# Universal fluorescent ballast

M

#### Mike Jackson

Jan 1, 1970
0
I understand that you have to hold the rapid start lamps at a constant
voltage, and the ballast dims them by adjusting the frequoency. Correct? I'm
trying to understand how a universal ballast would work for fluorescent
lamps. Does the voltage always stay the same but the lamp draws more current
or something because the resistance of the lamp is lower? I'm not quite
sure...? For example, how does a ballast run both a 55W and a 80W
fluorescent lamp? Can it? Thanks in advance, any online references explaning
this would be appreciated.

V

#### Victor Roberts

Jan 1, 1970
0
I understand that you have to hold the rapid start lamps at a constant
voltage, and the ballast dims them by adjusting the frequoency. Correct?

No. Lamps are operated from current sources and the lamp determines it
operating voltage, not the ballast. The ballast must be able to
accommodate the lamp voltage, but it does not set it.
I'm
trying to understand how a universal ballast would work for fluorescent
lamps. Does the voltage always stay the same but the lamp draws more current
or something because the resistance of the lamp is lower? I'm not quite
sure...? For example, how does a ballast run both a 55W and a 80W
fluorescent lamp? Can it?

There is no true "universal" yet. As stated above, lamps are run from
current sources and set their own voltage. All else being equal, a
longer lamp will have higher voltage, Therefore, you can design a
family of lamps that all have the same diameter, and are filled with
the same gas pressure and all operate at the same current. Longer
versions will have higher operating voltage and therefore will draw
more power when operated at the same current. The new "universal"
ballasts are designed to run a variety of lamps that all have the same
or very similar rated operating currents but a wide range of operating
voltages. As long as the ballast can accommodate the different lamp
voltages, it can operate lamps of different powers.

I said there is no true universal ballast because no way has yet been
developed for the ballast to figure out what the rated lamp current
is.
Thanks in advance, any online references explaning
this would be appreciated.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

M

#### Mike Jackson

Jan 1, 1970
0
My other question; is there a maximum frequency rating for fluorescent
lights, and they are dimmed (the rapid start) by changing the frequency
correct? Do you have any documents on line that are good references to this
subject?

V

#### Victor Roberts

Jan 1, 1970
0
My other question; is there a maximum frequency rating for fluorescent
lights, and they are dimmed (the rapid start) by changing the frequency
correct? Do you have any documents on line that are good references to this
subject?

known maximum frequency for operating fluorescent lamps. Fluorescent
lamp products are operated routinely between 50 Hz and 2.5 MHz. Other
prototype lamps operate at 13.56MHz and many people operate
fluorescent lamps at microwave frequencies for a variety of purposes.

The efficacy of fluorescent lamps increases as the frequency is raised
from 50 or 60 Hz to about 10 kHz. Once the frequency is above 10 kHz
the efficacy is just about independent of frequency.

So, how does changing the frequency dim a fluorescent lamp? The
changing frequency does not change the operation of the lamp. However,
all lamps need a current limiting device called a ballast. (Yes, in
the US we also call the entire system that operates the lamp a
"ballast", but technically this term should refer only to the current
limiting element.) If the ballast is an inductor, then increasing the
frequency will increase the impedance of the inductor and therefore
lower the current in the lamp. If the ballast is a capacitor then
reducing the frequency will increase the impedance of the capacitor
and reduce the lamp current.

You can find a discussion of fluorescent lamp operation at
http://members.misty.com/don/f-lamp.html#wd9a

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

M

#### Mike Jackson

Jan 1, 1970
0
I'm an american, so, my "ballast" just has to output X frequency and vary
some amperes to dim the fluorescent light? For example, if I were to use
some FETs to switch voltage -/+ and output a square wave AC signal, and I
just changed the rails to the switching devices, that would provide a
dimming function? So basically in european terms a ballast is some form of
impedence circuit to control the current to the lamp via frequency from a
power supply? Isn't it a little inefficient to change the current using an
impedence circuit? I mean shouldn't you just have switching transistors
switching the voltage from a switch mode power supply? Then you can just use
PWM to control the switch mode power supply to control the dimming. In the
end you would have a PWM switch mode power supply and transistors switching
its voltage (mins and plus so it is an AC square wave....). Then just
changing the voltage the transistors were switching would dim the lamp? This
is also how a rapid start lamp is suppose to be operated correct?

V

#### Victor Roberts

Jan 1, 1970
0
I'm an american, so, my "ballast" just has to output X frequency and vary
some amperes to dim the fluorescent light?

Yes, assuming it controls that current in some way and can supply the
current at the voltage needed by the lamp.
For example, if I were to use
some FETs to switch voltage -/+ and output a square wave AC signal, and I
just changed the rails to the switching devices, that would provide a
dimming function?

No! You have described a voltage source not a current source. If a
fluorescent or other discharge lamp is connected to a voltage source
it will draw as large a current as the source allows and either burn
itself out or burn out the voltage source.
So basically in european terms a ballast is some form of
impedence circuit to control the current to the lamp via frequency from a
power supply?

Not sure what you mean by "European terms". Everyone uses a device to
operate discharge lamps that provides starting voltage and operating
voltage and acts like a current source to control the operating
current. On this side of the pond we call this whole device a
"ballast", even though the term ballast was originally limited to the
current limiting element. The Europeans use the term "control gear"
for the same box that we call a ballast, and therefore avoid the
confusion created by the US terminology.
Isn't it a little inefficient to change the current using an
impedence circuit?

No. Perfect inductors have no power loss. Even imperfect ones can be
rather efficient if designed properly.
I mean shouldn't you just have switching transistors
switching the voltage from a switch mode power supply?

You could use PWM to create a current-controlled source, and this has
been done in certain ballasts. However, the lamp current can increase
to destructive levels in far less than 1 msec, even 1 usec depending
upon how much voltage is applied. If current pulses this short, and
with fast rise and fall times to get high efficiency in the switching
circuit, are applied to the lamp, it will generate a high amount of
EMI and be in violation of FCC regulations. You therefore need some
sort of filter on the output of the switching circuit to smooth out
the pulses.
Then you can just use
PWM to control the switch mode power supply to control the dimming. In the
end you would have a PWM switch mode power supply and transistors switching
its voltage (mins and plus so it is an AC square wave....). Then just
changing the voltage the transistors were switching would dim the lamp? This
is also how a rapid start lamp is suppose to be operated correct?

Not unless you have a current limiting device. See above.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

M

#### Mike Jackson

Jan 1, 1970
0
Ok I'm starting to understand it. I still have to buy a book on switch mode
power supplies I am saving up money for a really god one I have found. Is
there a site that has the operating specs for most fluorescent lamps that
you could refer me to?

V

#### Victor Roberts

Jan 1, 1970
0
Ok I'm starting to understand it. I still have to buy a book on switch mode
power supplies I am saving up money for a really god one I have found. Is
there a site that has the operating specs for most fluorescent lamps that
you could refer me to?

Operating specifications on fluorescent lamps are generally hard to
find without paying .

Commission) http://www.iec.ch/

IEC 60081 (2002-05) double-capped fluorescent lamps (what we call
linear fluorescent lamps). This can be purchased in PDF format for 296
CHF or about \$247. There also two amendments which are slightly less
expensive. There should be an IEC spec on single-capped fluorescent
lamps (CFLs), but I can't find it right now.

Since I don't think you want to spend all this money, you could also
and ANSI C78.901-2001, Single-Base Fluorescent Lamps, Dimensional and

These may not contain operating data on the newest lamp types. You can
sometimes find electrical data on the Philips Lighting or Osram Web
sites. Look on the European sites first instead of the US sites. For
some reason the lamp manufacturers think that European customers need
this data more than US customers. GE has also started publishing
electrical data for some of their lamps, but it is rather hard to
find.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

A

#### Andrew Gabriel

Jan 1, 1970
0
These may not contain operating data on the newest lamp types. You can
sometimes find electrical data on the Philips Lighting or Osram Web
sites. Look on the European sites first instead of the US sites. For
some reason the lamp manufacturers think that European customers need
this data more than US customers. GE has also started publishing
electrical data for some of their lamps, but it is rather hard to
find.

Bare in mind that the power ratings for many fluorescent lamps
are slightly different in the US. Particularly, T8's are mostly
lower powered in the US, and things like the 25W 4' T12 don't
exist outside the US.

M

#### Mike Jackson

Jan 1, 1970
0
Well do most fluorescent lamps like use the same voltage but different
currents are drawn? As far as limiting current, I read that the lamp has a
negative resistance, so wouldn't an isolation transformer take care of that?
I don't understand why you would use a coil. Plus a coil will act as a
voltage divider if it is in series with the lamp, so it would change both
the current and voltage. Of course if the isolation transformer was being
used with the wrong frequency that would affect the current to. Any ideas?

V

#### Victor Roberts

Jan 1, 1970
0
Well do most fluorescent lamps like use the same voltage but different
currents are drawn?

No. Lamp voltage and current varies quite a bit. However, it is
probably more common for a "family" of lamps to operate at the same
current, but to have different operating voltages and therefore powers
because some members of the family are longer than the others. This
allows one electrode design to be used for the whole family since they
all operate at the same current.
As far as limiting current, I read that the lamp has a
negative resistance,

Yes, that is the term used, but the absolute resistance is actually
positive, as it must be to absorb instead of generate power) while the
incremental resistance is negative. That is, as the current increases,
lamp resistance decreases, and so does the lamp voltage.
so wouldn't an isolation transformer take care of that?

No. I don't understand how an isolation transformer would solve
operating problems associated with negative incremental resistance.
I don't understand why you would use a coil.

The coil provides a current limiting function by "absorbing" "excess"
power supply voltage as the lamp current increases. This is best
explained graphically. I will have to post some graphs on my Web site
and then post the link back here.
Plus a coil will act as a
voltage divider if it is in series with the lamp, so it would change both
the current and voltage.

Yes, that is inherent in the current limiting function of the ballast.
An inductor is not a perfect current source, so it does allow the
current to change as the lamp voltage changes. However, if the
inductor is chosen to have a high enough impedance the lamp will
operate in a stable mode in spite of its negative incremental
resistance.
Of course if the isolation transformer was being
used with the wrong frequency that would affect the current to. Any ideas?

Ideal transformers are frequency independent.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

V

#### Victor Roberts

Jan 1, 1970
0
On Sat, 30 Oct 2004 09:50:08 -0400, Victor Roberts

[snip]
The coil provides a current limiting function by "absorbing" "excess"
power supply voltage as the lamp current increases. This is best
explained graphically. I will have to post some graphs on my Web site
and then post the link back here.

I have put a PDF file that explains negative incremental impedance on
my Web site at:

http://www.robertsresearchinc.com/F-Lamp/Need_for_Fluorescent_Lamp_Ballasts.html

This was extracted from a larger presentation and is not fully
any reasonable questions posted here.

The file is for personal, educational use only. it is not to be
reproduced or used for any other purpose.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

M

#### Mike Jackson

Jan 1, 1970
0
Well let me just first make it clear to myself. We are talking about rapid
start lamps right? I have been doing a lot of research on this fluorescent
stuff. I was understanding that only rapid-start ballasts and lamps should
dim because it will eventually damage other lamps? For example, you
shouldn't interchange most fluorescent lamp technologies. Thanks for the

Victor Roberts said:
On Sat, 30 Oct 2004 09:50:08 -0400, Victor Roberts

[snip]
The coil provides a current limiting function by "absorbing" "excess"
power supply voltage as the lamp current increases. This is best
explained graphically. I will have to post some graphs on my Web site
and then post the link back here.

I have put a PDF file that explains negative incremental impedance on
my Web site at:

http://www.robertsresearchinc.com/F-Lamp/Need_for_Fluorescent_Lamp_Ballasts.
html

This was extracted from a larger presentation and is not fully
any reasonable questions posted here.

The file is for personal, educational use only. it is not to be
reproduced or used for any other purpose.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

V

#### Victor Roberts

Jan 1, 1970
0
Well let me just first make it clear to myself. We are talking about rapid
start lamps right?

No. The fundaments, such as need for current limiting, effects of
frequency, etc. apply to all types of fluorescent lamps, and to most
other discharge lamps.
I have been doing a lot of research on this fluorescent
stuff. I was understanding that only rapid-start ballasts and lamps should
dim because it will eventually damage other lamps?

Yes, you are correct that dimming an instant start lamp beyond a
certain point will lead to short electrode life. But then again, even
dimming rapid start lamps on a rapid start ballast that is not
designed to heat the electrodes properly while the lamp is dimmed will
also lead to short electrode life. You need a properly designed rapid
start dimming ballast.
For example, you
shouldn't interchange most fluorescent lamp technologies. Thanks for the

Correct, and thanks.

--
Vic Roberts
http://www.RobertsResearchInc.com
or use e-mail address listed at the Web site.

Replies
0
Views
1K
Replies
24
Views
4K
Replies
39
Views
2K
Replies
2
Views
734
Replies
2
Views
5K