Maker Pro
Maker Pro

Adapting 40W ballast...

A

Artis Sideley

Jan 1, 1970
0
Dear all,

If anyone could tell me how to adapt a monotube 40W electronic ballast
to a 30W fluorescent bulb, how to limit the current Given that the
ballast is not self adaptative .
 
V

Victor Roberts

Jan 1, 1970
0
Dear all,

If anyone could tell me how to adapt a monotube 40W electronic ballast
to a 30W fluorescent bulb, how to limit the current Given that the
ballast is not self adaptative .

The ballast is a current regulator, though often a poor one.
In almost all cases you cannot change the output current.
However, tell us more about these lamps. Some lamp families
include lamps of different power that all operate at the
same current. In those cases the lamps have different
lengths, which will cause them to have different operating
voltages so they can all operate at the same current. In
that case, you can use lamps rated for different powers on
the same ballast.

--
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.
 
V

Victor Roberts

Jan 1, 1970
0
If you download a catalog from a ballast manufacturer such as Advance,
you will see that the same ballast will often be designated for a range
of lamps and combinations.

Most of these lamps are rated for the same or similar
current, but have different operating voltages. Electronic
ballasts have greater load voltage compliance than magnetic
ballasts, which is why they can be used to run lamps of
different power ratings as long as the rated currents of
these lamps are the same +/- 20%.
The difference in performance will be noted
in the specs as having a higher, or lower ballast factor (BF). In your
case, you may simply end up with a slightly brighter lamp than a 30w
ballast would normally give you.

I used to design my own power supplies for musical device projects, so
I have a bit of understanding of how an electronic ballast (not
magnetic) can supply extra current without any damage to the ballast.
The bulb won't care about the probable sight difference in load.

If the current is too high or too low then the lamp will
care.
I actually experimented with a large variety of lamp types and ballasts
and found a large number of lamp/ballast combos I could use to
overdrive lamps to the level of brightness desired by plant growers and
aquarists.

I was going to ask if you had run life tests, but I see
below that you have operated these lamps for longer than a
few hours. That's good.
All of the weird fixtures I created have been operating for years with
no ballast failure OR lamps going out yet.

Can you estimate the operating hours?
Although I am sure that your
intent is not quite along those lines, you can see how I have wired up
many ballast types, both Rapid and Instant start, by viewing the
diagrams I posted at:
http://members.iglou.com/rons0z/od/ovdr.htm

I certainly don't recommend overdriving any lamp since it
will shorten lamp life, but I know that this is done in some
applications where light level is more important than lamp
life.
The cheapest, most versatile ballast I have advised people to use comes
inside a very inexpensive ($8 or less) 2-lamp shoplight sold by Home
Depot - a Sunpark SL-15.

Is this one of those simple series L-C ballasts that creates
a terrible lamp current crest factor?
Your ballast will probably work just fine with the 30w lamp. I would
defer, however, to Victor's advice, as I learned a good deal of what I
know from reading all his posts over the years.

I think we need to find out from the OP what kind of lamp he
or she is using.

--
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.
 
V

Victor Roberts

Jan 1, 1970
0
Victor,

Initially I went through many Aquarium forums, and I read what those
people said about what they were doing.


A lot of them were overdriving by a factor of 3x-4x to get as much
light as possible - giving that they changed their lamps frequently,
regardless of the fact that they still had some life left. If you saw
my diagrams, you see it was a simple matter of parallelling outputs.
This technique only worked with ballasts of certain wiring schemes.

This practice was extremely common among some groups, and all of the
users were fanatically raving happy with the results. I tried it
mainly to advise the people in a website - The Gardenweb Forum's
Growing Under Lights section.
http://forums.gardenweb.com/forums/lights/

Although there were some informed contributers, so much really bad
advice was being given by some of the others there that I tried to
explain as much as I could to correct them. I often considered
inviting you, and the other regulars here in sci.engr.lighting to truly
explain fluorescent workings to the ill-educated. Too many people were
reading "Full Spectrum is Perfect Exact Sunlight" type advertising.
And they were trying to find that perfect magic mix of color
temperature lamps with no regard for the actual spectral content.

Overdriving by 3x or 4x required using a quad lamp ballast, usually
Instant Start, and was the rage of the aquarists. I found that just too
much, and I recommended no more than 2x to those who wanted to do it.
Before I did any of this I studied the websites of Icecap, and Fullham,
who both specialize in providing very high BF ballasts with versatile
diagrammed arrangements, so I tried overdriving out.
Here is Icecap's main page:
www.icecapinc.com/index.htm
A good example of their info/diagramsheets is here at:
www.marinedepot.com/aquarium_lighting_fluorescent_icecap_660_ballast.asp?CartId=

I can only repeat a variation of what I said before. It is
certainly possible to overdrive a fluorescent lamp. Unlike
an HID lamp, a fluorescent lamp will not explode when
overdriven. (DON'T OVERDRIVE ANY HID LAMP!) However,
operating a fluorescent lamp at a current substantially
higher than the lamp rated current will dramatically reduce
the life of the lamp electrodes and hence the lamp. (This
obviously does not apply to electrodeless fluorescent
lamps.) At high enough current the electrodes can be
destroyed in less than 1 second. So, yes, fluorescent lamps
can be overdriven, but anyone who suggests that this can be
done without a very substantial reduction in lamp life is
seriously misleading their audience.
Measuring a quad ballast with a 32wT8 I got these line amps
1x - .261A
2x - .462A
4x - .726A
Light output did not increase linearly. I do not own a calibrated
light meter, but someone who did measured about 169% light output at 2x
and 236% at 4x. With an inexpensive plantlight meter I noticed that
the output dropped slightly as the tube warmed to full extent...
definitely not optimum. By the way, since the original BF was .88 on
this ballast I figure I had created a 1.5BF ballast (.88x169)

The light fails to increase linearly for two reasons. The
simple one is that the mercury vapor pressure is driven too
high. Fluorescent lamps are designed to maintain the optimum
mercury vapor pressure when the lamp is operated in still
air at 25C (35 C for linear T5 lamps) and operated at the
lamp's rated current. If the lamp is driven at higher than
rated power without any additional cooling then it is
obvious that the lamp temperature will increase and that
will increase the temperature of the mercury or mercury
amalgam and that will increase the mercury vapor pressure
above the optimum value and that will reduce the percentage
of input power converted into 254 nm and 185 nm UV. I know
of one commercial application that uses overdriven
fluorescent lamps that also uses forced air cooling to keep
the mercury vapor pressure at the optimum value. However,
some sort of feedback system is ideally used to prevent the
mercury vapor pressure from being driven too low.

The second reason is that even if the mercury vapor pressure
is held at its optimum value the UV output of a low pressure
mercury-rare gas discharge will saturate as the discharge
current density is increased. For more on this I recommend
the book Electric Discharge Lamps by John Waymouth, MIT
Press, ~ 1971
I assume you mean one of those nearly invisible circuits hidden in the
endframes of those featherweight shop lights.

Yes.

[snip]

--
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.
 
D

Don Klipstein

Jan 1, 1970
0
Zink wrote said:
Too many people were reading "Full Spectrum is Perfect Exact Sunlight"
type advertising. And they were trying to find that perfect magic mix of
color temperature lamps with no regard for the actual spectral content.

"Full spectrum", although lacking any specific definition accepted
throughout the lighting industry, sounds to me like something not what one
wants anyway to best-satisfy lighting needs live organisms that have
lighting needs. This is despite replication of sunlight in spectral
properties satisfying most "proposed definitions" of "full spectrum".

Replication of sunlight in spectral properties is not the optimum for
artificial lighting to satisfy lighting needs of organisms that have
photochemical processes that make such organisms require light. The
reason is that such organisms do not make good use of all wavelengths that
are present in sunlight.

Green light is wasted on plants since plants do not much absorb and
therefore do not make much utilization of green wavelengths. This is why
fluorescent lamps made for growing plants specialize in producing
wavelengths that chlorophyll utilizes - mainly red and secondarily blue.

Live coral has a requirement of light best-satisfied by wavelengths in
the blue-violet to mid-blue range (near 410 to mid-400's nanometers).

Some members of the animal kingdom have a requirement of Vitamin D, and
a way of satisfying such is with ultraviolet of wavelengths somewhere in
the low 300's of nm. Humans are an example of members of the animal
kingdom that can have Vitamin D requirements satisfied by such UV or also
by ingestion of Vitamin D.

Beyond that, spectral sensitivities are mostly in known and suspected
photoreceptors in eyes of members of the animal kingdom - ranging from low
400's to close to 590 nm, in addition to one UV-centered one specific to
arthropods and centered around 350 nm (give-or-take and maybe). The
longest wavelengths needed for good color vision are wavelenths utilized
mainly by the longest wavelength vision photoreceptor. The longest
wavelength one known used in vision by anything with eyes (as far as I
know) is not centered/peaking at wavelengths much longer than those of the
longwave/"red" photoreception mechanism of humans. And triphosphor
fluorescent lamps with most "red" output in a single very narrow band
centered at 611 nm achieve a color rendering index of 82-86! I saw a
published spectrum of a fluorescent lamp with rated color rendering index
of 98, and most of its red/reddish output was at wavelengths well below
650 nm!

Bottom line - I advise against buying into some notion that you need
"full spectrum" or "spectral replication of sunlight" to satisfy
photochemical requirements of organisms, since organisms with light
requirements do not require nor can even utilize most of the wavelengths
in sunlight. As an extreme example of what is in sunlight ay Earth's
surface and useless for probably anything - the infrared wavelengths that
are able to penetrate the atmosphere but not water, generally most from
about or hardly over 1500 to low 2,000's nm.

- Don Klipstein ([email protected])
 
A

Artis Sideley

Jan 1, 1970
0
Thanks to you very helpful information, I'm giving more details hereafter:

I intend to use this ballast with a 30W T8 fluo' lamp. I don't know
anything about the current neither the output voltage of the ballast ,
but it seems to work well with a 38w T8 lamp because I've tried it
during a few minutes. Besides, I understand that overdriving a lamp
decreases its lifetime what I want to avoid. That's why I'm asking if
there is any mean to decrease the current while maintaining the
operating voltage.

I have no further tech specs about these ballasts I bought for a bargain.

About the light spctrum: I've noticed that full spectrum lamps delivers
more lumens than limited spectrum lamps(e.g plant culture lamps with
spectrum limited to the red and the blue ray) but if only interesting
rays are considered (blue and red), which kind of lamp delivers more ?
(for the same electric power)
 
T

TKM

Jan 1, 1970
0
Don Klipstein said:
"Full spectrum", although lacking any specific definition accepted
throughout the lighting industry, sounds to me like something not what one
wants anyway to best-satisfy lighting needs live organisms that have
lighting needs. This is despite replication of sunlight in spectral
properties satisfying most "proposed definitions" of "full spectrum".

Replication of sunlight in spectral properties is not the optimum for
artificial lighting to satisfy lighting needs of organisms that have
photochemical processes that make such organisms require light. The
reason is that such organisms do not make good use of all wavelengths that
are present in sunlight.

Green light is wasted on plants since plants do not much absorb and
therefore do not make much utilization of green wavelengths. This is why
fluorescent lamps made for growing plants specialize in producing
wavelengths that chlorophyll utilizes - mainly red and secondarily blue.

Live coral has a requirement of light best-satisfied by wavelengths in
the blue-violet to mid-blue range (near 410 to mid-400's nanometers).

Some members of the animal kingdom have a requirement of Vitamin D, and
a way of satisfying such is with ultraviolet of wavelengths somewhere in
the low 300's of nm. Humans are an example of members of the animal
kingdom that can have Vitamin D requirements satisfied by such UV or also
by ingestion of Vitamin D.

Beyond that, spectral sensitivities are mostly in known and suspected
photoreceptors in eyes of members of the animal kingdom - ranging from low
400's to close to 590 nm, in addition to one UV-centered one specific to
arthropods and centered around 350 nm (give-or-take and maybe). The
longest wavelengths needed for good color vision are wavelenths utilized
mainly by the longest wavelength vision photoreceptor. The longest
wavelength one known used in vision by anything with eyes (as far as I
know) is not centered/peaking at wavelengths much longer than those of the
longwave/"red" photoreception mechanism of humans. And triphosphor
fluorescent lamps with most "red" output in a single very narrow band
centered at 611 nm achieve a color rendering index of 82-86! I saw a
published spectrum of a fluorescent lamp with rated color rendering index
of 98, and most of its red/reddish output was at wavelengths well below
650 nm!

Bottom line - I advise against buying into some notion that you need
"full spectrum" or "spectral replication of sunlight" to satisfy
photochemical requirements of organisms, since organisms with light
requirements do not require nor can even utilize most of the wavelengths
in sunlight. As an extreme example of what is in sunlight ay Earth's
surface and useless for probably anything - the infrared wavelengths that
are able to penetrate the atmosphere but not water, generally most from
about or hardly over 1500 to low 2,000's nm.

- Don Klipstein ([email protected])

To Don's excellent information, I would only add that since light is energy,
there are factors in addition to spectrum that must be taken into account.
These are: intensity, duration, timing (when the light is applied) and
spatial distribution. It's the full "dose", which equates to energy
absorbed by the organism, that matters. Unfortunately, many think that for
both humans and other animals, installing a "full spectrum" lamp is all
that's needed. The reality, for humans and probably for other animals too,
is that color is one of the least important factors. Compared to what we
call, more or less, white light, any effects from "full spectrum" light are
hard to find.

What is the most important factor? No doubt about it; intensity.

Terry McGowan
 
T

TKM

Jan 1, 1970
0
Zink said:
Believe it or not, while perusing some of the Aquarium forums, I ran
into several descriptions of methods for overdriving HPS and MH lamps.

I have assembled and repaired a small range of HID fixtures, but I
would not attempt what they were suggesting. I did not quite
understand their explanations either. I think they were just preparing
to explode themselves. I have, on purpose, avoided exposing my flesh to
flying shards of hot glass all my life.

Ron Seadler


A very good policy!

Mercury and metal halide HID lamps are the most likely to rupture; HPS less
so. I visited a factory several years ago to trouble-shoot a problem of
cycling HPS lamps. New lamps installed in the fixtures went out after a few
minutes and then came on again a few mintues later only to brighten up and
go out again. After checking a few lamps, it was found that 250 watt HPS
lamps had been installed in sockets designed for 400 watt HPS. In such a
situation, the ballast will do its best to drive the lamp at the current
level of a 400 watt lamp. Light output and color were great -- for a few
minutes; however, what I learned was that overdriving the 250 watt lamps for
several days didn't ruin them and they didn't rupture although it apparently
did shorten lamp life.

I can't say that what I found applies to all HPS lamps plus dangerous things
can happen when any lamp is operated outside of its specification range; so,
as indicated above, don't do it.

Ron, a suggestion: Download the technical data sheets for some fluorescent
and HID lamps from the lamp manufacturers. Some will show lamp current,
voltage, etc.for a particular lamp. GE calls theirs "LSBs". Better yet
would be a copy of the lamp standards from ANSI where data for all lamps are
listed, but there is a charge for those. ANSI C78.81-2001 (double-ended
lamps) and C78.901-2001 (single-ended lamps) are the ones to look for. Vic
may know of other sources too.

Terry McGowan
 
V

Victor Roberts

Jan 1, 1970
0
Ron, a suggestion: Download the technical data sheets for some fluorescent
and HID lamps from the lamp manufacturers. Some will show lamp current,
voltage, etc.for a particular lamp. GE calls theirs "LSBs". Better yet
would be a copy of the lamp standards from ANSI where data for all lamps are
listed, but there is a charge for those. ANSI C78.81-2001 (double-ended
lamps) and C78.901-2001 (single-ended lamps) are the ones to look for. Vic
may know of other sources too.

Terry - good suggestion. Most lamp data provided on the
manufacturer's US web sites is incomplete. Philips and Osram
sometimes provide far more technical data on their European
sites than the do on their North American sites for the same
products. GE does not provide much technical data, but they
have recently posted a few data sheets in PDF format for
newer lamps.

The ANSI specs are pretty good and many of them are free at
either www.ansi.org or www.nema.org.

The most complete set of lamp data is found at www.iec.ch.
For example IEC 60081 - Double-capped fluorescent lamps -
Performance. But it costs 300 CHF for the PDF version
(about $235 at today's rate). There is also an IEC spec for
whet we call compact fluorescent lamps and they call
single-ended fluorescent lamps.

--
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.
 
A

Artis Sideley

Jan 1, 1970
0
Thanks a lot. I don't have the circuit diagram unfortunately, however, I
understand that a capacitor in serial upfront the lamp may decrease the
frequency and the current consequently ?!

About the spectrum point: Say we have 2 fluorescent lamps , one is "full
spectrump" , the other is a gro-lux, both have a same electrical power ,
but which one delivers more light power in the blue and red rays ?
 
V

Victor Roberts

Jan 1, 1970
0
Thanks a lot. I don't have the circuit diagram unfortunately, however, I
understand that a capacitor in serial upfront the lamp may decrease the
frequency and the current consequently ?!

Is this electronic ballast instant start of some variation
of what we in the Colonies call raid start? If it is not
instant start, that is if it has two leads from the ballast
to each end of the lamp then you cannot just add a series
impedance between the lamp and ballast because that will
severely reduce the electrode heating current.

Series impedance elements were used during the energy crunch
of the 1970's to reduce lap power, but when used with rapid
start lamps they always included a small transformer with a
1:1 turns ratio to carry the electrode heating power over
the additional series impedance.

--
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.
 
A

Artis Sideley

Jan 1, 1970
0
My ballasts look like your description. Now I'd better to purchase
adequate ballast.

Thanks.
 
V

Victor Roberts

Jan 1, 1970
0
Is this electronic ballast instant start of some variation
of what we in the Colonies call raid start?

My New Year's Resolution - MUST PROOFREAD BETTER.

The previous statement should have said "rapid start" not
"raid start".
If it is not
instant start, that is if it has two leads from the ballast
to each end of the lamp then you cannot just add a series
impedance between the lamp and ballast because that will
severely reduce the electrode heating current.

Series impedance elements were used during the energy crunch
of the 1970's to reduce lap power,

And this one should have said "lamp" not "lap".
but when used with rapid
start lamps they always included a small transformer with a
1:1 turns ratio to carry the electrode heating power over
the additional series impedance.

And this would have been clearer if it said "around the
additional series impedance" instead of "over".

--
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.
 
V

Victor Roberts

Jan 1, 1970
0
Hang on, s/he's trying to reduce it for a 30w tube on a 40w ballast, so
reducing heater power as well as run power is what's wanted.

The situation is different when youre underrunning a 40w tube, as you
then need the same heater power as a 40w tube, but lower run current.

Why do you think less electrode heating power is required?
Most rapid start lamps (in the US at least) use 3.6 volts to
heat the electrodes, though there are some that differ.
So series impedance should work fine.

I disagree. First with a rapid start system you would have
to either 1) add two inductors, one in each electrode lead
at the end of the lamp where you place the additional
impedance, or 2) leave one of the electrode leads
disconnected - which would eliminate the electrode heating
voltage that is the heart of rapid start. And since the lamp
operating voltage is ~ 100 volts and the electrode heating
voltage is ~ 4 volts, any set of two inductors that have
enough impedance to reduce the lamp current will have so
much voltage drop that the electrode heating voltage will be
reduced to zero. The use of a 1:1 transformer to carry the
heating voltage around the series impedance allows you to
use one series inductor or capacitor and still have the
proper electrodes heating voltage.

--
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.
 
V

Victor Roberts

Jan 1, 1970
0
OK, lets see. We're comparing a 30w tube with a 4' 40w. The 3' tube
takes less current while starting. The heater v might be the same, but
the v drops arent. With a simple L ballast, a 30w ballast has higher
impedance than a 40w ballast. It is almost entirely the ballast that
controls i during preheat, since R_filaments is so small, thus
i_preheat is lower with a 30w tube.

When you run a 40w tube at 30w, you need the same preheat current as a
40w tube, iow you need to reduce run current but not reduce preheat
current. When running a 30w tube at 30w on a 40w ballast, one needs to
reduce both i_run and i_preheat. This is normally done by higher
ballast impedance.

We were talking about a rapid start system, where the
electrode heating power is provided by a fixed voltage
transformer, not a preheat system where the electrodes are
heated by a the current that passes through the ballast
inductor when the "back" side of the two electrodes are
shorted together. Also, other than low cost integral CFL
ballasts, I have not seen any electronic ballasts that heat
fluorescent lamp electrodes using the arc current supply.
They all use separate windings on the output transformer.

Perhaps this discussion suffering from cross-pond language
differences?

--
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.
 
V

Victor Roberts

Jan 1, 1970
0
Victor,

You had asked about lamp lifetimes for overdriven lamps, which I could
not yet answer. I just went around the house today to look at various
lamps (looking to verify I had no PASTEL spectrums seen just recently
on a discussed website) and saw that I finally had an overdriven lamp
which had blackened ends and had gone out.

[big snip]

Thanks for posting the data.

--
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.
 
Top