Re: T8 benefits -- smaller can be better but more power might/might not

["S. Hieber" <shieber at yahoo_com>]

I'll add a couple of technical points to Wayne's excellent summary on
T8.

Point 1:  Smaller diameter bulbs can be (but not always are) more
efficient than similar length larger diameter bulbs.

Point 2:  Generally (i.e., oversimplified), increasing power to
fluorescent lamps reduces energy efficiency.

Re Point 1: Although a smaller diameter bulb has fewer electrons
inside, other things being equal, the gas stays warm more easily, which
helps the mercury vaporize and the bulb illuminate.  T5s can be made
more efficient than T8s.  Also the smaller diameter means multiple
bulbs tend less to block as much of each other's light.

But most T5s are either small very low wattage bulbs or high power,
high output bulbs like power compacts.  It seems that the low power
bulbs tend to be less efficient than their larger siblings -- if you
add to the equation any coiling of the tube, like with compact
fluorescent lamps, you get even less efficiency because the bulb is
blocking much of its own light.  


Re Point 2: Adding power to a bulb generally decreases efficiency as
light output increase.  As power is increased, energy losses increase
faster than light output increases.  High power bulbs, like VHOs and
PCs tend to be less *efficient* than more mildly powered bulbs because
a basic fact of life with fluorescents seems to be that, after a
certain point increases in power bring smaller and smaller increases in
light output.  Terribly oversimplified, as the current is increased,
which provides more electrons to knock more photons out of the
vaporized mercury inside the bulb, energy losses are also increased. 
(E.g., the mercury vapor pressure rises, which impedes the
electron-photon excitation process.)  Complicating all this, different
bulbs are designed for different operating (bulb) conditions.  

The original "trick" to making very high output fluorescent lamps was
to use a lighter gas (neon instead of argon) in the bulbs -- the
lighter gas meant fewer electrons were impeded when the mercury got
hotter at the higher levels of current.  Another change was to put in a
small plate to keep the mercury from overheating.   These design
changes drastically improved the ratio of increased light output to
increased energy losses.  The ration is still les than 1-to-1 but it
was much improved for the higher current levels.  But, by overdriving
fluorescents, you tend to defeat those and similar engineering feats. 
So you are back to increasing energy losses faster than you increase
output.

How a *given* bulb is driven, and whether it is at optimum energy
efficiency or maximum possible light output or maximum possible bulb
lifespan, depends on the ballast.  But there is a lot of leeway it
seems.  Many electronic ballasts compensate for different bulbs by
adjusting current and/or voltage somewhat.  (Some are constant current
inverters and some are constant voltage inverters.)

Suppose you find a ballast that gets you desired light level from a
desired bulb but it's less efficient than if you used a ballast that,
in combo with that bulb, gave less light.  Maybe so what.  Less
(energy) efficiency isn't necessarily bad.  What good is a more
efficient lamp set-up if you can't get enough light (lumens, PAR,
however you want to measure it) into your tank?  So overdriving is one
option, but twice the ballast circuits doesn't mean you get twice the
watts (Wayne has demonstrated this very well in his testing).  *And*
even if you double the watts, that won't translate into double the
light output.  Of course, if the bulb burns out in three weeks, like my
normal output bulbs did on my IceCap ballast, what good is efficiency
or light output effectiveness?  No bulb is cheap enough to be worth
replacing that often.

It would be nice to be able to predict energy consumption, light
output, and bulb life span for any given bulb and ballast.  Short of a
predictive model, it would be nice to have a lengthy list of combos and
the results.  So far all we have are manufacturer predictions, and most
of those do not include all 3 pieces of data just mentioned.
"Runs NO bulbs" might mean 50% brighter but 1/5 the lifespan and half
the efficiency compared to some other ballast.  

Scott H. 

__________________________________________________
Do You Yahoo!?
Sign up for SBC Yahoo! Dial - First Month Free
http://sbc.yahoo.com