Elk-930 doorbell troubleshooting

[jacdx]

Hey guys,

I'm FINALLY getting around to doing my Elk-930 install and I'm running into issues. So, it's a Spore LED lighted doorbell (specs here: http://sporedoorbells.com/storage/trueinfo.pdf) and I've only got one. I've got it wired up to a HC-800 contact and the Elk.

I wired it all up exactly as the Elk-930 instructions say, and have looked at prior threads on here. Programming-wise, I added a Doorbell, connected it to the HC-800 contact, programmed it to play one announcement when the doorbell is pressed, and a second when the doorbell is released (for testing). That was all sorts of erratic. I could only get the announcements to fire when I was handling the Elk module and twisting screws.

Then I ditched the annoucements and setup the events to turn a light on when doorbell pressed and turn it off when released. That shed some light on the subject (zing!). The light stays on constantly. Pressing the doorbell doesn't make it go off either. I've double checked and triple checked my wiring. Even swapped out the Elk module for a second one, same issue. Thought 'Invert Sensor' might help in Composer, but it didn't change anything.

Without the Elk module wired, I can cross the wires connected to the controller and they behave as expected, the contact closes and the light goes on. Uncross and the light goes out. I just can't seem to get the ELK to recognize the state of the doorbell correctly.

My transformer is a Broan 16.5V. I measured 19V on it. I'm thinking either the transformer is outputting too high so the Elk can't measure the voltage drop correctly...or the draw from the doorbell is too low for the Elk to detect and open contact state. Any ideas?

Edited January 12, 2013 by jacdx

[jacdx]

Hmmm... like good wine and every girl I've ever dated, it "needed some time". We went to dinner for an hour, came back and the contact is now closed (like it should be) until I press the doorbell. The doorbell now works through C4 correctly, announcements fire correctly. Now, if I can get the announcements to work faster... Currently it takes 3 seconds from doorbell press for the sound to start. When I had the light programmed to it, it would turn on/off instantly.

[LSDave]

With the ELK modules you should use a pull up/down resistor. Any type of sensor that has an open collector output being hooked up to Control4's contact inputs needs to be referenced to the 12v or GND terminals. a 10Kohm resistor should be fine. My Elk worked without it, but i would have "phantom" announcements due to the floating input on the controller.

[jacdx]

Thanks. I'll see if I can dig up a resistor. Where would it go on the circuit? I'm using 12+ and sig on the controller. Would it bridge those at the controller?

So, this issue isn't quite as cut and dry as I thought. I came into the kitchen this morning, the can lights turn on because of the motion sensor and the test light for the doorbell goes on with it!!!!! :o Surely mistaken, I turned the can lights off, doorbell light went off. Dimmed them up past 10% and the doorbell light would trigger a press and hold it until I turn off the cans. I turned the can lights on and off a few times just to make sure that I'm not crazy. And the doorbell test light would go on anytime the can lights were on. The doorbell test light is in a different room...not related in any way to my kitchen programming.

Late last night when I wrote the reply, the cans were never on, since the motion programming turns on an 'evening' scene instead at night.

My Elk & doorbell are high on a cabinet in the kitchen, about 4ft from a 75w halogen can light. The Elk has been just sitting out in the open while I get this figured out.

So, I turn the can lights on, doorbell press triggers and holds. I go up to the Elk and put my hand over it to block the light (not touching) and the doorbell release event fires. Move my hand, the doorbell presses and holds again. I took a scrap piece of opaque plastic and covered the Elk and left it for now. Now it's acting correctly, can lights on or off.

Could the Elk be light sensitive? Or is this setup so sensitive without a resistor that a nearby can light or halogen could trigger the contact via some sort of interference?

Edited January 12, 2013 by jacdx

[LSDave]

I would have to double check when i get home, but i think the resistor goes between the sig and gnd. It basically pulls the SIG line to a high impedance ground, then when the ELK module is triggered the SIG goes high(12v). Otherwise, the SIG line kind of floats in the middle.

[rosenqui]

For those (like me) who find this thread when doing a Google search for ELK-930 AND Control4, I wired mine up this way based on the above advice:

OUT  <----->  +12
NEG  <----->  SIG <--+
                     |
              GND <--+ 10k resistor

I added a doorbell contact device to the project and then connected it to CONTACT 1 on the HC800.

 

I found the /var/log/debug/driver_event.log file helpful to tell if the controller was seeing any change in the signal. If so, you get events like this (assuming the device is named "Doorbell"):

2015-05-07 08:14:48 -0400 home-controller-800-XXXXXXXXXXXX [3462] INFO: Driver Fan Speed Controller(94) fired event Unknown(4567)
2015-05-07 08:14:48 -0400 home-controller-800-XXXXXXXXXXXX [3462] INFO: Driver Keypad Dimmer(250) fired event Unknown(4568)
2015-05-07 08:14:51 -0400 home-controller-800-XXXXXXXXXXXX [3462] INFO: Driver Doorbell(396) fired event Unknown(1000)
2015-05-07 08:14:51 -0400 home-controller-800-XXXXXXXXXXXX [3462] INFO: Driver Doorbell(396) fired event Closed(2)
2015-05-07 08:14:51 -0400 home-controller-800-XXXXXXXXXXXX [3462] INFO: Executing command (EXECUTE_MACRO) on driver Macros(260)
2015-05-07 08:14:51 -0400 home-controller-800-XXXXXXXXXXXX [3462] INFO: Driver Macros(260) fired event Unknown(4)

I did this to get only the Doorbell events since you tend to get a lot of activity from switches etc. in that log file:

/var/log/debug# tail -f driver_event.log | grep Doorbell

So far I haven't had any false positives or missed button presses, but I'll keep an eye on it. If anyone else has wired theirs differently I'd be interested to hear how you wired it & why.

[rosenqui]

I'm posting on this old thread again because I've been getting phantom events from the doorbell contact lately. It was working quite reliably until the last few months, and the wiring hasn't changed so I'm trying to figure out what's going on.

See my post directly above for how I've wired the ELK-930 into my C4. The ELK-930 is wired like in "Method 1" from the instruction manual here. At least I think I've done it correctly that way.

I've swapped out the 10K resistor to see if that will help, but I'm wondering if something else might be wrong. When I measure the voltage at the ELK-930 across the OUT and NEG terminals in an idle state (everything wired up, but doorbell button not pressed) I get around 11.5VDC and it looks like about 2.2mA. The C4 is showing the contact as "Open", so I guess the pull-down resistor is doing its job. I don't have enough hands to measure the voltage when the doorbell is pressed, but I'm guessing it goes to the full 12V are several hundred mA.

I know that the open collector output on the ELK-930 can float a bit, but is 11.5VDC and 2.2mA normal? Should the voltage be lower or is the low amperage what makes the C4 see the contact as open?