NE-2H (Neon Lamp) Specifications

[Jim Thompson]

Anyone have specifications for NE-2H (Neon Lamp)?

Striking voltage, valley voltage, etc.?

Thanks!

...Jim Thompson

 

[Ian Stirling]

Gee, I used to use those all the time. Maybe 90 volts to fire one and
about 60 volts to drop back out - and they glow a very pretty blue when
you run RF thru them...

And emit quite a loud report when current exceeds a few amps.

Are there any variants that have decent (>20 years on) life?

 

[Luhan Monat]

Anyone have specifications for NE-2H (Neon Lamp)?

Striking voltage, valley voltage, etc.?

Thanks!

...Jim Thompson

Gee, I used to use those all the time. Maybe 90 volts to fire one and
about 60 volts to drop back out - and they glow a very pretty blue when
you run RF thru them...

 

[Jim Thompson]

Chicago Miniature has them in their catalog.

Link is at:

http://www.chml.com/assets/databookpdf/5-4.pdf

Good Luck,

Al

Thanks, Just what I needed.

...Jim Thompson

 

[Don Klipstein]

Anyone have specifications for NE-2H (Neon Lamp)?

Striking voltage, valley voltage, etc.?

Striking voltage varies so much and so wildly with age, condition,
manufacturer, etc... If it reliably strikes at 110 volts AC (peak voltage
155 volts) in the dark I would not consider it failed/defective. And they
have some photoelectric effect!

Valley voltage: Varies widely with manufacturer, age, condition...
Anywhere from about 54 volts to 80 volts or maybe even a bit more
(well-used). I would say typically 65 volts valley voltage.

Valley current - Design for the voltage drop to be close to the valley
voltage at any current from 2.5 to 4 mA. There may even be more than one
valley in the 2 to 4 mA range. Secondary valleys sometimes exist
sometimes at somewhat lower currents. The lamp may flicker if the current
is low enough to have the electrodes not fully or nearly fully covered
with glow (the glow may jump around like it does in "flicker flame" neon
lamps) - you need the current to exceed that of the highest-current valley
to avoid that. But have current exceeding that of the highest valley by
as little as practical - current in excess of that of the highest valley
results in "abnormal glow" or glow having "above-normal current density",
and that causes accelerated wear on the electrodes (and the bulb being
darkened by "sputtered" electrode material).

Life expectancy: At 120 volts AC, traditionally 20,000 or 25,000 hours
with a 33K resistor, 5,000 hours with a 22K resistor. At 120 volts AC, a
resistor of 39K or 47K usually results in current close to the main or
highest valley current - going higher extends life no more than
proportionately.

NE2H is sometimes AKA C2A. There is a mini-NE2H sometimes AKA A1C,
which is Radio Shack 272-1102, even though the catalog claims that it is a
C2A.

These are "high intensity" neon lamps, which contain nearly enough pure
neon.
"Standard intensity" neon lamps have 99.5% neon .5% argon, and these
include NE2 (AKA A1A) and the mini-NE2 (AKA A1B, which the Radio Shack
272-1100 usually is). The color of "standard intensity" neon lamps is a
"soft orange" rather than a brighter slightly reddish orange.
Although argon is expected to make the glow purpler, it does this more
in the main discharge column, which low pressure glow lamps having closely
spaced electrodes lack. In the electrode glow ("cathode/negative glow")
of a neon lamp, argon radiates pretty much only its strong infrared
wavelengths, and detracting less from a yellow wavelength of neon (585 nm)
than from neon's red and orange wavelengths. (My guess - but the effect
of half a percent argon is definitely making the color of the
cathode/negative glow yellower, as in a less-red shade of orange!)
The valley current of the NE2/A1A is less than a milliamp - the
traditionally recommended resistor for use at 120 VAC is 220K. The A1B
"mini-NE2" appears to me to want a 330K resistor at 120 VAC.
In these "standard intensity" neon lamps that have 99.5% neon .5% argon,
the valley voltage is a little lower - generally in the range of 50 to 65
volts, probably usually mid or upper 50's. The striking voltage is
usually under 100 volts, but varies widely with age, condition, incoming
light, and frequency or whatever can effect the electric field
distribution within a lamp that is trying to start. There are variants
with Krypton-85 added to assist starting (look for a very slightly pinkish
color, milder than usual of that of neon lamps that have had air leak into
them) - strinking voltage with those is probably reasonably reliably not
far from 75-80 volts, if you allow a few seconds for a beta particle to do
some good within the lamp when the peak voltage is that low. And Kr-85
has a halflife of only 9.4 years - give allowance accordingly to ones
manufactured a few decades ago back when this was not so politically
incorrect!

- Don Klipstein ([email protected])

 

[analog]

Anyone have specifications for NE-2H (Neon Lamp)?

Striking voltage, valley voltage, etc.?

Ok, how about an LTspice model that starts with the following parameters:

V_strike = 130V (strike voltage in fully non ionized gas)
I_ionization = .5ma (threshold current between strike and hold regions)
Tau_ionization = 1ms (gas ionization time constant)
V_hold = 80V (Zener-like hold voltage above ionization current)
R_hold = 1k (series resistance within hold current range)
I_abnormal_glow = 4ma (current beyond which series resistance increases)

I think this should cover all of the parameters important to model the
normal operation of a neon bulb. The decrease in voltage beyond the
abnormal glow voltage rise could easily be added onto this if it were
required for some reason.

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* NE-2H Small Neon Lamp Behavioral Model by [email protected]
* Vs: strike voltage (bi-directional)
* Ii: threshold current boundary between strike and hold voltage
* Ti: ionization time constant of gas
* Vh: hold (regulation) voltage within normal current range
* Rh: on resistance within normal current range
* Ia: transition current to the increasing resistance of abnormal glow
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
..subckt NE-2H A1 A2 params: Vs=130 Ii=.5m Ti=1m Vh=80 Rh=1k Ia=4m
Varc A1 arc 0 ; current sense
Barc arc A2 I= sgn(V(arc,A2))*uramp(abs(V(arc,A2))-V(ref))/(Rh*V(abn))
Carc arc A2 5p ; stray terminal capacitance
Bref 0 ref I= Vh+{Vs-Vh}/(1+V(ion)**2) Rpar=1 ; voltage transition
Cref ref 0 1n ; tiny capacitance here aids convergence
Bion 0 ion I= abs(I(Varc)) Rpar={1/Ii} ; measure of free ions
Cion ion 0 {Ti*Ii} ; gas ionization time constant
Babn 0 abn I= {Ia**2}+I(Varc)**2 Rpar={1/Ia**2} ; abnormal glow
Cabn abn 0 1p ; tiny capacitance here aids convergence
..ends NE-2H

Note that all static calculations involving only parameters (and not
dynamic variables from the circuit matrix) are enclosed within curly
braces so that they are evaluated once only as the subcircuit is
first read in and not at every step during the transient analysis.

Also, note that the model has only one significant time constant and
this represents the lifetime of conductive ions within the gas. Thus,
the neon lamp could theoretically be represented by only two sources,
one to model the ionization time constant and one to model all of the
non-linear effects combined (which are not time dependent).

In fact, I found that breaking these non-linear calculations into
several parts actually makes the model run faster and more reliably
even though the resulting circuit matrix is somewhat larger. I
suppose this is because the matrix solver requires quasi-linear
behavior to get from one time point to the next, so that extreme
non-linear behavior forces extremely small time steps.

I find the understanding and modeling of glow and arc discharges to
be fascinating because of the unusual negative incremental resistance
characteristic of arcs at low frequencies. A number of very good
papers on the subject of modeling fluorescent lamps are contained
within several of the IEEE power electronics conference proceedings
published within the last decade. As I recall, these papers
concentrated on modeling the ionization time constants and negative
(V = 1/sqrt(I)) arc resistance stemming from the area and density
modulation of the conductive column within the gas in the arc path.
But I don't recall coming across a paper exclusively addressing the
very much lower current density glow discharge processes within a
neon bulb.

With a neon bulb there isn't enough current to heat up the closely
spaced electrodes so the cathode operates cold and also there is no
discharge column to contribute to the voltage drop. As long as some
ions are already present and the surface of the cathode isn't saturated
with current, the voltage required to generate additional conductors is
largely independent of current and consists of what is known as the
cathode fall voltage. Initial striking voltage is higher because no
conductors yet exist to be accelerated by the electric field into
regenerating additional conductor through collisions. The first loose
electrons must either be ripped from their orbits by electric field
alone or with photoelectric assist.

Once the surfaces of the electrodes are entirely covered with glow, it
takes a higher energy to crowd in additional conductors and the neon
bulb's normal glow voltage begins to rise with current. This is the
abnormal glow region of operation and can damage the electrodes by
blasting off metal atoms. Pushing the current even higher will heat
up the electrodes to the point of thermonic emission, leading to a
reduced on voltage as the cathode fall disappears. This is the
negative resistance region of operation of an arc discharge and cannot
be sustained for any length of time by a tiny neon bulb (boom!).

What this all means is that a good spice model for a neon bulb may not
be the same as a model one for a fluorescent bulb or high pressure
discharge lamp. An all encompassing model may be possible, but would
probably be too complicated to run as quickly as a simpler, application
specific model would.

 

[Ian Stirling]

Striking voltage varies so much and so wildly with age, condition,
manufacturer, etc... If it reliably strikes at 110 volts AC (peak voltage
155 volts) in the dark I would not consider it failed/defective. And they
have some photoelectric effect!

Even on 240V failed neons re quite common.

 

[Jim Thompson]

Ok, how about an LTspice model that starts with the following parameters:

V_strike = 130V (strike voltage in fully non ionized gas)
I_ionization = .5ma (threshold current between strike and hold regions)
Tau_ionization = 1ms (gas ionization time constant)
V_hold = 80V (Zener-like hold voltage above ionization current)
R_hold = 1k (series resistance within hold current range)
I_abnormal_glow = 4ma (current beyond which series resistance increases)

[snip]

Analog, I_ionozation seems quite large... order of magnitude higher
than with your NE-2 model; and gives rather bizarre behavior when used
in a "multivibrator" type circuit.

...Jim Thompson

 

[Tim Shoppa]

And Kr-85
has a halflife of only 9.4 years - give allowance accordingly to ones
manufactured a few decades ago back when this was not so politically
incorrect!

I have some OA2 (gas voltage regulator) tubes from the early 70's which
have a disclaimer on the side about Krypton in them that I never understood
before. I understand a little more now!

Tim.

 

[Jim Thompson]

I have some OA2 (gas voltage regulator) tubes from the early 70's which
have a disclaimer on the side about Krypton in them that I never understood
before. I understand a little more now!

Tim.

I learned the hard way about the behavior of gas voltage regulator
tubes that *don't* have the Krypton...

In 1961, I had my big-ass "toooobz" amplifier in the attic (I rented
the third floor of an old house in Cambridge).

Went home to WV for a few weeks that summer and I shut down the
normally always-on power supply.

When I came back I flipped the switch and ¡KA-BOOM!, every lytic went
off like a rocket.

Seems the striking voltage rises when Krypton-less regulator tubes sit
in the dark for a while :-(

...Jim Thompson

 

[Oppie]

Anyone have specifications for NE-2H (Neon Lamp)?

Striking voltage, valley voltage, etc.?

Thanks!

...Jim Thompson

Jim,

If you would also like, I can repost the scan of 'The Ubiquitous Neon' from
the "Electrical Experimenter's Handbook" (1969)

Great applications for the NE-2 series of indicators. Useful for everything
from voltage regulators to bistable logic elements and just about everything
in between.

Oppie

 

[Jim Thompson]

Jim,

If you would also like, I can repost the scan of 'The Ubiquitous Neon' from
the "Electrical Experimenter's Handbook" (1969)

Great applications for the NE-2 series of indicators. Useful for everything
from voltage regulators to bistable logic elements and just about everything
in between.

Oppie

Hi Oppie, Sure would appreciate that! Thanks!

...Jim Thompson

 

[Chuck Harris]

I think I even saw a "ring oscillator" once. That's like an
astable, but with more than 2 stages. Haven't built one yet,
however.

I built one for my dad's BD when I was a kid. It used 3 scrap
neon indicators from an old IBM mainframe, and just went blink,
blink,blink... in a triangle pattern.

It sat on my dad's night stand for an awfully long time blinking
away. At one point, many many years later, the bulbs grew dim,
and he asked me if I could fix it for him, so I did.

Funny the things one's parent's just like... fond memories, I
guess.

-Chuck

 

[JeffM]

they have some photoelectric effect

Don Klipstein

Yup. Baldwin used NE-2's in the triggering of the percussion section
of a generation of their organs.
The circuitry had a row of Type 47 bulbs along its length
to optically bias the neons.

 

[John Woodgate]

I read in sci.electronics.design that Oppie <[email protected]>

If you would also like, I can repost the scan of 'The Ubiquitous Neon'
from the "Electrical Experimenter's Handbook" (1969)

How do ubiquitious neons differ from ordinary ones?

 

[Tim Shoppa]

I think I even saw a "ring oscillator" once. That's like an
astable, but with more than 2 stages. Haven't built one yet,
however.

There's a 5-stage one illustrated at

http://members.misty.com/don/sillyne2.html

I've built it and it works well: even though 150V and 200V
B+ batteries aren't around anymore 9-volts stack very nicely.

It's worth noting that the design is perfectly symmetric; with
well-matched components it'll go left-to-right just as likely
as right-to-left. The symmetry is broken at startup by the firing of the
first and second adjacent neon bulbs. Sometimes a third bulb will fire
out of sequence at startup, but after that it'll be in sequence.

I think this is an example of "spontaneous symmetry breaking". After
building this neon circuit I now understand sentences in my graduate
physics books like "it can be seen how the spontaneous symmetry breaking
of the U(1) symmetry caused by the degenerate energy minimum of the
Lagrangian (2.10) creates a pertubative theory with a massive scalar boson."

I've tried extending it to 10 neons but it doesn't run so stably there...
sometimes bizarre patterns pop up seemingly at random, and other
times it will run in sequence. The crosswire
that "inhibits" the non-adjacent bulb from going off isn't quite so
effective compared to the random ionization levels, I guess.

Tim.

 

[Chuck Harris]

How do ubiquitious neons differ from ordinary ones?

They're simply *everywhere*.

-Chuck

 

[Jim Thompson]

Hi Oppie, Sure would appreciate that! Thanks!

...Jim Thompson

See attached PDF [early Christmas Present]
Regards to all

Thanks, Oppie!!

...Jim Thompson

 

[Louis Scheffer]

The first digital clock I ever saw (about 1965) used Nixie tubes as both
the output display and the storage elements. Each digit was wired in a
ring-oscillator like (but stable) arrangement. When the input pulse came
in the current number was turned off and the next on. Clearly there must
have been some trick to get it into the right initial state, or at least
some initial state with only one lamp lit. but at the time (7th grade) I
didn't know enough to ask about that. (and I hope I'm remembering it correctly)

There were neon tubes optimized for this sort of operation, with a single
small dot for each state, rather than a whole digit.

http://www.wps.com/projects/decimal-tubes/index.html#DECIMAL

but as best I remember the design used the Nixie (with perhaps a single
transistor/digit) for the state storage as well as the display. Of course
it did not need to count very fast....

Lou Scheffer

 

[Oppie]

The first digital clock I ever saw (about 1965) used Nixie tubes as both
the output display and the storage elements. Each digit was wired in a
ring-oscillator like (but stable) arrangement. When the input pulse came
in the current number was turned off and the next on. Clearly there must
have been some trick to get it into the right initial state, or at least
some initial state with only one lamp lit. but at the time (7th grade) I
didn't know enough to ask about that. (and I hope I'm remembering it correctly)

There were neon tubes optimized for this sort of operation, with a single
small dot for each state, rather than a whole digit.

http://www.wps.com/projects/decimal-tubes/index.html#DECIMAL

but as best I remember the design used the Nixie (with perhaps a single
transistor/digit) for the state storage as well as the display. Of course
it did not need to count very fast....

Lou Scheffer

Similar to the early nuclear monitoring equipment. the earliest Gieger
counter with electronic counters had several columns of indicators labeled
0-9. they were neon type indicators that advanced with each count. at '9'
reset to 0 and cary to next column. Awkward to read but from what I
understand, implemented only with the neon indicators as the counting
elements.

Oppie