Discussion:
Can you increase fluorescent lamp efficiency by 10% by running on square wave?
(too old to reply)
ColdFusion
2012-05-04 03:20:10 UTC
Permalink
Raw Message
X-No-Archive: Yes

I know that mercury arc actually becomes about 10% more efficient when
it is operated on high frequency. I believe it has to do with the arc
not extinguishing at zero crossing as it does on sinusoidal operation
on line frequency.

Low frequency square wave is utilized for filming video to avoid
flickering as well by electronic HID ballast as HIDs can not tolerate
high frequency operation.

If fluorescent lamps were to be operated at 60 or 400Hz square wave,
would it provide 10% gain? How about lamp life? Since with square
wave, the peak and RMS are the same, I would think that it would
provide a LCCF of one which should be beneficial to lamp life.
Andrew Gabriel
2012-05-04 10:50:08 UTC
Permalink
Raw Message
Post by ColdFusion
X-No-Archive: Yes
I know that mercury arc actually becomes about 10% more efficient when
it is operated on high frequency. I believe it has to do with the arc
not extinguishing at zero crossing as it does on sinusoidal operation
on line frequency.
Yes - the electrons have not had time to decay back, so the gas
is still conducting and doesn't have to restrike. This happens
from around 5kHz upwards (with a sine wave), I believe.
Post by ColdFusion
Low frequency square wave is utilized for filming video to avoid
flickering as well by electronic HID ballast as HIDs can not tolerate
high frequency operation.
If fluorescent lamps were to be operated at 60 or 400Hz square wave,
would it provide 10% gain? How about lamp life? Since with square
wave, the peak and RMS are the same, I would think that it would
provide a LCCF of one which should be beneficial to lamp life.
Sounds likely to me.

A problem with square wave is RF noise generation, as the square
corners have large high frequency components. It's also more of
a challenge if transformers are involved, as there isn't one
choice of core materials which work well at both the low frequency
and high frequency components of a square wave.

However, given that the gain comes at about 5kHz sine wave, in
theory a square wave passed through a filter to cut off from 5kHz
up (in effect, rounding off the sharp corners), would still give
the benefits you suggest. Could even go higher than 5kHz, given
that electronic ballasts operate up to around 50kHz.
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
Don Klipstein
2012-05-05 04:40:03 UTC
Permalink
Raw Message
Post by ColdFusion
X-No-Archive: Yes
I know that mercury arc actually becomes about 10% more efficient when
it is operated on high frequency. I believe it has to do with the arc
not extinguishing at zero crossing as it does on sinusoidal operation
on line frequency.
Low frequency square wave is utilized for filming video to avoid
flickering as well by electronic HID ballast as HIDs can not tolerate
high frequency operation.
If fluorescent lamps were to be operated at 60 or 400Hz square wave,
would it provide 10% gain? How about lamp life? Since with square
wave, the peak and RMS are the same, I would think that it would
provide a LCCF of one which should be beneficial to lamp life.
There are 3 factors here.

One is that a usual low pressure mercury vapor arc has a nonlinearity in
converting electrical power to 254 nm UV, favoring lower and more-constant
current.

The 2nd, which I consider more significant, is that most hot cathode
fluorescent lamps have "oscillatory anode fall" if operated at DC or
frequencies below several hundred Hz. Using AC of frequencies a few KHz
or more usually largely eliminates that loss.

A 3rd factor is that if frequency gets very high, the longish time it
takes (could it be as much as ac few microseconds?) to generate a 254 nm
photon and get it to the wall of the envelope, through numerous
absorptions and re-emmissions, could smooth effects of non-squarewave
current waveforms so that a non-squarewave waveform has efficiency like a
squarewave as long as the electrodes find the waveform as favorable.

Bgds,

- Don Klipstein (Jr) (http://donklipstein.com)
Victor Roberts
2012-05-26 11:51:58 UTC
Permalink
Raw Message
On Thu, 3 May 2012 20:20:10 -0700 (PDT), ColdFusion
Post by ColdFusion
X-No-Archive: Yes
I know that mercury arc actually becomes about 10% more efficient when
it is operated on high frequency. I believe it has to do with the arc
not extinguishing at zero crossing as it does on sinusoidal operation
on line frequency.
Low frequency square wave is utilized for filming video to avoid
flickering as well by electronic HID ballast as HIDs can not tolerate
high frequency operation.
If fluorescent lamps were to be operated at 60 or 400Hz square wave,
would it provide 10% gain? How about lamp life? Since with square
wave, the peak and RMS are the same, I would think that it would
provide a LCCF of one which should be beneficial to lamp life.
Sorry, I've been away too long.

I a 4-foot, T12 fluorescent lamp, the lamp used most often for high
frequency measurements and modeling, the gain from high frequency is
about 10%, and half of that, about 5%, comes from a decrease in
electron density modulation, while the other half, another 5%, comes
from a decrease in anode fall losses at high frequency. Nine of the
gain is due to "starting" but I think you mean the modulation of
electron density.

So, if you run this same lamp at DC, that will give you a 5% efficacy
gain. There was an excellent paper published in 1972 by Drop and
Polman at Philips that shows this effect. I will try to post that
later today. A square wave will be like DC as far as the electron
density is concerned.

Since part of the gain comes from a reduction in end losses, the anode
fall, you can see that the high frequency efficacy gain is a function
of the lamp voltage (length). An infinitely long lamp would have a 5%
gain, because the reduction in end losses is insignificant. A short
lamp, such as a 2-foot, T12, would have a greater gain, about 15%,
because the importance of the end losses is double.
--
Vic Roberts
http://www.RobertsResearchInc.com
http://www.cflfacts.com
sci.engr.lighting Rogues Gallery http://www.langmuir.org
To reply via e-mail:
replace xyz 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
2012-05-26 13:14:19 UTC
Permalink
Raw Message
On Sat, 26 May 2012 07:51:58 -0400, Victor Roberts
Post by Victor Roberts
On Thu, 3 May 2012 20:20:10 -0700 (PDT), ColdFusion
Post by ColdFusion
X-No-Archive: Yes
I know that mercury arc actually becomes about 10% more efficient when
it is operated on high frequency. I believe it has to do with the arc
not extinguishing at zero crossing as it does on sinusoidal operation
on line frequency.
Low frequency square wave is utilized for filming video to avoid
flickering as well by electronic HID ballast as HIDs can not tolerate
high frequency operation.
If fluorescent lamps were to be operated at 60 or 400Hz square wave,
would it provide 10% gain? How about lamp life? Since with square
wave, the peak and RMS are the same, I would think that it would
provide a LCCF of one which should be beneficial to lamp life.
Sorry, I've been away too long.
I a 4-foot, T12 fluorescent lamp, the lamp used most often for high
frequency measurements and modeling, the gain from high frequency is
about 10%, and half of that, about 5%, comes from a decrease in
electron density modulation, while the other half, another 5%, comes
from a decrease in anode fall losses at high frequency. Nine of the
gain is due to "starting" but I think you mean the modulation of
electron density.
So, if you run this same lamp at DC, that will give you a 5% efficacy
gain. There was an excellent paper published in 1972 by Drop and
Polman at Philips that shows this effect. I will try to post that
later today. A square wave will be like DC as far as the electron
density is concerned.
Since part of the gain comes from a reduction in end losses, the anode
fall, you can see that the high frequency efficacy gain is a function
of the lamp voltage (length). An infinitely long lamp would have a 5%
gain, because the reduction in end losses is insignificant. A short
lamp, such as a 2-foot, T12, would have a greater gain, about 15%,
because the importance of the end losses is double.
The Drop and Polman paper has been posted on my website:

http://www.robertsresearchinc.com/Papers/Drop_and_Polman_J_Phys_D_Appl_Phys_5_1972.pdf
--
Vic Roberts
http://www.RobertsResearchInc.com
http://www.cflfacts.com
sci.engr.lighting Rogues Gallery http://www.langmuir.org
To reply via e-mail:
replace xyz 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.
Loading...