular DC 20V pin on the controller down. This is meant for a (key) switch to control operate the door or turn on the LED lamp.
			I've wired the pin through two relays. One pulls the pin low enough (using a 100 Ohm resistor to form a voltage divider with a pull-up inside the controller) to activate the lamp, the other shorts it to ground, which will activate the door actuator. The control is rather crude, basically a pulse will stop a moving door, or starts a non-moving door. Direction is reversed from whatever was the last direction., although this is not 100% guaranteed it seems. Normally not an issue as you'd see the direction of the door and can just operate the switch to compensate.
			Of course, the idea is that the relays can be operated from anywhere, to let people in when unable to open a door by hand. So it's useful to be able to 'see' if the door is open or closed. This is sensed via a reed relay and a magnet stuck to the door. if the door opens, the magnet travels away from the reed relay, changing the voltage over it, which is measurable by the relay-controller.

			The relay-controller runs on an ATT45, which talks over (software) serial to a host computer (GLInet router in my case) and from there to the rest of the world. Access control is handled on the host computer, for flexibility and to save space on the ATT45


Line 18:		Line 23:
	* invert logic on door open detection reed contact so open circuit when door closed?		* invert logic on door open detection reed contact so open circuit when door closed?
	* print a box for mains->USB power brick and connect as separate power supply for coils?		* print a box for mains->USB power brick and connect as separate power supply for coils?
			* add more reed contacts at different voltages to detect e.g. fully opened door, or checkpoints along the way. maybe some kind of ultrasound distance meter to detect how far the door is opened?

Luteijn: document work with typical relay board for arduino to control garagedoor opener.

2018-01-19T16:42:25Z

document work with typical relay board for arduino to control garagedoor opener.

← Older revision		Revision as of 16:42, 19 January 2018
Line 1:		Line 1:
	~~Relais bordje~~		===Project "Microcontrolled relay board":===
			{{Project
			\|Name=Luteijn/Relay
			\|Status=In progress
			\|Contact=Luteijn
			\|Picture=
			}}



			====DONE:====
			* control typical 'arduino' 2-relay board with NodeMCU.
			* control typical 'arduino' 2-relay board with ATTiny 45.
			* garage-door opener controller for NodeMCU as PoC - web frontend.
			* garage-door opener controller for ATT45, serial 'front'end.

			====TODO:====
			* invert logic on door open detection reed contact so open circuit when door closed?
			* print a box for mains->USB power brick and connect as separate power supply for coils?

Luteijn: Created page with "Relais bordje"

2018-01-19T16:31:40Z

Created page with "Relais bordje"

New page

Relais bordje