Automated Dimming of Fluorescent Lights

[rbroders]

I understand that some fluorescents are dimmable, but you need a special ballast and a special dimmer (three-wire or four-wire). Are there any automated lighting systems which support such dimmers?

Thanks -- Bob

 

[hockeypuck]

I understand that some fluorescents are dimmable, but you need a special ballast and a special dimmer (three-wire or four-wire). Are there any automated lighting systems which support such dimmers?

Thanks -- Bob

Colorado VNet had a floro dimming module. I would image SQD's calipsal does as well.

 

[rfdesq]

I'm not sure what you mean by automated lighting systems, but UPB can control dimmable fluorescents.

 

[Mike]

I bought a few to test, some standard bulbs and some R30.

The standard bulbs were microbrite, the larger ones Neptun. I have Insteon dimmer switches throughout the house.

For the standard bulbs, they dim fine, but I did not pay attention to the initial lumens. A higher value would have been better as it starts out fairly dim and then gets brighter. Dimming function is nice and smooth, no flicker.

The others are being taken out. They flicker at low dim rates and are not as smooth on the dimming. The wife hated the test bulbs (the large ones at least). I had planned to hit the primary use areas and replace them but I am going to stick with normal bulbs for now. This is all personal preference though, you might want to order a few bulbs and try them.

I think the issue is that standard dimmers can be used, but the bulbs must be dimmable (not sure about the larger flourescents if options exist for those, which may have been your real question).

Check out this link (someone here recommended the link originally and is who I used for a trial):
http://www.1000bulbs.com/category.php?category=3876

 

[hult]

I understand that some fluorescents are dimmable, but you need a special ballast and a special dimmer (three-wire or four-wire). Are there any automated lighting systems which support such dimmers?

Thanks -- Bob

For the last 3 (4? 5?) decades, a standard control signal for commercial and architectural dimmable fluorescent ballasts has been 0-10vdc analog. Google 0-10vdc fluorescent

An architectual/entertainment lighting standard protocol that readily supports 0-10vdc analog is DMX-512 (aka DMX512a ) in part because DMX was a follow-on to when 0-10vdc was itself the standard for dimming. I have some references to DMX and its use in HA on my website in various places at www.ECOntrol.org

0-10vdc analog is a convenient method for HA because :

1) it is easy to trouble-shoot,
2) easy to design and construct DIY hardware including analog sensors
3) can be used for a variety of purposes (see eg www.ECOntrol.org/audio.htm ) and
4) prices for DMX have plummeted as DIY and Far East supplies and suppliers supplement US and European manufacturers.

For example, I use American DJ Multi-Data/Paks which provide 72 (seventy-two) channels of 0-10vdc analog output in a well-built 1U rack-mount enclosure based on DMX input. eBay for ~ $100. New street price is ~ $300; That's only ~1.40 to $4.00 per channel.

A single DMX universe can have 512 channels which can change (controlled) as rapidly as 44 times per second.

DMX can also be used for 0-10vdc *input* at 44 samples/sec of 512 channels, but this is far less frequently used.

... Marc

 

[rbroders]

Let me clarify my requirements a bit. I want a fully dimmable (ideally 1% - 100%) architetectural (tubes hidden in valences) fluorescent lighting system which can be controlled by my automation system (Elk) using standard switches.

Check out: http://www.lutron.com/product_technical/fluorescent.htm

They have a Tu-Wire ballast which I believe is compatible with the standard triac dimmers provided by most automation manufacturers. Has anyone used this product? Are you happy with it? Does it dim enough?

Otherwise, the standard is 4-wire, which as Marc described is two 120VAC lines for power, and two 0-10VDC lines for control. I am not aware of any standard commercial switches which provide this output though X10, ALC, iLine, Lutron?

There is also a 3-wire standard, one 120V Neutral, one 120V switched hot, and one 120V dimmed hot (again I assume the triac dimmers can supply this). Do you leave the 120V switched hot on all the time, or dedicate another switch to turn that leg off when the dimmer leg is off?

Now that codes require more and more fluorescent use I'm surprised more people aren't grappling with this issue. Also, do you really need to season the lamps at full power for 100hours???

Thanks -- Bob

P.S. Marc, I'm sure your system is really cool, but my wife won't let me go that far off the reservation if you know what I mean.

 

[greenup]

Let me clarify my requirements a bit. I want a fully dimmable (ideally 1% - 100%) architectural (tubes hidden in valences) fluorescent lighting system which can be controlled by my automation system (Elk) using standard switches.

Well, I'm not an Elk user, and I couldn't afford those spiffy 1% ballasts, but I did get some Sylvania Quicktronic Powersense ballasts (QTP3x32T8/UNV DIM-TCL) to work with my Insteon (Switchlinc V2) switches. these are 100%..5%, and, as I'm sure you have seen elsewhere, the light-power curve for florescent means that 5% power means more like 25% light, but I can cope; particularly because I think the wallet performance factor was huge. (something like $50 vs $200; *10 fixtures. that would hurt me, at least.)

Otherwise, the standard is 4-wire, which as Marc described is two 120VAC lines for power, and two 0-10VDC lines for control. I am not aware of any standard commercial switches which provide this output though X10, ALC, iLine, Lutron?

My original plan was to do some complicated fiddling with a dimmer, controlling a 10V transformer, also paired to an in-linelinc to switch the power. I was QUITE relieved that the switch seems to control the ballast directly with minimum fuss. Of course, the bottom dim levels on the switch do not result in light from the fixture, and it's probably murder on the ballast when that happens.
Also, I don't have any long term metrics on whether or not this will result in premature bulb or ballast failure. I did a lot of research before trying, and had found one message from an Insteon tech suggesting that the triac they used was configured to cut the voltage phase in the same way that "The Real" dimmable controls did.

I'm surprised more people aren't grappling with this issue.

Yeah, it surprised me too... Not because of code, but just because florescent provides MUCH more light, and there are more options for light rendering, spectrum, etc. Florescent WAS horrible in the past... poor CRI, bad ballasts that flickered 60Hz, ... but these days, even the good stuff is cheap.

My only relationship with Insteon or Sylvania is "Customer". Specific items mentioned only for reproducibility.

-greenup

 

[rbroders]

I understand that some fluorescents are dimmable, but you need a special ballast and a special dimmer (three-wire or four-wire). Are there any automated lighting systems which support such dimmers?

Thanks -- Bob

Colorado VNet had a floro dimming module. I would image SQD's calipsal does as well.

I found the Colorado VNet module: 8 0-12VDC outputs (configurable to 0-10VDC supposedly)

The SQD clipsal unit is 4 0-10VDC outputs, and they have a DALI gateway (DALI is a digital control interface used on some dimmable ballasts).

The good news is that neither of the 0-10VDC solutions appear to include switched 120V outputs, so I assume that dimming to 0V is sufficient and there isn't too much waste involved in leaving the ballast hot all the time (though you might want to throw the breaker before changing bulbs).

I wonder how hard it is to convert a standard 120VDC triac dimmer output to 10VDC? Do you need a transformer? How about a simple rectifier, capacitor and resistor ladder?

Does this also mean that the hot leg of the 3-wire ballasts can be left on all the time and our standard triac dimmers can be used to dim the 120V control leg? Lutron Hi-Lume calls it standard line voltage phase-control technology, but I'm not sure its compatible.

I also found some Crestron products:
The CNLFDB-4 is a four channel dimmer for Lutron Hi-Lume or ECO-10 ballasts. It seems to have three wire terminals, so perhaps it powers off the ballast when the light is off.

Also the CNLDAB-4A is a four channel dimmer for Advance Mk VII ballasts. It definitely has both the 0-10VDC output and a relay for switched 120VAC.

Hmmm, maybe its possible to use some Elk outputs to do this instead - arghhh.

Any experience with these products would be very helpful! Thanks -- Bob

P.S. It seems Advance also sells the Mark 10 ballast which uses only two wires and claims 100%-5% dimming performance. Not totally sure it will work with standard triac dimmers though...

 

[rbroders]

Okay, I think I found the solution: Leviton Power Extenders. The normal power extender, model PE100-10W is used to extend a low power dimmer (600W typ) to 1920W. The PE100-10W's output matches the dimmer controls output, and I believe it supports standard triac dimmers.

The cool thing is that Leviton also sells "Power Extenders" which support three and four wire dimmable ballasts! The PE200-10W controls three-wire (Leviton Hi-Lume) dimmable ballasts up to 1920VA, and the PE300-D0W controls four-wire (Advance Mark 7) dimmable ballasts up to 20A.

Of course these devices add $170+ to the cost.

I may purchase a set of these to test it out, and then I'll post an epitath to this thread -- Bob

 

[rbroders]

Okay, to fully test the dimming fluorescent problem, I upgraded my kitchen lights with HiLume ballasts. I used an Acenti dimmer which natively supports the HiLume but does not support automation (this old house has no automation). The ballasts work great. The tubes dim to a very low level, and seem quite reliable. There is a brief flash as they startup, but otherwise they start directly at the desired dim level. The Acenti dimmer can dim down to the point where one or both tubes in the fixture go out even though the system is still on (the tube ends are glowing, but there is no illumination). Unlike incandescant bulbs, the fluorescents seem to maintain their color temperature throughout the dim range.

I also purchased the Leviton PE200-10W Power Extender, and temporarily wired the system to a couple of standard dimmers (including an EDT i-ON-D). The Power Extender flawlessly drove the HiLume ballast with the standard dimmers. The PE module automatically turns the ballast on when the dimmer reaches a certain threshold (there is an audible click, and an LED indicator). There is also a minimum level adjustment which allows you to dial the flourescent tubes to their minimum level at the dimmer turn-on point, so you don't have to worry about the light being energized without illuminating.

The i-ON-D's "soft start" feature seems unnecessary with the fluorescents because of the initial flash. I do like the soft off though.

Overall, I believe dimmable fluorescents are viable, but even though they provide reduced illumination and you can use your kitchen at night without blinding yourself, they don't provide the mood you get from dimmed sconces or even halogen undercabinet lights. Point sources do a better job of providing the light and shadow interplay that you expect when entering a dimly lit room.

--Bob

P.S. The biggest improvement to this kitchen light was throwing out the crappy 4200K 60CRI tubes and replacing them with 3000K 85CRI tubes.

 

[JonR]

Thanks for the report on the Power Extender. I have a new home with several tube fixtures in a cove installation using Hi-Lume ballasts and all driven from one dimmer. I had done some research several months ago about how to control those lights, preferably with UPB, but came up empty. The rest of the building craziness has had me distracted for several months so I missed this thread but it looks like the PEs might be just what I need.

We ran the lights for 100 hours using an on/off switch before the dimmer was installed - it really wasn't that bad considering we're still under construction and needed the light anyway.

JonR