Problem - Open Collector output to a negative trigger relay not working

[Desert_AIP]

I have the ELK-930 dorobell detector.
I'm trying to get it to trip an HAI zone.
 
If I wire the Open Collector output of the 930 directly to the HAI zone, and place a 1K EOL resistor acros the zone, it properly triggers the zone, so I know the 930 is operating correctly.
The problem is the Omni has a fixed 300ms response time, so unless someone holds down the button a bit the zone does not trip reliably.
So I tried wiring the OC output from the 930 through a negative triggered time delay relay to extend the pulse, with the relay wired to the Omni.
 
I have used this same relay and technique to extend the 50ms pulse of the Hydreon rain sensor and an IR door threshold lighting trigger, and they work perfectly.
But one has a relay output and the other has a positive voltage output, not OC.
 
I have the zone wired across the NO terminals of the relay with a 1k EOL bridging them.
With 12VDC power applied to the relay, if I use a jumper to short the "-" side to the negative input terminal, the relay operates as expected.
 
I have the "-" side of power wired to the "negative" terminal of the 930. 
I have the "ouput" side of the 930 wired to the negative input of the relay.
After reading several threads here I also added a diode on the "-" side of the relay power.
 
When I press the doorbell button the doorbell rings, so the OC output should be bridged, but the relay does not operate.
Even if I hold the button down like I did when the OC output was directly wired to the Omni, it does not operate.
If I place a jumper across the "negative" and "output" terminals of the 930, the relay operates as expected.
 
 
I am stumped.
 
If I understand simple transitors and OC outputs correctly, negative power should flow from the negative terminal to the collector.
So I believe I have it wired correctly.
 
Do I need some kind of load resistor on the OC output to produce a current?

 

[pete_c]

I had issues with the HAI OPII zone thing (debounce) a while ago and added the Elk-960 debounce circuit board to the mix (Elk-930). 
 
Not sure if the attached diagrams will help you.  Analog reading shows 148.
 
I do have a diode in place today.
 
I have also recently been playing with some "el cheapo" $10.00 debounce circuits for mailbox sensors.
 
Recently also installed a little LED doorbell.
 

 

 
Curious about your Hydreon rain sensor and your methods of connecting it? 
 
Here I still have two in place; one is connected to a dual 1-wire counter and the other is just an adjusted RS-232 cable.
 
Mostly using the newer Davis tipping bucket and the older Dallas Instruments tipping bucket (its sort of big).

 

[Desert_AIP]

Pete, thanks. 
I actually used your diagram as a model to wire mine up.  I took the diode idea from your diagram.
 
I'm not using the Elk 960 board,
It's a Chinese e-bay board.
 
But I've used it before and it appears to be operating correctly and so does the 930, it's just not getting the trigger input from the OC output.
That's what has me so purplexed.
 
Even if the board I am using is not as sensitive as the Elk 960, it doesn't trigger even if I hold the button down.
I don't have an assistant to hold down the button while I meter things.

 

[pete_c]

Yeah disconnect the door bell wires and use a jumper maybe.  
 
You can test the button too by connecting an X-10 DS10A to the end of the wires of the doorbell and look at the button status outside with your cell phone?
 
I sometimes utilize those little tiny jumper cables with little teeth on them for circuit board testing.
 
My old old analog doorbell would get wet and play havoc with the OPII panel (leaving it in a water induced closed mode). 
 
The "el cheapo" debounce board has a test button on it. 
 

 

[Desert_AIP]

This is the Chinese relay board I'm using.
It looks similar to your el cheapo.
It seems to be operating correctly, it just doesn't like the OC input.
 

 
http://www.ebay.com/itm/160986663204#ht_5219wt_869

 

[mdesmarais]

I don't know the gear you are talking about, but I do know OC.
 

 
This is what OC typically means- the output will float. You then add a resistor from the output to the +VDC, which gives you a high. (to VDC) When the output turns on, it will pull the OUTPUT pin LOW, but only if it can sink enough current to overcome your pullup resistor.
 
Hope that helps?
 
Markd

 

[pete_c]

Wondering too how your doorbell & doorbell wires are doing? 
 
The wires to the doorbell here are in the frame of the inside panel of the door and I am more careful with these as it would be a real pita to try to run these from scratch. 
 
Here I use the Elk boards for the doorbell and the above mentioned board for my outside mailbox sensor.  It is inside the structure in a plastic sandwich bag now for a couple of years or so.  It still works fine. 

 

[RAL]

If I understand simple transitors and OC outputs correctly, negative power should flow from the negative terminal to the collector.

 
 
In simple terms, an open collector transistor acts sort of like a switch,  so on the 930, it is connecting the OUT terminal to the NEG terminal through the transistor.   But unlike a real switch, which will pull the input of the relay to 0 volts, the transistor can only pull it down to around 0.7 volts.    For the negative trigger input of the relay board, this may not be low enough to cause it to trigger.  Adding the diode to the negative power supply input of the relay has the effect of raising the reference point for the relay board slightly to a level that should work, depending on the characteristics of the diode you chose.   If it doesn't, you could try adding a second diode in series with the first one to raise it a bit more and see if that does the trick.

 

[Desert_AIP]

In simple terms, an open collector transistor acts sort of like a switch,  so on the 930, it is connecting the OUT terminal to the NEG terminal through the transistor.   But unlike a real switch, which will pull the input of the relay to 0 volts, the transistor can only pull it down to around 0.7 volts.    For the negative trigger input of the relay board, this may not be low enough to cause it to trigger.  Adding the diode to the negative power supply input of the relay has the effect of raising the reference point for the relay board slightly to a level that should work, depending on the characteristics of the diode you chose.   If it doesn't, you could try adding a second diode in series with the first one to raise it a bit more and see if that does the trick.

 
Hmmm...
I'll give that a try.  I'll do some metering.
 
The Elk 930 detector only has 4 terminals.
Two terminals are for the AC doorbell transformer, you splice in to one leg as depicted in pete's diagram above.
I assume it is a current sensor.
The other two terminals are "Negative and "Open Collector Output"
 
So I believe the Negative terminal is connected to the common/emitter side of the transitior, the Open Collector is the collector pin of the transistor (not connected to anything else), and the base is connected internally to the current sense circuit of the board.
When the doorbell operates and pulls current, the transitor turns on and essentially shorts the negative and OC pins.  or that is what I expect to happen.

 

[Desert_AIP]

Wondering too how your doorbell & doorbell wires are doing? 
 
The wires to the doorbell here are in the frame of the inside panel of the door and I am more careful with these as it would be a real pita to try to run these from scratch. 
 
Here I use the Elk boards for the doorbell and the above mentioned board for my outside mailbox sensor.  It is inside the structure in a plastic sandwich bag now for a couple of years or so.  It still works fine. 

I assume the wires to the push button work fine.
It is illuminated and the doorbell operates properly.
When I hooked just the 930 to it it operated OK and the doorbell still rang. 
The pulse was just too short to reliably trip the Omni.

 

[Desert_AIP]

I don't know the gear you are talking about, but I do know OC.
 

 
This is what OC typically means- the output will float. You then add a resistor from the output to the +VDC, which gives you a high. (to VDC) When the output turns on, it will pull the OUTPUT pin LOW, but only if it can sink enough current to overcome your pullup resistor.
 
Hope that helps?
 
Markd

 
After a bit of googling around this may be the issue.
I don't have a pull up resistor in place.
I may need a resistor from the +12VDC terminal to the OC output, making the input to the relay +12VDC when off.
When the detector detects, it connects the collector to the emitter and pulls the terminal to ground, creating the negative going transition. 
 
Since the Omni passes current from the "+" to the "-" side of the zone, this may have been providing the current to generate the transition when I had just the board hooked up to the zone.
 
If so, I'm wondering how pete got his circuit to work.
Maybe the Elk 960 has an internal connection to keep the negative trigger terminal high?

 

[pete_c]

Yup that "shock delay" pot referenced on my board changes the debounce time up to 60 seconds or so.  I do not recall what I adjusted it too.  The board came with a vibration sensor and some reference to being a motorcycle alarm board.
 
You got me looking at mine.  Its in the rafters in the basement.  It does have a resistor in place.  I am assuming it's just an EOL resistor.  I'll take a picture in a little bit.  It looks to be going across the terminals rather than in series.   A close up picture will validate the connection as its mounted high up and there is no light there.

 

[RAL]

I don't have a pull up resistor in place.
I may need a resistor from the +12VDC terminal to the OC output, making the input to the relay +12VDC when off.
When the detector detects, it connects the collector to the emitter and pulls the terminal to ground, creating the negative going transition. 
 

 
 
I'll hazard a guess that the trigger input of the relay board has an internal pullup on it.  Without that, it would just be floating, which could cause problems even when a regular switch was used as the trigger mechanism and the switch was open.
 
It won't hurt anything to try adding a pullup, though. 

 

[Desert_AIP]

I'll hazard a guess that the trigger input of the relay board has an internal pullup on it.  Without that, it would just be floating, which could cause problems even when a regular switch was used as the trigger mechanism and the switch was open.
 
It won't hurt anything to try adding a pullup, though. 

 

You are correct.
I metered the boards when I got home.
The relay board supply voltage meters right at 12VDC.
The voltage between negative power terminal and the negative input is 10.5-11VDC with nothing but power connected to the board.
 
I disconnected both the negative and output from the Elk board, but left the telephone transformer wires in place.
I metered between the negative and output terminal and it was a positive voltage (10V or so) that steadily declined. 
So I assume the cap on the board was discharging as I was metering it.  It had to have been charged by the 10VDC from the relay board when I had it hooked up.  The transformer wires are on the same leg, the board just splices in.  0 VAC between them.
 
As I said, a jumper wire from the negative power terminal to the negative input triggers the board.
 
I tried wiring a blocking diode in numerous orientations between the terminals of the OC output and the negative and trigger inputs of the relay board.  Nothing worked.
 
I'm going to need to jumper the button wires so I can meter the OC terminals when they are active.
They should pull low, but I don't understand why they aren't triggering the relay board.

 

[RAL]

That's some helpful information.  Here's what I think would be helpful to measure next.
 
Connect the OUT terminal of the 930 connected directly to the negative trigger input of the relay board, and the NEG terminal of the 930 connected to the negative power supply terminal. 
 
You could wire the negative power terminal of the relay board directly to the negative power supply terminal, or through a diode. Might be interesting to try it both ways.
 
Then, with the doorbell button shorted,  measure the voltage at the negative trigger input of the relay board.   I would expect this to be less then 1V.    If you connect the diode to the negative power terminal of the relay board, also measure the voltage across the diode.  It should be about 0.7V.