The situation is maybe slightly different in Europe:
Old Pearl (acid etched) 0.5% absorption
New Pearl (powder coat) 2% absorption
Satin / Soft White / Opal 7% absorption
The Satin types give a much better diffusion and a softer light due to the
use of a thicker coating. However these types are almost exclusively
offered with Krypton gas instead of the usual argon. This boosts the
filament efficacy by 6-7% and compensates for the coating loss, so in
practice their lumen output is the same as with clear bulbs. The light is
also whiter, owing to the higher filament colour temperature that Kr
permits. The Krypton lamps are usually re-shaped from the traditional
A-bulb (pear shape) to the K-bulb (mushroom shape). The latter have a
smaller volume and contain much less of the expensive Kr gas filling, to
keep the costs down.
Incidentally for HPS lamps a thick coating is needed (about 5% absorption)
only to achieve the same diffusion effect as the fluorescent powder coating
on a mercury lamp. Then the two lamps will then give the same light
distribution pattern from the fixture.
Terry - you made a great comment about bouncing bulbs off a concrete floor.
I recently saw a similarly impressive demo at a ribbon bulb plant, where
they dropped a small piece of broken glass into the bulb and gently rattled
it around. With little effort it causes the bulb to shatter, nicely
demonstrating the stress distribution in the glass. The external surface is
thermally stressed to put the glass into compression, which makes it very
strong. However the internal surface goes into tension, when glass is at
its weakest. It was surprising to see how little shock on the inside will
break the glass.
James.
TKM said:
Ioannis said:
[snip]
This question brings back many memories. Historically, there was much
debate about light losses due to lamp coatings and the lamp engineers
struggled to gain diffusion with losses of only fractions of a per cent.
What I remember is:
- Inside frost (which is an acid etch process) is the most efficient --
no
light loss compared to a clear finish.
- Powder coatings (some manufacturers refer to these as "soft white") --
about 1% loss
- Paint or enamel coatings -- about 4% loss typically
[snip]
If I also remember well, the inside frost finish is, as you say, an acid
etching process. I think it's HF. But I've heard it can also be sand
blasting.
Seems to me both acid and powder coatings hope to simulate a Lambertian
surface and thus reduce brilliance. I think the powder coatings are a
little
closer to this, since with the inside frosting the visibility of the
glass
cover is higher.
In any case, both my PHILIPS and Sylvania catalogs, have separate lumen
output
tables for coated and clear HID's, but not on inside frosted lamps.
Taking the 400W SON and SON-T high pressure sodium lamps, I see:
SON-T (clear) 38000 lm
SON (coated) 40000 lm
This is 2000 lumens lower in 40000. A loss close to 5%. So your figures
look
about right.
I took the HPS example because with HIDs which contain mercury the
diffusion
coating sometimes is fluorescent, so it might add lumens to the final
output.
Right. The acid is HF. The environmental problems involved in disposing
of used HF have hastened the end of inside frost lamp manufacturing. Sand
blasting works just fine too; but it weakens the glass. The HF frosting
process involves a series of acid baths proceeding from strong to weak
acid content. The idea is to etch the inside of the bulb; but then "round
off the edges" of the etched pits with weaker acid concentrations. That
helps strengthen the glass surface too.
One of the astonishing demonstrations in a bulb-blowing plant is to bounce
a glass lamp bulb off of a concrete floor. Take a clear bulb (no insides
or base) and drop it on its rounded end about a meter from the floor. It
will indeed bounce like a rubber ball.
Terry McGowan
Terry McGowan