GE's new "High Efficiency Incandescent"...HIR or??

Main Page | Report this Page

Science Forum Index » Engineering - Lighting Forum » GE's new "High Efficiency Incandescent"...HIR or??

Page 1 of 2 Goto page 1, 2 Next

Author

Message

Daniel J. Stern

Posted: Tue Feb 27, 2007 11:54 pm

Guest

So, just in time for legislators' incandescent bans & CFL mandates, GE
announce they're devising a "high efficiency incandescent" lamp with
efficacy of up to 30 lpw:

http://tinyurl.com/2yzmhc

The press release claims the new lamps will incorporate new materials
presently under development. H'mmm. Anyone have any ideas (or
knowledge) of what they plan to do that'll double the efficacy of the
common incandescent lamp? The first thought that comes to mind is some
development of HIR. Just using some common volume-production
automotive lamps I'm familiar with, I come up with:

Tungsten: 12 lpw
Halogen: 18 lpw
HIR: 34 lpw

-DS

Don Klipstein

Posted: Tue Feb 27, 2007 11:54 pm

Guest

In article <1172634865.678850.84500@a75g2000cwd.googlegroups.com>, Daniel
J. Stern wrote:

Quote:

These do vary with voltage, wattage and design life expectancy.

For example, the 1156 produces close enough to 400 lumens (32 spherical
candlepower) from 26.9 watts (12.8 volts * 2.1 amps). That's about 14.9
lpw. A 120V 100A19 designed to last 750 hours gets 17.1 lpw. The 12V
100A21 is rated to achieve 20.5 lpw and is designed to last 1,000 hours.
A fair number of halogen lamps achieve 20 lpw, and I know of a 50 watt
12V one rated 19 lpw and designed to last 2000 hours (Philips
50W/12V/Capsule). I seem to think that a better brand 9005 achieves at
least 22 lpw at 12.8 volts, though I have not yet dug deep enough to back
that one up.

34 lpw sounds a bit on the high side for HIR to me, and that only
improves over some favorable non-HIR lamps by a little noticeably short of
doubling. I somewhat recall figures tossed into this thread before for
HIR were trending to the upper 20's lpw. I do somewhat recall a 350W 120V
T3 HIR halogen lamp being a replacement for the 500W one and with very
slightly less light output as I recall - that means a roughly 40% gain on
the efficacy of the 500W one that may achieve 21 lpw. I seem to think
close to 10,000 lumens from 350 watts - roughly 29 lpw.

OK, now I cite what appears to me to be a GE catalog page:

(My usual newsreader forces me to split into a few lines)

http://genet.gelighting.com/LightProducts/
Dispatcher?REQUEST=RESULTPAGE&CHANNEL=Consumer&FILTER=FT0010:
Specialty_Residential_Torchiere&CATEGORY=Lamps&BREADCRUMP=
Specialty_Residential_Torchiere%230

This is for torchiere halogen lamps/"bulbs".

The 225 watt HIR one achieves 5950 lumens, same as the non-HIR 300 watt
one. Lpw are 26.4 and 19.8 respectively, for 33.3% efficacy gain.

The 350 watt HIR one achieves 10,000 lumens, or 28.6 lpw. That is a 36%
gain over the 21 lpw of a 500W one achieving 10,500 lumens.

If HIR technology advances further or has recently advanced further, I
would surely like to know! And I think a major application would be
improving the efficacy of those popular halogen torchieres!

- Don Klipstein (don@misty.com)

Victor Roberts

Posted: Wed Feb 28, 2007 10:53 am

Guest

On 27 Feb 2007 19:54:25 -0800, "Daniel J. Stern"
<dastern@engin.umich.edu> wrote:

Quote:

Dan - see the thread New GE Incandescent Lamp Technology for
a more complete discussion.

They may be using low voltage HIR, which we believe can
already achieve 30 lm/W and is the technology that Osram
Sylvania is using in the e-PRO reflector lamps - OR, they
could have developed a selective emitter that can be used on
or in place of tungsten.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

Jim Hawkins

Posted: Wed Feb 28, 2007 1:43 pm

Guest

"Daniel J. Stern" <dastern@engin.umich.edu> wrote in message
news:1172634865.678850.84500@a75g2000cwd.googlegroups.com...

Quote:

What does HIR stand for ?

Jim Hawkins

Daniel J. Stern

Posted: Wed Feb 28, 2007 3:09 pm

Guest

On Feb 28, 12:43 pm, "Jim Hawkins" wrote:

Quote:

What does HIR stand for ?

Halogen Infrared Reflective. A dichroic multilayer coating (see
http://en.wikipedia.org/wiki/Dichroism ) is applied to the lamp
envelope such that visible light is passed and infrared (heat) is
reflected. The envelope shape and filament placement within the
envelope are (theoretically) carefully coördinated such that the
reflected IR is focused on the filament, increasing its temperature
beyond that reached by dint of the electrical characteristics of the
filament, thus increasing filament luminance without increasing power
consumption.

Daniel J. Stern

Posted: Wed Feb 28, 2007 3:42 pm

Guest

On Feb 28, 12:19Â am, Don Klipstein wrote:

Quote:

Just using some common volume-production
automotive lamps I'm familiar with, I come up with:

Tungsten: 12 lpw
Halogen: 18 lpw
HIR: 34 lpw

Â These do vary with voltage, wattage and design life expectancy.

Well, sure. And yes, I misremembered the flux value for the HIR lamp I
used for the above figures. Here are the two HIR headlight lamps:

9011/HIR1: 2300 lm Â± 15% @ 70w max at 12.8v (65w nominal)
9012/HIR2: 1700 lm Â± 15% @ 60w max at 12.8v (55w nominal)

So, fixing the wattage at the mean of nominal and max, we get efficacy
ranges for these two lamps as follows:

9011/HIR1: 29 to 39 lpw (optimised for flux)
9012/HIR2: 25 to 34 lpw (optimised for life)
Average of above ranges: 27 to 36.5 lpw
Mean of above average range: 31.75 lpw

The directly analogous headlight lamps without HIR are:

9005/HB3: 1700 lm Â± 12% @ 70w max at 12.8v (65w nominal)
9006/HB4: 1000 lm Â± 15% @ 60w max at 12.8v (55w nominal)

Same conditions and calculations as for the HIR1/HIR2 gives us:

9005/HB3: 22 to 28 lpw (optimised for flux)
9006/HB4: 15 to 20 lpw (optimised for life)
Average of above ranges: 18.5 to 24 lpw
Mean of above average range: 21.25 lpw

So, the HIR technology in this case gives an efficacy advantage of
around 49%.

Quote:

For example, the 1156 produces close enough to 400 lumens (32 spherical
candlepower) from 26.9 watts (12.8 volts * 2.1 amps). Â That's about 14.9
lpw.

I'd rather keep the comparisons constant in as many ways as possible,
so let's use the 12v version of the European R2 tungsten nonhalogen
headlight lamp for this part of the comparison:

High beam: â‰¥860 lm* @ 57w max at 13.2v (45w nominal @ 12.0v**)
Low beam: 675 lm Â± 15% @ 51w max at 13.2v (40w nominal @ 12.0v**)

* No tolerance range is specified for high beam flux, only a minimum
allowable value. A tolerance of Â±15% is more or less standard in this
sort of application, so if the minimum allowable value is 860 lm, we
can posit a maximum value of 1118 lm and derive a nominal value of 989
lm Â±15%.

** A calculation difficulty arises from the use of different voltages
for nominal and actual values. Here for the sake of consistency I'll
plug in half the 5v span between same-voltage nominal and max values
as in the calculations above, realising this introduces the
possibility for a small amount of error.

So:
High beam: 16 to 20.5 lpw (optimised for flux)
Low beam: 12 to 16 lpw (optimised for life)
Average of above ranges: 14 to 18.25 lpw
Mean of above average range: 16.125 lpw

Halogen advantage vs: plain tungsten: 32%
HIR advantage vs. halogen: 49%
HIR advantage vs. plain tungsten: 97% multiplicative, 81% additiveâ€”
citing both figures because press releases often incorporate claims
based on random maths!

Quote:

Â This is for torchiere halogen lamps/"bulbs".
Â The 225 watt HIR one achieves 5950 lumens, same as the non-HIR 300 watt
one. Â Lpw are 26.4 and 19.8 respectively, for 33.3% efficacy gain.

Â The 350 watt HIR one achieves 10,000 lumens, or 28.6 lpw. Â That is a 36%
gain over the 21 lpw of a 500W one achieving 10,500 lumens.

My understanding is that HIR technology works in these linear lamps,
but not as well as it can work where you can have a relatively small
filament at the centre of a spherical or quasispherical IR-reflective
envelope. Within the long tubular envelope, the long linear filament
of the torchiere lamp tends to sag between the supports out of the IR-
reflection focus, making it difficult to attain and sustain the
benefit. That said, my experience with the (hard to find!) GE HIR
linear halogen lamps has been quite favourable.

DS

Ioannis

Posted: Wed Feb 28, 2007 5:49 pm

Guest

"Daniel J. Stern" <dastern@engin.umich.edu> wrote in message
news:1172689752.436447.250890@k78g2000cwa.googlegroups.com...

Quote:

Isn't it the opposite?

I thought dichroic reflectors passed *heat* and reflected visible light
instead, so that heat could escape in all directions, whereas light would be
focused by the reflector.
--
I.N. Galidakis
http://ioannis.virtualcomposer2000.com/

Victor Roberts

Posted: Wed Feb 28, 2007 6:11 pm

Guest

On Wed, 28 Feb 2007 23:49:07 +0200, "Ioannis"
<morpheus@olympus.mons> wrote:

Quote:

"Daniel J. Stern" <dastern@engin.umich.edu> wrote in message
news:1172689752.436447.250890@k78g2000cwa.googlegroups.com...

On Feb 28, 12:43 pm, "Jim Hawkins" wrote:

What does HIR stand for ?

Halogen Infrared Reflective. A dichroic multilayer coating (see
http://en.wikipedia.org/wiki/Dichroism ) is applied to the lamp
envelope such that visible light is passed and infrared (heat) is
reflected.

Isn't it the opposite?

I thought dichroic reflectors passed *heat* and reflected visible light
instead, so that heat could escape in all directions, whereas light would be
focused by the reflector.

They can be designed either way.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

Victor Roberts

Posted: Wed Feb 28, 2007 6:12 pm

Guest

On 28 Feb 2007 11:09:12 -0800, "Daniel J. Stern"
<dastern@engin.umich.edu> wrote:

Quote:

On Feb 28, 12:43 pm, "Jim Hawkins" wrote:

What does HIR stand for ?

Halogen Infrared Reflective. A dichroic multilayer coating (see
http://en.wikipedia.org/wiki/Dichroism ) is applied to the lamp
envelope such that visible light is passed and infrared (heat) is
reflected. The envelope shape and filament placement within the
envelope are (theoretically) carefully coördinated such that the
reflected IR is focused on the filament, increasing its temperature
beyond that reached by dint of the electrical characteristics of the
filament, thus increasing filament luminance without increasing power
consumption.

I would prefer to say that reflection of otherwise wasted IR
back on the filament allows it to operate at the SAME
temperature with LESS electrical power.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

Daniel J. Stern

Posted: Thu Mar 01, 2007 12:25 pm

Guest

On Feb 28, 5:11 pm, Victor Roberts wrote:

Quote:

On Wed, 28 Feb 2007 23:49:07 +0200, "Ioannis" wrote:

Halogen Infrared Reflective. A dichroic multilayer coating (see
http://en.wikipedia.org/wiki/Dichroism) is applied to the lamp
envelope such that visible light is passed and infrared (heat) is
reflected.

Isn't it the opposite?
I thought dichroic reflectors passed *heat* and reflected visible light
instead, so that heat could escape in all directions, whereas light would be
focused by the reflector.

They can be designed either way.

Actually, dichroic reflectors can be designed with numerous different
pass/reflect characteristics, that's why so many different colours of
dichroic filter are available. Usually when the dichroism is between
two ranges of visible light, you have complementary pairs (pass
yellow, reflect blue; pass red, reflect green; pass amber, reflect
turqoise, etc.) but I am not certain if this is necessarily the case.
The HIR lamps' dichroic reflectors operate as described above: Pass
visible, reflect IR.

DS

Daniel J. Stern

Posted: Thu Mar 01, 2007 12:28 pm

Guest

On Feb 28, 5:12 pm, Victor Roberts wrote:

Quote:

On 28 Feb 2007 11:09:12 -0800, "Daniel J. Stern" wrote:

Halogen Infrared Reflective. A dichroic multilayer coating (see
http://en.wikipedia.org/wiki/Dichroism) is applied to the lamp
envelope such that visible light is passed and infrared (heat) is
reflected. The envelope shape and filament placement within the
envelope are (theoretically) carefully coördinated such that the
reflected IR is focused on the filament, increasing its temperature
beyond that reached by dint of the electrical characteristics of the
filament, thus increasing filament luminance without increasing power
consumption.

I would prefer to say that reflection of otherwise wasted IR
back on the filament allows it to operate at the SAME
temperature with LESS electrical power.

Erm...why would you prefer that? It's no more or less correct. The
lamps can be designed either way, just as the efficacy advantage of
halogen over vacuum or inert-gas fill can be used to provide greater
flux from the same power or equal flux from lower power. If you'll
read back up the thread, you'll find real-world examples of both
design optimisations: The HIR1/9011 and HIR2/9012 automotive headlight
lamps use the same power and produce more light than the analogous non-
HIR HB3/9005 and HB4/9006, while the GE linear HIRs use less power to
give equal light compared to the analogous non-HIR halogen linear
lamps.

DS

Victor Roberts

Posted: Thu Mar 01, 2007 4:04 pm

Guest

On 1 Mar 2007 08:28:49 -0800, "Daniel J. Stern"
<dastern@engin.umich.edu> wrote:

Quote:

On Feb 28, 5:12 pm, Victor Roberts wrote:
On 28 Feb 2007 11:09:12 -0800, "Daniel J. Stern" wrote:

Halogen Infrared Reflective. A dichroic multilayer coating (see
http://en.wikipedia.org/wiki/Dichroism) is applied to the lamp
envelope such that visible light is passed and infrared (heat) is
reflected. The envelope shape and filament placement within the
envelope are (theoretically) carefully coördinated such that the
reflected IR is focused on the filament, increasing its temperature
beyond that reached by dint of the electrical characteristics of the
filament, thus increasing filament luminance without increasing power
consumption.

I would prefer to say that reflection of otherwise wasted IR
back on the filament allows it to operate at the SAME
temperature with LESS electrical power.

Erm...why would you prefer that? It's no more or less correct. The
lamps can be designed either way, just as the efficacy advantage of
halogen over vacuum or inert-gas fill can be used to provide greater
flux from the same power or equal flux from lower power. If you'll
read back up the thread, you'll find real-world examples of both
design optimisations: The HIR1/9011 and HIR2/9012 automotive headlight
lamps use the same power and produce more light than the analogous non-
HIR HB3/9005 and HB4/9006, while the GE linear HIRs use less power to
give equal light compared to the analogous non-HIR halogen linear
lamps.

I agree you can do either. I made that comment because that
is the way that IR reflecting films have typically been used
in general lighting to make improved incandescent lamps. If
they were used to increase the filament temperature then
life would be greatly reduced, which most consumers would
not find to be desirable.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

Daniel J. Stern

Posted: Fri Mar 02, 2007 9:17 pm

Guest

On Mar 1, 3:04 pm, Victor Roberts wrote:

Quote:

Halogen Infrared Reflective.reflected IR is focused on the filament, increasing its
temperature beyond that reached by dint of the electrical characteristics of the
filament, thus increasing filament luminance without increasing power
consumption.

I would prefer to say that reflection of otherwise wasted IR
back on the filament allows it to operate at the SAME
temperature with LESS electrical power.

The lamps can be designed either way

I agree you can do either. I made that comment because that
is the way that IR reflecting films have typically been used
in general lighting to make improved incandescent lamps. If
they were used to increase the filament temperature then
life would be greatly reduced, which most consumers would
not find to be desirable.

H'mm...not sure I agree this is necessarily the case.

G-E claim a B50 of 850h @ 14.0v for their HB4/9006 (1000 lm).
G-E claim a B50 of 900h @ 14.0v for their HIR2/9012 (1700 lm - this
lamp is NLA)
Toshiba claim a B50 of 800h @ 14.0v for their HIR2/9012 (1700 lm -
current production)

Other than the HIR aspects (envelope shape & coating, lumens) these
bulbs are essentially identical. Same filament geometry, same power
consumption, same architecture.

....?

DS

Victor Roberts

Posted: Sat Mar 03, 2007 10:00 pm

Guest

On 2 Mar 2007 17:17:22 -0800, "Daniel J. Stern"
<dastern@engin.umich.edu> wrote:

Quote:

On Mar 1, 3:04 pm, Victor Roberts wrote:

Halogen Infrared Reflective.reflected IR is focused on the filament, increasing its
temperature beyond that reached by dint of the electrical characteristics of the
filament, thus increasing filament luminance without increasing power
consumption.

I would prefer to say that reflection of otherwise wasted IR
back on the filament allows it to operate at the SAME
temperature with LESS electrical power.

The lamps can be designed either way

I agree you can do either. I made that comment because that
is the way that IR reflecting films have typically been used
in general lighting to make improved incandescent lamps. If
they were used to increase the filament temperature then
life would be greatly reduced, which most consumers would
not find to be desirable.

H'mm...not sure I agree this is necessarily the case.

G-E claim a B50 of 850h @ 14.0v for their HB4/9006 (1000 lm).
G-E claim a B50 of 900h @ 14.0v for their HIR2/9012 (1700 lm - this
lamp is NLA)
Toshiba claim a B50 of 800h @ 14.0v for their HIR2/9012 (1700 lm -
current production)

Other than the HIR aspects (envelope shape & coating, lumens) these
bulbs are essentially identical. Same filament geometry, same power
consumption, same architecture.

...?

DS

But not the same filament.

You do raise some interesting questions and perhaps my
answer was too simplistic.

We all know that other things being equal, if we increase
the temperature of a given tungsten filament it will
generate more light and also have a higher evaporation rate
and hence shorter life.

So, if we start with a 90-watt halogen filament tube and add
an IR reflecting coating we need to lower the input power,
for example, by lowering the voltage, in order to keep the
tungsten from overheating. This results in lower power
consumption for the same amount of light.

In the real world we want the HIR lamp to operate at the
normal input voltage so we modify the filament, usually by
making it longer, so it has higher resistance and draws less
power at the standard power line voltage.

One example is a 60-watt, 120-volt HIR filament tube that
produces about the same amount of light as a 90-watt,
120-volt halogen filament tube of similar design.

Now, let's look at the lamps you present, since different
design rules were used.

If we start with the 55-watt HB4/9006 and add ONLY an
IR-reflecting coating the filament temperature will go up
and we will get more light, but the life will go down,
perhaps drastically, since we could heat the tungsten to the
melting point.

However, if the HIR2/9012 design starts with thicker and
hence lower resistance tungsten, for example the wire size
that might be used for a 100-watt lamp if the filament were
the same length as the HB4/9006, and then coils the filament
so the length of the tungsten is longer and the net
resistance is the same as the 55-watt version, this lamp
would also draw 55 watt at 14 volts but would have a cooler
filament than the HB4/9006. If we then add the IR
reflecting coating, the reflected IR will raise the filament
temperature. Since the tungsten wire is thicker it can now
be operated at a higher temperature for the same or slightly
longer life, which will produce more light. The greater
surface area, due to both the larger diameter and longer
length, also contributes to the higher light output.

Of course, since tungsten resistance is a strong function of
temperature, the design is not done serially as I have
described but in an integrated manner. In the end the
HIR2/9012 has a longer length of thicker wire that has the
same hot resistance as the HB4/9006 but can be run at higher
temperature due to its greater thickness. This filament,
the one in the HIR2/9012, produces more light than the
filament in the HB4/9006 due to both its higher temperature
and its larger surface area.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

Daniel J. Stern

Posted: Mon Mar 05, 2007 7:13 pm

Guest

On Mar 3, 9:00 pm, Victor Roberts wrote:

Quote:

G-E claim a B50 of 850h @ 14.0v for their HB4/9006 (1000 lm).
G-E claim a B50 of 900h @ 14.0v for their HIR2/9012 (1700 lm - this
lamp is NLA)
Toshiba claim a B50 of 800h @ 14.0v for their HIR2/9012 (1700 lm -
current production)

Other than the HIR aspects (envelope shape & coating, lumens) these
bulbs are essentially identical. Same filament geometry, same power
consumption, same architecture.

But not the same filament.

Mmm...good point. Though the differences are probably small in
absolute scale, considering the specifications for the two lamps'
filament coils are quite close:

HIR2: 1.6mm max diameter, 5.3mm max length,
HB4: 1.6mm max diameter, 5.1mm max length.

Quote:

In the real world we want the HIR lamp to operate at the
normal input voltage so we modify the filament, usually by
making it longer, so it has higher resistance and draws less
power at the standard power line voltage.

Understood, but I have to wonder what degree of significant filament
alteration is possible within the specs given above. I suppose a
different gauge and pitch might be used.

Regards,

DS

Page 1 of 2 Goto page 1, 2 Next All times are GMT - 5 Hours The time now is Mon Oct 13, 2008 5:44 am Science Forum Index