solid state timer switch kills CFLs?

[vtpeaknik]

A query that may have implications for some home-power situations:

On several occasions I've left a running lamp on a timer while away from my on-the-grid home for  a few days.  A CFL lamp of course.  My old electro-mechanical timers are kind of cranky, so I was happy when I picked up at a yard sale a digital timer, with no moving parts, simple but with some nifty features such as on-for-x-hours-after-dusk (it has a light sensor too).  But, at least twice, when I came back home the CFL was dead.  So it seems that it kills CFLs.  But why?  I know standard CFLs are not to be used with dimmers, but this thing is either totally on or totally off, presumably using a triac but with an almost 100% duty cycle.  Ideas?

[sPuDd]

Don't use the light sensor option. It will "flicker" between on & off at dusk & dawn.

Thats what usually kills them. A commercial light sensor switch will come with the

warning "Not to be used for CFLs".

sPuDd..

[DamonHD]

Is the timer in-line with the lamp and drawing its own power through the lamp?  (And does it have a battery.)  If so it will have only two wires rather than live, neutral and lamp.

When I've done designs on that basis, for an incandescent load, I've snuck the drive current through a capacitor to get it even while the voltage crosses zero.

It is possible that a CFL may not like that reactive element in series with it.

Rgds

Damon

[dnix71]

If it was made for incadescent lights, it may also be a diode on purpose. The half wave is enough for light but greatly increases the life of the bulb.

My first guess would be the post above about flicker. I have been forced to modify porch lights with a dusk/dawn sensor because they would flicker for hours if facing the wrong way. When the light came on it would increase the ambient light enough to shut the light off again.

[Electron Skipper]

Considering the power for the timer itself is drawn in parallel as it is plugged into the wall, the flicker issue is the most realistic.  A simple screw in module that screws into the bulb socket would be the same.  If however, you have one that is hardwired in place of the wall switch, the unit draws low current through the light bulb itself even in the "off" setting.

I do happen to have a similar timer as the one described, and I have used it on conventional T-12 lighting with no issues at all.  Most likely both are using a Triac for the control device for simplicity and cheapness of circuit construction, I highly doubt any issues would arise from the timer aspect directly. or it's circuitry.  

Which leaves the matter of the flickering as the most likely culprit.  Without testing in actual conditions, conjecture is all that can be done.  Set it up while you are at home to watch how it behaves is the best suggestion if you are wedded to CFL's in that fixture, or that that timer for that fixture.

As an alternate program idea, if your timer has a "random" setting, use that instead.  It randomizes the actual turn on period using the time setting as an "approximate start point."

[Opera House]

Triac circuits tend to have high leakage currents so the CFL may never actually turn off and be operating in a low voltage brown out condition.  I have responded to this question many times on engineering forums.  One solution is to place a capacitive load across the CFL, a 220 ohm resistor in series with a .47uF capacitor often corrects the situation.  If you have a voltmeter, measure the voltage on the lamp when the timer is off. You need to be under 10V. A relay to control the load is guaranteed to save the CFL, but you may still need the capacitor to insure the relay turns off after being turned on.

[vtpeaknik]

Thanks for your reply.  Can you explain what the capacitor (plus resistor) does?

[Opera House]

Most triac circuits need a low resistance load to work because they always are sourcing some current. A high wattage resistor could provide this load.  The capacitor provides a load without generating any heat.  The resistor in series is more of a safety item to limit transient currents in capacitors that are not rated X2.  he leakage current of 5mA may be enough to operate the electronic ballast in a region where the fet gate never fully turns on and overheats or the filaments are on all the time and eventually burn out.  You can test for leakage by plaving only a 1000 to 470 ohm resistor as the load in the off state.  Measure the voltage and determine the leakage current by E/R.

[slickone987]

Triacs will kill the ballast circuits in the CFL bulbs. Use a timer with solid-state relay switching such as the Swylite brand. These timers are designed for and tested to work with CFL & LED bulbs down to a 1 watt total power draw. Most other timers need at least a 40 watt power draw to work, which means you need to have 120W or above in CFL's to make them work.

I have been using them for years on my CFL's (inside 12W and out 7W) and haven't had any problems whatsoever.