JVDE X10 Forum

General Discussion => X10 General Topics => Topic started by: bkenobi on December 05, 2013, 11:41:33 PM

Title: Socket rocket failure
Post by: bkenobi on December 05, 2013, 11:41:33 PM
I have a few socket rockets that usually sit in my miscellaneous bin.  I use one of them during the holidays to control exterior Xmas lights using a bulb to plug adapter.  Last couple years this worked great, but this year the module is unresponsive after being turned on for a while (gets stuck on).  Is this common for socket rockets?  Is there a way to fix them or do I just toss it?  I'm going to swap it out this weekend, but I'm curious where I should put the current one (garbage or project bin).
Title: Re: Socket rocket failure
Post by: Brian H on December 06, 2013, 05:52:39 AM
If it was stuck ON as soon as power is applied. The Triac is shorted. A common problem if the bulb burns out and there was a power surge from the filament opening up.

I would also suspect a power supply issue if it works cold and gets flaky when On for awhile.

Biggest challenge is getting it apart with out breaking things so it can't be reassembled. Maybe informative to give it a look inside but reassembly may not be easy. I have one opened up and would never try and put it back together.

There is some data on this web site on disassembly of a LM15A.

Title: Re: Socket rocket failure
Post by: bkenobi on December 06, 2013, 06:14:47 AM
Sounds like a power supply then since it works initially and then fails to respond after a while.  It actually stays on if I cycle the switch which is not the way it should work from my recollection.  I'll drop in the replacement and put this one in the tinker pile.
Title: Re: Socket rocket failure
Post by: bkenobi on December 19, 2013, 06:55:15 PM
Soon after installing the second (brand new) socket rocket, I found the lights acting abnormally again.  This time, the lights turn on and off with the switch but do not respond to any x10 control.  1x could be a faulty module.  2x is most likely something else.  I thought the light strands were using a total of 90W which would be easily handled by the socket rocket.  My clamp meter indicates around 2A though (240W).  Well, there you go...I toasted the modules!  Oh well, I'll do something different next year.
Title: Re: Socket rocket failure
Post by: Brian H on December 20, 2013, 03:25:47 AM
Do the lights have any type of power supply that may have messed things up?

If you have a chance. See if the socket rocket works after cooling down. I have done some stress tests on Lamp Modules. If over heated they frequently stopped responding to commands before destroying themselves.
Title: Re: Socket rocket failure
Post by: bkenobi on December 20, 2013, 08:20:45 AM
The original module would work correctly after cooling down.  It was scheduled to turn on at dusk and off at 11pm.  When I went to bed, I would have to manually flip the switch on that set of lights.  In the morning, if I turned the switch back on, they would stay off but would turn on when a command was sent.  This second module turns back on in the morning with the switch.  I have not tried to send a command though, so perhaps the Socket Rocket is supposed to remember the state prior to power loss.

As for the lights, these are icicle style Xmas lights that have 150 bulbs each.  I thought they were 30W each, but now that I think about it, that was for a different strand.  ~4 strands (a couple sections intentionally disabled) is actually pulling 2A which equates to 240W.  So, I guess each strand is pulling close to 60W.  60W/150bulbs=0.4W/bulb.  According to www.christmaslightsetc.com:

Mini Lights

LED minis draw only 0.069 watts, while incandescent mini lights draw approximately 0.408 watts.

I'm using incandescent, so I'm right were I should be.  These are straight incandescent, so there's no power supply.
Title: Re: Socket rocket failure
Post by: bkenobi on December 21, 2013, 09:15:41 PM
I pulled the last failed unit this evening and feel pretty lucky, actually.  The outer casing was completely melted and charred around the heat sink.  Since the module was obviously dead, I dissected it and things were way worse than outside.  The triac was charred and cracked and the plastic casing black and very brittle.  The board it was attached to was black and fell apart when removed from the casing.  One of the capacitors (c1 if you have a schematic) was inside out.

Basically, the module failed just soon enough such that it didn't burn down the house.  The lesson to take away from this? The module says it's good for 150W Max...don't double that!

The only good thing is that I found a better way to disassemble these so they are more likely to be reassembled.  If you have a soldering iron and desoldering iron, you can remove the solder on the tip, the two sides (just above the threaded section), and the solder blob inside the female threaded region.  After these are removed, the whole thing will slide apart without damage to the circuit or casing.
Title: Re: Socket rocket failure
Post by: Brian H on December 22, 2013, 07:20:26 AM
Wow. Thanks for sharing the burning issues.  ::)

I think I saw on an X10 related web site. Someone else had found a similar way to open a LM15A.
Good tip that should assist others.
Title: Re: Socket rocket failure
Post by: bkenobi on January 28, 2014, 11:25:58 AM
I came up with a permanent solution for this issue.  I changed the outlet for one of the strands and installed a new outlet for the other.  I'm now going to be using OutletLinc's set with an X10 address to control these outdoor lights going forward.  I shouldn't have any more melted modules due to overloading the modules now!