Using Dimmer as Switch

[io_guy]

Is anyone here using Insteon Dimmers as Switches?

I currently use a combination of both (as expected) with switches connected to my non-dimmable CFLs.

I'm thinking long term, every light in my house will be dimmable. So why not use dimmers on the current CFLs, just setting ramp to 0.1s and output to 100%?

 

Anyone try this with success? I can't see why the full output dimmer would cause an issue.

[LeeG]

There is a recent post on the Smarthome forum where a user had LEDs connected to a dimmer. They were understood to be non-dimming LEDs but at 100% the user thought it would be okay. Burned up several LEDs. I've not put a dimmer output on a scope but I believe the sine wave is being chopped even at 100%. Only comment I can remember seeing about dimmers causing damage at 100%. There is always the possibility that someone will dim by accident or visitors not aware of the need to keep at 100%.

 

http://www.smarthome.com/forum/topic.asp?TOPIC_ID=10881

[IndyMike]

IO_Guy,

 

LeeG is exactly correct - Insteon dimmers do not provide a full sine wave at 100%. There is roughly 20% of the leading waveform chopped. You may get lucky and find non-dimmable CFL's that can handle the chopped waveform (I've tried some), but you're risking your dimmer and the CFL.

 

In addition to the chopped waveform, here's a couple of additional issues to consider:

 

• Ramp time: Insteon devices list the min ram time as 0.1 second. From a brightness standpoint, this may be correct. From a waveform standpoint (time for the chopped waveform to stabilize), the time is closer to 0.5 seconds.
  
• Noise: All dimmers produce impulse noise due to the fast rise time of the chopped output. CFL's present a capacitive load which dramatically increases powerline noise.
  

 

The following plots illustrate the "chopped" nature of the dimmer output and the noise induced by incandescent and CFL loads. Note that using a relay switch eliminates the impulse noise.

 

Chopped Waveform at 100% Brightness (Incandescent Load)

 

Powerline Noise Induced by 40W Incandescent at 100%

Red trace: Dual Band LL Dimmer Output @100%

Yellow trace: Powerline induced impulse noise (960 mv p-p)

 

 

Powerline Noise Induced by 12W Dimmable CFL at 100%

Red trace: Dual Band LL Dimmer Output @100%

Yellow trace: Powerline induced impulse noise (11.6 V p-p)

 

[io_guy]

Lee / IM,

Thanks for the info. I'm a little surprised at the waveforms coming out at 100%. I guess I'm stuck still buying switches.

I'll have to pull out the scope and tear apart a couple old dimmers - curious to see what guts they're using now.

Interesting how it clips/stalls the first 50% after zero crossing.

[apostolakisl]

I have a set of 4 non-dimming cfl's on an Insteon dimmer switch. We are now at 4 years on those guys without incident. These are kitchen lights so they are on/off all the time. I was aware of the reasons not to do this when I did it, but I did it anyway feeling fairly comfortable that in reality it really wouldn't make a difference. And it hasn't. But this is just one brand of bulb and one switch. I have no others.

[Brian H]

Depends on the CFL model.

Some like you have no problems and others have very short lived CFLs or smoke.

I am testing some dimmable LED bulbs and preliminary tests look good.

[ELA]

Here is another scope trace to add to IndyMikes pictoral of what happens when a CFL is dimmed. I tested several last year and as Brian said it varies from one CFL to another. I post this as it also shows the current waveform ( green) so you can see it is the current spikes ( or fast rate of change in current) that cause the "noise" in the voltage waveform.

 

 

This waveform also demonstrates how CFLs may not represent much of a load (energy wise) when first switched on at 100% ( other than to charge the input cap). Note how the wider duration current demand does not occur until a few milliseconds after initial turn on ( lamp firing).

 

Another possible reason for issues when running non dimmable CFLs at 100% on a dimmer is a lack of "holding current" at initial turn on. All triacs require a minimum current to remain on. With a single small wattage CFL(not designed for dimming) the triac may not "see" enough of a load at initial turn on and may turn on/off repeatedly resulting in damage. So in that scenario having several CFLs as a load may actually be better than just one small CFL.

 

Here is another trace of the same CFL but at 75% on:

 

Here you see both the large capacitive spike in current immediately followed by longer duration current required to fire the tube.

[Brian H]

Thanks to both IndyMike and ELA for the test results.

Very informative and explains why my particular CFL got real hot and flaky.