Neon and fluorescent lights explained

Newsgroups: sci.electronics
Subject: SUMMARY: How do neon lights work?
Date: 9 Mar 1994 17:50:07 GMT
Organization: Brussels Free Universities (VUB/ULB), Belgium
X-Newsreader: TIN [version 1.2 PL2]

Hello again,

Last week I posted a question about neon lights, I was wondering how they worked and (especially) what components were required to make these lights function.

Well, thank you very much for all your replies. They were very enlightening, primarily because they made me realize how badly I had stated my question :-(

Neon lights

When I talked about "neon lights" I was actually referring to fluorescent tubes and had made the mistake of literally translating the popular french name for these lights into English, thus giving "neon lights". This lead some people to give comprehensive explanations about lamps other than fluorescent tubes. (For which I am grateful too!)

Let's start with those (some messages have been edited for clarity):

1) Neon indicator lamps.
2) Neon "display" or advertising lamps.

1 and 2 are low density neon gas in a sealed container. A voltage, related to the distance between the electrodes (small in the first case, large in the second) "breaks down" (ionizes) the neon[1]. As the electrons fall back, the radiate in the visible, color characteristic of the gas.

The external stuff provides the voltage, and limits the current. Ionized gases had negative resistance (really), so without a limiter, they would destroy themselves or the supply. In the case of (2), supply is a transformer, typical US values are around 6,000VAC and up at a few mA.

[1] In display work, other gases, and combinations of gases, are used to get variant colors. Some colors also result from colored glass tubing.

and also...

I'll assume you're talking about the little indicator lamps, not the fancy store signs. A sufficiently high (circa 90V) voltage ionizes the neon gas and makes it conductive. Once it's ionized a lower voltage will keep it ionized and keep current flowing through the lamp. The light comes from the excited neon ions. The only other component involved is a series resistor to limit the current to something like a few milliamps.

If you mean the fancy store signs, the only other component involved is a high voltage transformer. The "breakdown" voltage must be much higher (10's of KV) because the two electrodes are meters apart instead of 1 mm apart. Different colors come from different combinations of gases.

Neon lamps and neon signs are filled with fairly low pressure gas. It's not a vacuum, but it's at much lower than atmospheric pressure.

And finally...

Neon lights are usually glass filled with an inert gas like neon, argon, xenon, krypton... (noble gases). There is a cathode and anode, usually one at each end of the tube. What happens is that the potential between the anode and cathode (electrical potential) causes the electrons in the gas molecules to be promoted to a higher energy. I'm not sure if you're familiar with the Bohr radius model of electron shells, but it states that electrons in atoms are (for the most part) in finite energy levels. When the electrons are excited by a potential, they are placed on higher energy levels. But, this promotion to the higher energy level does not last very long, and the electrons comes crashing down to it's original state, giving off the absorbed energy in the process in the form of a photon (light). The frequency (and wavelength) of the electron, and hence the color, is determined by the energy difference between the upper (excited) and lower (ground) states. So, while you might say a "neon" light, neon might be just one molecule in there contributing to the color you see.

Fluorescent tubes

Now we'll move on to fluorescent tubes. Here I have summarized their workings myself drawing on some of the emails I received and also (mostly) from an article I found in Elektor magazine (French edition) of May 1989.

The Inner Workings

A fluorescent tube is a glass tube [1] filled with a low pressure mixtures of gases among which are mercury and argon. At each end of the tube are two electrodes [2] with are connected by a filament [3] inside the tube.

             --------------------------------------------- - - - -
             ||                              [4]
       [2]   ||   \  [3]
             ||                              [4]
             --------------------------------------------- - - - -
The inside of the tube is coated with a layer of fluorescent material [4]. Light is produced in a similar fashion as in the emails above about neon lamps, (i.e. a high voltage ionizes the gas, etc) except that here the mercury atoms give off ultraviolet light, causing [4] to emit visible light.

The Circuit

The following circuit is used get the tubes to "light up"

        [1] ~-----------
                       )| [2]
                       |    -------------------------
                       |    |            |          |
                     ----------          |          |
                     |  |  |  |          |          |
                     |  /\/\  |          |          |
                     |        |      [4] |          |
                     |        |        +++++        |
                [3]  |        |        + S +      -----  [5]
                     |        |        +++++      -----
                     |        |          |          |
                     |        |          |          |
                     |        |          |          |
                     |  /\/\  |          |          |
                     |  |  |  |          |          |
                     ----------          |          |
                       |    |            |          |
                       |    -------------------------
       [1] ~------------
When mains voltage [1] is applied to the circuit, the starter [4] (which is basically a switch) allows current to flow through the electrodes of the tube [3]. The current causes the starter's contacts to heat up and open, thus interrupting the flow of current. This causes a great increase in voltage in the inductor [2] (called a ballast here, in the UK & US too I presume :-) which eventually (it doesn't always work first time!) causes the tube to light. Since the gases in the tube have negative resistance, the inductor now plays the role of a current limiter. The capacitor [5] is sometimes used to get voltage and current back in phase.

On smaller fluorescent tubes and lamps (i.e. the ones you can plug in to incandescent lamp sockets) the inductor is sometimes replaced by a resistor.

Additional Info

I thought I'd share some more data from the Elektor article with you...

+ Light output: incandescent lamps output around 10 to 20 lm/W whereas some of the tubes (generally the biggest ones) give off up to 70 lm/W. Thus showing that the tubes are cheaper to run than incandescent lamps.

+ Lifetime: the lifetime of fluorescent tubes and lamps depends greatly on their mean "on-time" (i.e. how long they are emitting light for on average).

Mean on-time (hours)        Mean lifetime (h)        Mean lifetime (h)
                            Fluor. tubes             Fluor lamps
        .25                     2500                     1250
        .5                      5000                     2500
       1                        7000                     3500
       2                        9000                     5000
       5                       12000                     6500
      10                       14000                     7500
(For comparison, the article states the incandescent lamps have a mean lifetime of around 1000 hours, irrespective of on-time)

+ % emitted light: this one's a bit weird (and there wasn't any explanation either!)... It seems that in fluorescent lamps the %age of light they emit depends not only on their age but also on the ambient temperature and on their orientation! For instance, fluor. lamps of around 10W power give off 100% of light at 10 degrees celsius when they are "pointing upwards" (i.e. vertically mounted and socket being the lowest part of the lamp). At 40C the same lamp would only emit at 80% efficiency. If the same lamp is lying sideways it gives off 100% of light at around 25C but only 70% at 10C. Don't ask me why, they didn't say... but I'd like to know if this is true or not (and also why!).

Well, that's all folks. Thanks again for all the info, I hope some of this is useful to you. Oh, and BTW, the "Additional Info" part is *not* meant to start another of our periodical "bulb wars", don't flame me for it, and, yes, I too wish that Elektor had listed their sources of data.

All the best,

Paul-L. Clip            Unit of Behavioural Ecology,
                        Center for Non-Linear Phenomena and Complex Systems,
Tel: +32-2-650-5796     CP 231, Campus Plaine ULB, B-1050 Brussels, Belgium.
Fax: +32-2-650-5767     Email: (no NeXTmail please)

Lamps with wto fluorescent tubes

From: (Andrew Gabriel)
Subject: Re: tubes in series
Date: 29 Sep 1996 14:40:06 GMT
Organization: home

In article <>, (Sparrow) writes:
>This may be a very stupid question, but: Is there any reason why
>flourescent tubes may not be run in series, ie. connecting the two
>pins of one tube directly to those of another, running the starter and
>ballast gear across the whole lot?

Can't read your ASCII-art, but on 220-240v supplies, what you suggest is actually very common for tubes up to 40W. However, each tube has it's own starter, and the ballast will be specially designed for the purpose...

                       ____________        _____________
    ----uuuuuuuu------|            |------|             |-----
        ========     -|____________|-    -|_____________|-
        ballast     |       _        |  |        _        |
                     ------(_)-------    -------(_)-------
                         starter              starter
In most tube failure modes, death of one tube will prevent the other from working too.
Andrew Gabriel                        Home:
Consultant Software Engineer          Work:

Don Klipstein's comments on circuit above

Unless you are fairly exactly duplicating something that works, this may not work. I tried this once, and it largely didn't work. If one starter opens while the other is glowing, the tube seemed to almost never start, and the starting process seemed to have to repeat until one starter opens while the other is closed. Unless the starters do this particular job well, the filaments in the tubes may suffer excessive wear from numerous failed starting attempts.

Except for possible cranky starting problems, this should work if the line/mains voltage is high enough. This voltage should generally be at least 30 percent higher than the normal operating voltage drops of the two tubes for this to work at all, preferably at least something like 80 percent higher for reliable smooth operation.

What is common (at least in the US) are two-tube rapid start ballasts. The tubes are in series with current-limited high voltage applied to them. Additional ballast windings apply voltage to the filaments to heat them up. No starters are used in fixtures with such ballasts.

For more info, please check out the fluorescent lamp FAQ at:

 - Don Klipstein ( or

Wiring for preheat fluorescent fixtures:

The following is the circuit diagram for a typical preheat lamp - one that uses a starter or starting switch.

              Power Switch    +-----------+
 Line 1 (H) o------/ ---------|  Ballast  |-----------+
                              +-----------+           |
                      .--------------------------.    |
 Line 2 (N) o---------|-       Fluorescent      -|----+
                      | )         Tube         ( |
                  +---|-         (bipin)        -|----+
                  |   '--------------------------'    |
                  |                                   |
                  |          +-------------+          |
                  |          |   Starter   |          |
                  +----------| or starting |----------+
                             |   switch    |