Monitoring Power Status via LED

[rbroders]

Here is an update on an old topic...
 
I have an Onkyo TX-NR708 and I need to monitor power status to turn on my SubWoofer AMP and raise my center channel speaker.  I used to use this current sensor http://www.phenixcontrols.com/Current_Sensor/9993/, but when I turned on the Ethernet based automation interface, the Onkyo's power usage became unreliable.
 
Plan Z (all the easier plans exhausted - I couldn't even find a video output monitor) became monitoring the status of the StandBy light on the front panel.

This tiny little light goes on when the stereo is off, so I also had to invert the status.
 
The heart of the project is my favorite amplified relay: http://www.elkproducts.com/product-catalog/elk-924-sensitive-relay
 
After a little reverse engineering I have the schematic:

 
Next step is finding a photo transistor sensitive enough to pick up the light from the tiny standby led.
www.mouser.com/ds/2/311/pt80a_Pb_free_2008_08_14-37737.pdf
The LPT 80 A turned out to be perfect.  First it is sensitive to visible light (most of the phototransistors I found are IR only), and second, the transparent case makes it easy to position over the LED.
 
While the LPT 80 A was able to turn on/off with the StandBy LED, the transistor wasn't closing enough to control my sensitive relay.  To solve this, I added a 2N3904 in a common collector darlington arrangement to amplify the signal.  I literally soldered the 2N3904 to the LPT 80 A so the signal across the wires is strong enough for decent noise immunity.
 
Be sure to use enough black electrical tape to prevent stray light from activating the sensor.
 
To get the reverse effect, I used a 47K resistor to hold T- low and used the phototransistor darlington to pull T- up.  Thus when the sensor sees light, it raises T- and keeps the relay off.  When the stereo is on and the standby LED is off, the sensor turns off and allows the 47K resistor to trigger the relay.

 
 
--Bob

 

[BraveSirRobbin]

Hi Bob and thank you very much for taking the time to post this detailed explanation.  I would like to create an official "How-To" (giving you full credit for the project of course) when I get some time.
 
This method has me thinking about alternatives to a method I used to monitor an appliances status light as I used a CdS photo-resistor to trigger a transistor.
 
For instance, I'm going to have to look into (or maybe you know off hand) the sensitivity of the photo-transistor compared to the photo-resistor in the visible range (it would be nice to see sensitivity vs spectral range of the two on one graph).  Reason is I was having a great deal of trouble having the photo-resistor pick up the light from a dispersed source such as an LCD display (more specifically a Dexcom Diabetes Monitor).
 
I'm now curious if that, or another photo-transistor would have been sensitive enough to detect that (though its light is in a far different spectral range than your application).
 
Also thanks for posting the schematic to the sensitive relay from Elk.  (I've also always wondered why they put two relays on that board with only one trigger).
 
Another question I have is what are you triggering with your final contact closure output?  Reason I asked is have you ever looked into the possibility of using the collector-emiter junction of the transistor driving the relay directly into that digital input (bypassing the relay all-together)?
 
Regards,
 
BSR

 

[rbroders]

I saw your CDS photoresistor and defintely considered it.  The spectral match is much more toward the center of the visible.  However the devices tend to be much larger and clunkier with more of a 60's kind of vibe.  They are also slower than phototransistors, but not significant for our application.
 
I am driving a speaker lift which needs a Form C input (short A-B and it goes up, short A-C and it goes down), so the Elk-924 works perfectly for that.  I used the other relay for a 12V trigger to my SubWoofer amp.  Probably the transistor could have driven that directly.
 
Yeah I wonder why Elk didn't just get a DPDT relay instead of two SPDTs.  It would have halved the power consumption of the device.  Maybe they were on sale.
 
--Bob