X10 Community Forum
🔌General Home Automation => Automating Your House => Troubleshooting Automation Problems => Topic started by: bigmac on July 30, 2012, 05:53:58 PM
-
Just checked out my daughter's light switch after she told me it quit working this weekend. It only controls two 40 watt incandescent bulbs in a ceiling fixture. I took the switch cover off and smelled that noticeable burnt smell. The switch has a burn hole and is melted in the bottom right corner (have pic, not sure how to post????).
Thinking of pulling all my X10 switches/modules.
THIS COULD HAVE BEEN BAD!
Has anyone else seen this issue?
Thanks,
Merle
-
Burning smell with a hole in the case does not sound good.
Was it in service for a long time?
Not sure if a Newbie status allows a picture but you may want to see if the additional options tab near the bottom left of the reply screen give you any choices.
-
It's been in service about 18 months or so.
-
At about 18 months. It was probably a soft start version where it ramped On and Off.
There isn't a whole lot of user information on the new soft start ones.
We have seen older ones in photos and could maybe have made an educated guess to what failed from the approximate location of the failure.
-
I replied once to this thread, but it seems to have disappeared.
Bigmac
If any peace of mind, it is very rare for an X10 module to go incendiary. I have been a regular since the forum first opened (2003?) and this is only second or third report of fire. Also, I believe X10 uses a flame retardent plastic in their line operated modules
Interesting that it was a WS467 which does not have a neutral line, meaning there was no direct short between line and neutral. I doubt it is a triac problem. Would be nice if you could post pictures and perhaps we could figure what part went poof.
-
Thank you Dave. Gives me a liitle sense of security....................
I have a pic but not sure how to post it? The "insert image" icon just adds the (http://) hypertext????
-
Photo didn't make it. The Additional Options Tab near the bottom left of the reply screen didn't have an attach feature where you could add your photo?
Since the WS476 is UL approved. I believe they require a flame retardant plastic case.
I am thinking the safety slide switch. Used to break the circuit for bulb changing. May have been failing. As all the current for the load and the electronics passes through it it got warm and got worse. Kind of like a domino effect.
-
The "insert image" icon just adds the (http://) hypertext????
You have to supply a URL to the image.
-
Let's try this.................... (http://i1255.photobucket.com/albums/hh625/bigmac4442/X10_WS467_fried.jpg)
-
WOW that is one nasty looking WS467.
-
Yes, it either got so hot or flamed out that it deformed one of the wire nuts!
-
I don't have a WS467 to pop apart, but since it has two burn spots, I'm going to guess one of the big power supply reactor caps shorted.
-
here's the inside. Caps look OK???
(http://i1255.photobucket.com/albums/hh625/bigmac4442/2c19574a.jpg)
(http://i1255.photobucket.com/albums/hh625/bigmac4442/909b26d0.jpg)
-
The caps dave w was talking about are not electrolytic.
One of them looks like it is just left of the big burned blob and below the left tuning coil.
Was one or both of the light bulbs in the fixture burned out?
Almost looks like the load was shorted to neutral or ground and the excessive current just went wild.
-
It looks like the larger NON electrolytic cap might have shorted. (it ain't there no more!!!) WOW!! :o :o :o
OR
The "lamp change" switch was making poor contact, and over heated the board and surrounding components.
This looks like a NON softstart wall switch.
The newer softstart wall switches have the processor in a different location, a daughterboard (AGC circuit), and two 10 ohm resistors in series rather than the one 22 ohm resistor.
(top picture is old style wall sw., bottom is new softstart)
Note the two 1/2 watt resistors tied together (extreme right in picture), and the processor is near the top and horizontal.
My picture of the new wall switch I have modified for the neutral wire connection (grey wire next to the larger blue cap).
The larger blue cap is part of the power supply and the smaller blue cap is part of the PLC coupling network from the 120Khz transformer.
-
Thank you for the photo of the new Soft Start.
Until your post. I have never seen the new ones insides.
Helpful from me.
-
Here is a schematic of the non softstart wall sw module.
-
Thank you for the photo of the new Soft Start.
Until your post. I have never seen the new ones insides.
Helpful from me.
You are welcome! Just note the differences... they are minor but obvious and when you see one, you can tell which one you have.
-
Here is a picture of the case cracked open and where the big blue cap is in relation to your burn.
-
I probably will not see one open.
I have no problem opening one. Closing again is a different story. rofl
Now that X10 modules are getting scarce and I have moved to a different protocol but still having X10 devices. I would not be needing any WS467s. I still have a few older ones in a storage tub.
-
I probably will not see one open.
I have no problem opening one. Closing again is a different story. rofl
Now that X10 modules are getting scarce and I have moved to a different protocol but still having X10 devices. I would not be needing any WS467s. I still have a few older ones in a storage tub.
Yes, seeing that these are rather pricey now, any modification must be done with CARE!!!
Closing them up isn't real hard, just a little tricky!
As far as the cap failing... even IF the .68uf were directly across the line, it shouldn't fail, so I think that the cap shorted internally (failed part), OR something exceeded the breakdown voltage and punched through the dielectric and cooked the cap. After all... the cap is GONE! Only the leads remain! WOW WOW WOW!!!!
-
It looks like the larger NON electrolytic cap might have shorted.
OR
The "lamp change" switch was making poor contact, and over heated the board and surrounding components.
The larger blue cap is part of the power supply and the smaller blue cap is part of the PLC coupling network from the 120Khz transformer.
I'm sticking with my "power supply reactor cap shorting" guess. If the switch was standard vertical mount in a wall box I think if it was the power disconnect bar switch over heating, you would have more damage to the top of the switch, not two burn spots on back and side. $0.02. Unfortunately I would have preferred it be the switch causing the problem....
One has to wonder what would cause a non electrolytic, mylar cap, in service for almost two years, to short. Too bad the damage is too severe to see the original markings on the cap.
-
The one reason ANY cap could short would be due to a high voltage spike that would punch a hole through the dielectric. Physical damage could make it fail as well, but not likely for a mylar like this as it would be obvious and take alot of damage.
Electrolytics can fail due to the dielectric drying out (usually results in very low value or open).
Of course, age could do that too, but I doubt the mylar cap, being sealed with an epoxy coating, would do that.
Like I said also, even if the cap were to go directly across the line, being a high voltage rating cap, it should survive. After all, these kind of caps are used for noise suppression as well (see X10 signal suckers) and are across the line.
Yes, I doubt the switch got hot and failed... just a guess.
-
It looks like the larger NON electrolytic cap might have shorted.
OR
The "lamp change" switch was making poor contact, and over heated the board and surrounding components.
The larger blue cap is part of the power supply and the smaller blue cap is part of the PLC coupling network from the 120Khz transformer.
I'm sticking with my "power supply reactor cap shorting" guess. If the switch was standard vertical mount in a wall box I think if it was the power disconnect bar switch over heating, you would have more damage to the top of the switch, not two burn spots on back and side. $0.02. Unfortunately I would have preferred it be the switch causing the problem....
One has to wonder what would cause a non electrolytic, mylar cap, in service for almost two years, to short. Too bad the damage is too severe to see the original markings on the cap.
When you say power supply reactor cap, do you mean the 220uf electrolytic? IF that shorted, the only thing that would happen is the module wouldn't work. IF it opened, same thing... no workee.
The 22 ohm resistor MAY cook but the V- is about 18volts dc, and another 10% drop across the 22 ohm shouldn't make the resistor get that much warmer (even though it runs hot as is, probably why they went with 2 10 in series, to reduce the power dissipation on the resistors).
But we will never know exactly what happened... too bad!
BTW.... are you SURE that ONLY two 40watt incandescent lamps were on the dimmer??? no CFL's or ceiling fan?
Those are reactive loads and COULD create higher voltage spikes on the load side of the triac that COULD damage things.
I have many old style wall switches for years, and have (gasp) run transformers and tried cfl's on them, and have never had this problem.
The ONLY time I cooked an old style appliance module is when I had a floating neutral, and the line voltage went WAY over 120.... popped the varistor in the module due to overvoltage.... THAT was EXCITING and spectacular!!!
-
I would also say. Most manufacturers should use a X2 rated cap made for across AC power line use. For their signal suckers.
-
True the voltage rating should be 2x higher for safety margin.
The .68uf is essentially across the line (through the transformer, 22 ohm, 18volt zener diode).
I just took a look at the big blue cap, and it doesn't have a voltage rating on it, but the part number is 2E684... wonder if the 2 is 200 volts and 684 is 680nfd (or .68uf) (or 68 with 4 zeros, picofarads)(or 680000pf)(or 680nf)... ok, I'll quit rofl rofl rofl
It would still be curiously nice to find out what the failure mode was..... another of life's mystery. ;D
REALLY glad that it didn't do more damage!!!!
-
I may not have been clear.
The capacitor with an X2 designation. Is designed specifically for power line filters and being across AC Power Lines.
I bought some when I was using a capacitor across the phases of the house. For a short time.
Also I bought a few surplus Home Settings ZWave modules when Intermatic dropped all their ZWave stuff. Since they are strictly RF. They had a X2 rated cap directly across the AC Power Input Prongs. Kept their noise inside the module but also was a great X10 and Insteon signal sucker.
-
Ok, thanks Brian for clearing that up!
-
Just to clear up a couple of qustions raised......
1: There are only two 40 watt bulbs on this switch.
2: Neither bulb was burned out, connected the wires and both came on as expected.
I know have soft-start decorator switch in there. i'll keep my fingers crossed!
Merle
-
I didn't doubt that you only have the incandescent load on the switch... just wanted to make sure that there wasn't anything else.
Still curious as to why the switch failed????
-
One more thought.
Some homes have outlets controlled by wall switches.
Your WS467 isn't controlling an outlet where someone can accidentally plug in something like a vacuum cleaner or fan?
-
When you say power supply reactor cap, do you mean the 220uf electrolytic?
Nope.
I am talking about the big mylar. The caps reactance at 60Hz is used to drop the voltage/current.
-
That is what I thought... they are used to drop MOST of the line voltage, and the zener diode does the regulation. The 22 ohm is also there to limit current.
-
My soft start LM465's don't have the zener diode anymore. Made the series cap much smaller in value and raised the series resistor.
The new processor uses around 5 volts logic and as the extra current for driving the triacs gate turns on. The voltage goes down. ???
-
My soft start LM465's don't have the zener diode anymore. Made the series cap much smaller in value and raised the series resistor.
The new processor uses around 5 volts logic and as the extra current for driving the triacs gate turns on. The voltage goes down. ???
The softstarts that I have use 2, 10 ohm 1/2watt resistors in series. Are the resistors in yours 10 ohm?
-
Mine have a 330 Ohm 1/2 watt and a .22uf series cap.
Sounds like another revision. :o
-
My soft start LM465's don't have the zener diode anymore. Made the series cap much smaller in value and raised the series resistor.
Interesting. You can design capacitive or resistive transformerless power supplies but I haven't seen one combining both approaches. Microchip AN954 covers both approaches.
- http://ww1.microchip.com/downloads/en/AppNotes/00954A.pdf
-
Mine have a 330 Ohm 1/2 watt and a .22uf series cap.
Sounds like another revision. :o
330 ohm?????? Is that resistor near the edge of the board? Sure sounds like, yet, another version of the SAME model number.... OH BOY here we go AGAIN!!!
-
Mine is Date Code 09J44 with the Sonix SN8P2501B Surface Mounted Controller.
One of the production runs that had reported problems of going On to 50%
I had one where the triac only conducted on the negative half of the AC waveform
Was a flaky triac as I replaced it with one from another LM465 that I had used for a SSR experimental output.
I will look at the PC Board and report findings.
-
Neutral Pin---Fuse Wire--C3 .22uf--R1 330 Ohms 1/2W--Signal transformer-- Two 1N4004 Diodes. One to a 100uf/10v cap the other to Line pin to shunt the other half of the AC.
Also mine has the inductor coil for the outlet on the output side of the triac to line out. Not the neutral input pin to the neutral output pin like the older ones.
Silk Screen on PCB 11314A
Later tonight I will double check my notes to verify a brain malfunction has not happened. ;D
-
Neutral Pin---Fuse Wire--C3 .22uf--R1 330 Ohms 1/2W--Signal transformer-- Two 1N4004 Diodes. One to a 100uf/10v cap the other to Line pin to shunt the other half of the AC.
Also mine has the inductor coil for the outlet on the output side of the triac to line out. Not the neutral input pin to the neutral output pin like the older ones.
Silk Screen on PCB 11314A
Thanks for tracing that out! YES it IS another version! The SS ones that I have do have the choke on the triac output (hot) and not the neutral, so yours IS a later version..... hhmmmmm
Maybe they are using the two 1n4004's to create a (approx.) 3.3 volts for the processor, or are just relying on the cap/resistor/diode divider for the uP supply???
-
The coil on the neutral was in the old before soft start with the DIP IC controllers.