So, an electronic ballast has a rating, as well as, in some cases a lamp
type. OK, thats simple. But then, there are so many lamp types and
wattages, that keeping a stock of ballasts in the van for each type of
lamp would be costly.
Testing a few yesterday, it seems they cannot be reliably inter-changed
between types.
Why wouldn't a 36w linear ballast work with a 28w 2D lamp, and vice
versa?
OK, rating at 8w over rating may be asking too much, but why not run a
28w off a 36w ballast?

And a 2 x 36W works with one lamp lit, the other not working, but
wouldnt work with a single, good, lamp fitted?



--
To reply by e-mail, change the ' + ' to 'plus'.

On Sunday, December 23, 2012 9:15:57 AM UTC, A.Lee wrote:

So, an electronic ballast has a rating, as well as, in some cases a lamp
type. OK, thats simple. But then, there are so many lamp types and
wattages, that keeping a stock of ballasts in the van for each type of
lamp would be costly.
Testing a few yesterday, it seems they cannot be reliably inter-changed
between types.
Why wouldn't a 36w linear ballast work with a 28w 2D lamp, and vice
versa?
OK, rating at 8w over rating may be asking too much, but why not run a
28w off a 36w ballast?
And a 2 x 36W works with one lamp lit, the other not working, but
wouldnt work with a single, good, lamp fitted?

A fluorescent tube requires the correct power, or close to it, the right V & i which depend on tube width, length and design features, and also the correct heater power, v and i. Hence ballasts that will by design run more than one tube type are a minority.

A 36w ballast on a 28w lamp would overrun the lamp, reducing life expectancy.

Its certainly possible to do some ballast/tube type swapping. Iron ballasts are easily adapted to run lower power tubes by adding the right capacitor, but the calculation of its value is nontrivial and correct calculation is essential. CFL ballasts often run small fl tubes happily enough. In principle modern electronic ballasts can run lower power tubes with 2 capacitors on the output, but its too far from trivial to be practical.

Perhaps someone should make a ballast with multiple settings.


NT

On 23/12/2012 09:43, wrote:
On Sunday, December 23, 2012 9:15:57 AM UTC, A.Lee wrote:

So, an electronic ballast has a rating, as well as, in some cases a lamp
type. OK, thats simple. But then, there are so many lamp types and
wattages, that keeping a stock of ballasts in the van for each type of
lamp would be costly.
Testing a few yesterday, it seems they cannot be reliably inter-changed
between types.
Why wouldn't a 36w linear ballast work with a 28w 2D lamp, and vice
versa?
OK, rating at 8w over rating may be asking too much, but why not run a
28w off a 36w ballast?
And a 2 x 36W works with one lamp lit, the other not working, but
wouldnt work with a single, good, lamp fitted?

A fluorescent tube requires the correct power, or close to it, the right V & i which depend on tube width, length and design features, and also the correct heater power, v and i. Hence ballasts that will by design run more than one tube type are a minority.

A 36w ballast on a 28w lamp would overrun the lamp, reducing life expectancy.

ITYM it might.
It depends on the tube.
eg. you can get different length 28W T5 tubes for aquaria, so they
require different ballasts to give 28W.
The effective resistance of the tube will vary according to length,
cross sectional area and the gas mix.


Its certainly possible to do some ballast/tube type swapping. Iron ballasts are easily adapted to run lower power tubes by adding the right capacitor, but the calculation of its value is nontrivial and correct calculation is essential. CFL ballasts often run small fl tubes happily enough. In principle modern electronic ballasts can run lower power tubes with 2 capacitors on the output, but its too far from trivial to be practical.

Perhaps someone should make a ballast with multiple settings.

Its probably possible to design a ballast that has voltage and current
sensors to measure the actual power delivered and adjust accordingly.
It wouldn't sell unless it only cost an extra few pence.

In article ,
(A.Lee) writes:
So, an electronic ballast has a rating, as well as, in some cases a lamp
type. OK, thats simple. But then, there are so many lamp types and
wattages, that keeping a stock of ballasts in the van for each type of
lamp would be costly.
Testing a few yesterday, it seems they cannot be reliably inter-changed
between types.
Why wouldn't a 36w linear ballast work with a 28w 2D lamp, and vice
versa?

A filament lamp is designed to run at a partcular power rating
when connected to specific voltage, and this is governed by its
operating resistance. To run it at the designed power rating,
you feed it a constant voltage, and it will decide what current
it's going to draw.

A fluorescent lamp is different in two respects. Firstly, they
have to be run at a constant current in order to run at the
designed power rating. Secondly, the resistance of a fluorescent
tube decreases dramatically as the current increases, and in
effect, the tube decides what voltage it will operate at.
Given that our mains supply (and battery supplies) are effectively
constant voltage supplies, they cannot be connected directly across
a fluorscent tube, because as the current increased, the tube
resistance would drop, and the current would increase, and the
tube resistance would drop further, until something goes bang.
Consequently, a tube requires something to limit the current.

OK, rating at 8w over rating may be asking too much, but why not run a
28w off a 36w ballast?

Originally, each ballast was designed to run just a single tube,
or possibly a small number of tubes with similar operating
characteristics. This required a large number of different ballasts
to be manufactured and stocked, which was not a good thing.

There are families of tubes which all operate at the same current
and the same loading (watts per unit length) and vary in power
rating proportionally to the tube length (and hence operating
voltage). If you build a ballast which generates a truely constant
current, you can run any of these because the tube will decide on
its operating voltage which will give the correct rated power output,
and that's what's done today. Thus you can get a ballast which will
run all tubes in one family because they operate at the same current.

As ballasts have gone digital (controlled by micro controllers),
they have been able to get even more sophisticated by looking at the
characteristics of the lamp, and working out which family its from,
and can thus drive different lamps from different families, adjusting
the current appropriately. As an example, I just grabbed one off the
shelf and it can drive:
All the T5HE family lamps (14W, 21W, 28W, 35W)
The longest T5H0 family lamps (49W, 80W)
DL55W, DL80W (folded 55W and 80W tubes)
FC55W (Circular 55W T5 tube)

My guess is that it initially drives the tube at the tube current of
170mA for T5HE, and checks the operating voltage. If it's not what's
expected of the T5HE lamps, it will then increase the current to that
of the T5HO lamps (don't know what that is off-hand), and possibly
some more checks for the other 3 if they have different operating
currents.

This probably replaced 9 separate ballasts in the catalogues.

However, back to your 28W and 36W examples, they may have quite
different operating characteristics in terms of tube current and
voltage. Even if they did both work, the lamp and/or ballast may
be operating well outside it's designed rating.

And a 2 x 36W works with one lamp lit, the other not working, but
wouldnt work with a single, good, lamp fitted?

Microcontroller ballasts check for open circuit filaments before
trying to start the tubes, so they don't start one with worn out
filaments and end up running it in cold cathode mode, which will
overheat the tube ends.

The ability to operate less than the full compliment of tubes varies
by ballast - most commonly they operate them all in series, so this
usually won't work. Some 3 and 4 lamp ballasts have two series chains.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

And there do seem to be what I'd call cold cathode starters that appear not
to use the heaters at all, and just hike up the voltage till it breaks down
and strikes, then reduce it to the nominal level again. I have even got one
of those to run two small 8 watt tubes with the heaters of on ent connected
together so it has to strike in both to operate. Good fun, but now I can't
see its all a bit academic, and probably dangerous to do such things. Note
I'm not advocating this mode, just saying it can be done.
Brian

--
Brian Gaff....Note, this account does not accept Bcc: email.
graphics are great, but the blind can't hear them
Email:
__________________________________________________ __________________________________________________ __________


wrote in message
...
On Sunday, December 23, 2012 9:15:57 AM UTC, A.Lee wrote:

So, an electronic ballast has a rating, as well as, in some cases a lamp
type. OK, thats simple. But then, there are so many lamp types and
wattages, that keeping a stock of ballasts in the van for each type of
lamp would be costly.
Testing a few yesterday, it seems they cannot be reliably inter-changed
between types.
Why wouldn't a 36w linear ballast work with a 28w 2D lamp, and vice
versa?
OK, rating at 8w over rating may be asking too much, but why not run a
28w off a 36w ballast?
And a 2 x 36W works with one lamp lit, the other not working, but
wouldnt work with a single, good, lamp fitted?

A fluorescent tube requires the correct power, or close to it, the right V &
i which depend on tube width, length and design features, and also the
correct heater power, v and i. Hence ballasts that will by design run more
than one tube type are a minority.

A 36w ballast on a 28w lamp would overrun the lamp, reducing life
expectancy.

Its certainly possible to do some ballast/tube type swapping. Iron ballasts
are easily adapted to run lower power tubes by adding the right capacitor,
but the calculation of its value is nontrivial and correct calculation is
essential. CFL ballasts often run small fl tubes happily enough. In
principle modern electronic ballasts can run lower power tubes with 2
capacitors on the output, but its too far from trivial to be practical.

Perhaps someone should make a ballast with multiple settings.


NT

In article ,
Tim Streater writes:
So if I have a tube that, when power is supplied, flashes a bit with a
glow at both ends and then eventually comes on, is that the tube going,
the ballast, or the starter?

That sounds like a switch-start system which is working fine!

If it takes a long time to come on, it could be:
Starter failing,
Wrong starter,
A long tube which is cold.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

(Andrew Gabriel) wrote in
:

In article ,
(A.Lee) writes:
So, an electronic ballast has a rating, as well as, in
some cases a lamp type. OK, thats simple. But then, there
are so many lamp types and wattages, that keeping a stock
of ballasts in the van for each type of lamp would be
costly. Testing a few yesterday, it seems they cannot be
reliably inter-changed between types.
Why wouldn't a 36w linear ballast work with a 28w 2D lamp,
and vice versa?

A filament lamp is designed to run at a partcular power
rating when connected to specific voltage, and this is
governed by its operating resistance. To run it at the
designed power rating, you feed it a constant voltage, and
it will decide what current it's going to draw.

A fluorescent lamp is different in two respects. Firstly,
they have to be run at a constant current in order to run
at the designed power rating. Secondly, the resistance of a
fluorescent tube decreases dramatically as the current
increases, and in effect, the tube decides what voltage it
will operate at. Given that our mains supply (and battery
supplies) are effectively constant voltage supplies, they
cannot be connected directly across a fluorscent tube,
because as the current increased, the tube resistance would
drop, and the current would increase, and the tube
resistance would drop further, until something goes bang.
Consequently, a tube requires something to limit the
current.

OK, rating at 8w over rating may be asking too much, but
why not run a 28w off a 36w ballast?

Originally, each ballast was designed to run just a single
tube, or possibly a small number of tubes with similar
operating characteristics. This required a large number of
different ballasts to be manufactured and stocked, which
was not a good thing.

There are families of tubes which all operate at the same
current and the same loading (watts per unit length) and
vary in power rating proportionally to the tube length (and
hence operating voltage). If you build a ballast which
generates a truely constant current, you can run any of
these because the tube will decide on its operating voltage
which will give the correct rated power output, and that's
what's done today. Thus you can get a ballast which will
run all tubes in one family because they operate at the
same current.

As ballasts have gone digital (controlled by micro
controllers), they have been able to get even more
sophisticated by looking at the characteristics of the
lamp, and working out which family its from, and can thus
drive different lamps from different families, adjusting
the current appropriately. As an example, I just grabbed
one off the shelf and it can drive:
All the T5HE family lamps (14W, 21W, 28W, 35W)
The longest T5H0 family lamps (49W, 80W)
DL55W, DL80W (folded 55W and 80W tubes)
FC55W (Circular 55W T5 tube)

My guess is that it initially drives the tube at the tube
current of 170mA for T5HE, and checks the operating
voltage. If it's not what's expected of the T5HE lamps, it
will then increase the current to that of the T5HO lamps
(don't know what that is off-hand), and possibly some more
checks for the other 3 if they have different operating
currents.

This probably replaced 9 separate ballasts in the
catalogues.

However, back to your 28W and 36W examples, they may have
quite different operating characteristics in terms of tube
current and voltage. Even if they did both work, the lamp
and/or ballast may be operating well outside it's designed
rating.

And a 2 x 36W works with one lamp lit, the other not
working, but wouldnt work with a single, good, lamp
fitted?

Microcontroller ballasts check for open circuit filaments
before trying to start the tubes, so they don't start one
with worn out filaments and end up running it in cold
cathode mode, which will overheat the tube ends.

The ability to operate less than the full compliment of
tubes varies by ballast - most commonly they operate them
all in series, so this usually won't work. Some 3 and 4
lamp ballasts have two series chains.


Thank you.

A most informative and educational post.


--- news://freenews.netfront.net/ - complaints: ---

Andrew Gabriel wrote:

In article ,
(A.Lee) writes:
So, an electronic ballast has a rating, as well as, in some cases a lamp
type. OK, thats simple. But then, there are so many lamp types and
wattages, that keeping a stock of ballasts in the van for each type of
lamp would be costly.
Testing a few yesterday, it seems they cannot be reliably inter-changed
between types.

A filament lamp is designed to run at a partcular power rating
when connected to specific voltage, and this is governed by its
operating resistance. To run it at the designed power rating,
you feed it a constant voltage, and it will decide what current
it's going to draw.

snip. Cheers, confirming my experimentation that swapping around the
cheap ones supplied with light fittings is a waste of time, and I need
to buy the correct ones.
Ta
Alan.

--
To reply by e-mail, change the ' + ' to 'plus'.

On Sunday, December 23, 2012 11:42:09 AM UTC, Brian Gaff wrote:

And there do seem to be what I'd call cold cathode starters that appear not
to use the heaters at all, and just hike up the voltage till it breaks down

If you're into odd ways to light a tube, try rubbing it with pads that produce static electricity. Or for something more sensible, drive it with high frequency via bits of foil on the glass, the capacitance acts as the current limiting element. This can make most dead tubes work fine.


NT

On 23.12.2012 12:34, Andrew Gabriel wrote:
In ,
(A.Lee) writes:
So, an electronic ballast has a rating, as well as, in some cases
a lamp type. OK, thats simple. But then, there are so many lamp
types and wattages, that keeping a stock of ballasts in the van
for each type of lamp would be costly. Testing a few yesterday,
it seems they cannot be reliably inter-changed between types. Why
wouldn't a 36w linear ballast work with a 28w 2D lamp, and vice
versa?
A filament lamp is designed to run at a partcular power rating when
connected to specific voltage, and this is governed by its operating
resistance. To run it at the designed power rating, you feed it a
constant voltage, and it will decide what current it's going to
draw.

Today I used a filament lamp for testing of my multimeter and became a
bit desperate when I found that its resistance did not agree with Ohms
Law. Its reading was not only wrong, it was wrong by 800 %.

What do you do when a physical law like Ohms Law fails?
I tried some Google searches and calmed down when I found this:
http://wiki.answers.com/Q/Why_the_fi...ollow_Ohms_law
A filament lamp does not obey Ohm's Law because the ratio of voltage
to current isn't a constant over a range of voltage increments. This
is because the resistance of metals such as tungsten increase
significantly as its temperature rises to its operating temperature.

There seems to be a major misconception surrounding Ohm's Law. Ohm's
Law simply states that 'the current flowing along a wire, at constant
temperature, is directly proportional to the potential difference
across its ends'.

The equation R = V/I applies whether Ohm's Law is followed or not,
because the ratio of V/I will always indicate what the resistance
happens to be for that particular ratio. But the fact that you can
use this equation under any circumstance doesn't mean Ohm's Law is
applicable.
--
jo
"Anti-intellectualism has been a constant thread winding its
way through our political and cultural life, nurtured by the
false notion that democracy means that 'my ignorance is just
as good as your knowledge.'" -- Isaac Asimov

On Fri, 28 Dec 2012 20:59:56 +0100, Jo Stein
wrote:

What do you do when a physical law like Ohms Law fails?

Are you actually a monkey in an AI lab?

On 30/12/12 12:15, Grimly Curmudgeon wrote:
On Fri, 28 Dec 2012 20:59:56 +0100, Jo Stein
wrote:

What do you do when a physical law like Ohms Law fails?

Are you actually a monkey in an AI lab?

No, he's a norwegian.


--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On 30/12/2012 12:15, Grimly Curmudgeon wrote:
On Fri, 28 Dec 2012 20:59:56 +0100, Jo Stein
wrote:

What do you do when a physical law like Ohms Law fails?

Are you actually a monkey in an AI lab?

Artificial insemination? :-)

Jo - there is no such thing as a failed physical law. If it does not
work, it is no longer a law.

--
Rod

On 30/12/12 12:30, polygonum wrote:
On 30/12/2012 12:15, Grimly Curmudgeon wrote:
On Fri, 28 Dec 2012 20:59:56 +0100, Jo Stein
wrote:

What do you do when a physical law like Ohms Law fails?

Are you actually a monkey in an AI lab?

Artificial insemination? :-)

Jo - there is no such thing as a failed physical law. If it does not
work, it is no longer a law.

Ohms law is in any case more a rule of thumb for metallic conductors.

It never applied to gas discharge tubes or to semiconductors.

Or rather you can say that there are things whose resistance depends on
current and on applied voltage. And in the case of valves and
semiconductors, what's happening elsewhere in the device.

That's the trouble with these 'know a very little knowitalls'.

They stumble on something a bit more than basic and think they have won
the Nobel Prize. They always believe in global warming, too.


--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On 30.12.2012 13:34, The Natural Philosopher wrote:
....
Ohms law is in any case more a rule of thumb for metallic
conductors.

It never applied to gas discharge tubes or to semiconductors.

Or rather you can say that there are things whose resistance depends
on current and on applied voltage. And in the case of valves and
semiconductors, what's happening elsewhere in the device.

That's the trouble with these 'know a very little knowitalls'.

They stumble on something a bit more than basic and think they have
won the Nobel Prize. They always believe in global warming, too.

I believe in global warming because I was a very clever student.
When I graduated from my university my qualifications in physics
could not be measured. I broke the scale an got 1.0.

I told you about the misery of Ivar Giæver in an earlier message.
It is repeated below because more research is needed to explain
his failure.

One of the first things we were told at university was to remember
that professors are only human and that we should not assume they
are right, simply because they are professors.

None of my professors said that. They all knew that they were at the
top. I am at the top without being a professor.

Accidents can happen. When the The Nobel Prize winner Ivar Giæver
graduated from my university, he got the grade 4.0 i math and
physics. 1.0 was top and 6.0 was bottom. He became a professor by
emigrating to the US where 4.0 was the top.

In the long run the system has a self cleaning property. Quality
floats to the top and stay there, garbage sinks to the bottom and
stay there. Today Ivar Giæver stays at the bottom as a climate
skeptic: http://en.wikipedia.org/wiki/Ivar_Giaever
On 13 September 2011 Giaever resigned from the American Physical
Society over its official position that "the evidence is
incontrovertible."

More research is also needed to explain the failure of
The Natural Philosopher.

Now back to Ohms law:
http://en.wikipedia.org/wiki/Ohm%27s...rature_effects
Ohm's law has sometimes been stated as, "for a conductor in a given
state, the electromotive force is proportional to the current
produced." That is, that the resistance, the ratio of the applied
electromotive force (or voltage) to the current, "does not vary with
the current strength ." The qualifier "in a given state" is usually
interpreted as meaning "at a constant temperature," since the
resistivity of materials is usually temperature dependent. Because
the conduction of current is related to Joule heating of the
conducting body, according to Joule's first law, the temperature of a
conducting body may change when it carries a current. The dependence
of resistance on temperature therefore makes resistance depend upon
the current in a typical experimental setup, making the law in this
form difficult to directly verify. Maxwell and others worked out
several methods to test the law experimentally in 1876, controlling
for heating effects.

Because of the excellent work of Ohm, Maxwell, Heaviside,
Croll, Arrhenius, Milankovitch and others, our world become
very predictable for those that have big brains.
--
jo
€œThere is a tension between short-term, individual welfare
and long-term, group welfare or world welfare.
If it were left to Darwinism alone, there could be no hope.
Short-term greed is bound to win. The only hope lies in
the unique human capacity to use our big brains with our
massive communal database and our forward-simulating imaginations.€�
€”From Dr. Dawkins' acceptance speech at the 2001 Kistler Prize Banquet