Dimmer 1 unresponsive after blown bulb

[cjard 1]

I've 8 halogens in one room, running on a dimmer 1. After contact with a high speed football (oops) the RCD for the circuit* tripped. After resetting the RCD the lights came back on, apart from the bulb that was struck.. I've since noted though, that they won't turn off! Switch continuity is good. Using a jumper wire to briefly connect Sx and S1 had no effect. Using the web app/mobile app to control the dimmer has no effect; the settings look like they work in that it accepts a change to the slider for brightness, and the dimmer gives readings/can be queried in the admin pages and responds to commands to alter parameters, but never varies its output to the bulbs - it's locked to "always on max brightness"

 

is the dimmer ruined?

Edited July 23, 2017 by cjard

[petergebruers 1,265]

The electronic switch used in "Dimmer 1" is incapable of interrupting short circuit currents. Once the switch is "on", it stays "on". Although it is fairly robust (for an electronic switch) a short circuit can cause it to get destroyed. In my experience and confirmed by numerous reports, this component has a very high (99 %) probability of failing as a dead short, because the very high current will locally melt the semiconductor structure, giving you a conductive block of molten, doped silicon. So after a gross overload, it is quite possible that the Z-Wave part of the dimmer still functions, but the load will be "always on".

 

There's lots of technical stuff behind that simple event (football -> broken lamp -> shorted dimmer) and as you've probably guessed, it fascinates me. Why would a broken lamp cause a very high over-current? As the glass breaks, the low pressure gas in the lamp gets replaced by air. This burns the filament, which makes it thinner and hotter. In the blink of an eye, the filament will break and then something interesting happens: the hot metal forms a conductive gas, the gas goes everywhere and causes almost dead short. This conductive plasma thing also happens when the filament breaks when the bulb is at the end of its life. You can see this has happened: the glass envelope will be greyish, and when you look closely you'll see a metal shine, a mirror like surface. A beautiful failure mode, if you ask me, that makes all possible failure modes of LED lamps rather boring. Except maybe, an exploding electrolytic capacitor, but that will be contained in the base of the lamp or the power supply...

 

I'd expect your circuit breaker to interrupt the fault current, not your RCD. Two possible explanations: either your RCD is actually an RCBO or when the glass ruptured, the hot plasma touched the grounded metal parts of the lamp holder, triggering an earth fault.

 

That is to say, "it's dead, Jim".

 

Would "Dimmer 2" have survived? I'm not sure. Possibly. At least, chances are better (than Dimmer 1). I know for certain that Dimmer 2 will not get damaged, if you turn it on with a very high current load (very crude measurement: it turns off in microseconds if I > 40 A). But turning on a high current load is not the same as "interrupting a current that occurs when the switch is already on". I do not know if it is safe to short circuit the output of Dimmer 2, while it is on and I am not prepared to test that

 

 

 

[cjard 1]

Peter, I'm glad your brain visits the internet; it holds a fascinating cornucopia of electrickery!

 

the bulb in question is a gu10, and is intact physically but I'll admit I haven't inspected the inner halogen unit to see if it's burnt silver. I suspect you're right, as you've also correctly surmised that all my circuits are RCBOs (I put an asterisk in my original post as I was going to mention this but forgot)

 

oh well, if the defective component can't be swapped out I'll have to bin that one. I got a load of them from Poland off eBay years ago for a tenner each.. of course, now the only option is to replace with a dimmer 2 and actually I don't think it can handle the load from 8 x 50w halogens, so I might have to go led in the hall.. shame cos I really like the way halogens dim to red, the only led that replicate this are hideously expensive and the hall lights are only used occasionally so don't really warrant led/will never repay investment..

Edited July 24, 2017 by cjard

[petergebruers 1,265]

@cjard thank you. I never opened a FGD-211 but I dare to guess... the TRIAC is a surface mount type and if you've never tried to desolder a similar component, you might find it very difficult. There's also the possibility that the TRIAC destroyed the output of the microcontroller when it died.

 

Dimmer 2 is 250 W resistive indeed. It also uses a different bypass, you'll probably need one if you convert to LED.

 

I like halogen, I still have a few. The only lamp that gets close enough, for me, is the "Philips Warm Glow Dimmable LED".

 

It's expensive, but if you love halogen, I can really recommend you try one:

 

Please login or register to see this link.

Edited July 27, 2017 by petergebruers