Replace old fluorescent tube with brighter?

[Andy Wade]

Andrew Gabriel wrote:

[UK]

Commercially, T12's haven't been used for a long time except for
special applications.

Except of course for the 8 ft. length (now 100 W), which are still
fairly common. I've always assumed this is for reasons of mechanical
robustness - an 8 ft. T8 tube would be just too fragile - is that right?

 

[Andrew Gabriel]

Electronic components are getting less expensive and sturdier, more
robust. Circuit designs are improved upon those that failed the test of
time.

Bigger iron core items of older technology have not enjoyed as much
reduction in production cost.

I think there are significant differences between US and Europe here
which dramatically change the economies of electronic ballasts between
the continents.

In Europe...
o Electronic ballasts are still mostly very expensive, and special
order, and very limited selection.
(As someone who has made a number of fluorescent fittings, I wish
this wasn't the case, but it is.)
o Iron core ballasts are very cheap (possibly much cheaper than in
the US as ballasts for 230V mains operation are very much simpler).
o My impression from reading US newsgroups is that iron core ballasts
in europe are very much more reliable -- failures are pretty much
unknown, but seem more common in the US (possibly in part because
it's easier to use the wrong lamp with the wrong ballast, and
possibly because they are more complex due to low mains voltage).
Electronic ballasts just can't touch Iron core ballasts for
reliability, as Iron core ballasts effectively last forever here.
o Iron core ballasts in Europe are getting much more efficient due to
EU regs (regs which were thought to phase them out, but just forced
manufacturers to make them higher efficiency as that is still much
cheaper that switching to electronic ballasts). They probably were
in any case always more efficient than US ballasts, because they
don't need the extra complexity for 120V operation.
o Europe (at least the parts I know) moved away from T12 tubes decades
ago, so the savings you are talking about with respect to switching
to T8's we already did 25 years ago.

Also, all popular F32T8 lamps made for "general lighting purpose" are

This is a USA-only tube. See my other post.

 

[Clive Mitchell]

1. high speed sharp bladed food processors that always have interlocks.
Interlocking prevents any hand contact with moving blades. There is
also the fact that in use one does not see spinning blades, but rather
a mass of moving food. And the food does not move fast enough to apear
stationary

Not all.

2. low speed blunt paddle devices that are open. These move so slowly
it is not possible to encounter a strobing problem with fl lighting on
50 or 60Hz. The blades move such a small amount between each half cycle
that this simply can not be a problem.

So there is no risk at all in using mag ballasted fl lighting with cake
mixers.

Oh I rather think not. The CakeMaster 3000 has a variable speed DC drive
with digital speed feedback and clearly has blade settings at 100 and
120 Hz. The motor is rated 3HP and the blades are barbed to facilitate
good reduction of cake batter without lumps.

Several wealthy old ladies have quite literally been pureed to death in
these kitchen monsters. In one instance the only clue to what had
happened was that the mixing bowl was fuller than normal and had a
fluffy slipper perched on it's edge.

 

[Victor Roberts]

This is correct, but it is not the relevant fact. What is relevant is
that the market is more cost driven and aggressive than in the past,
and many electronic goods that were once reliable have become much less
reliable, some to the point of disposable. This is primarily
competition driven. Electronic reliability depends on design margins,
and design margins depend on money. hence, not surprisingly, cheap
electronic goods are not known for reliability.

To replace a mag ballast with en electronic one and expect greater
reliability would be optimism over fact.

I don't think anyone has claimed that electronic ballasts
are more reliable than EM ballasts. After all, EM ballast
have only a few parts and perhaps 20 connections, while
electronic ballasts have perhaps 30 to 50 components and
perhaps 70 to 100 connections.

EM ballasts have lives of well over 20 years, at least in
the US, and it is not necessary for electronic ballasts to
have better reliability than EM ballasts in order for them
to be a good idea. If the life of an electronic ballast is
20 years, the room will probably be renovated before the
ballasts die.

I can't quite figure out why you believe that electronic
ballasts have serious reliability issues. It sounds like
you are stuck in 1980, a few years after the introduction of
electronic ballasts. I assume you own a TV set, some sort
of music system, perhaps a DVD player, own a car with an
electronic control system, travel in airplanes that use
electronic control systems, and, yes, that computer that you
are using to post your messages. I have not seem any
mechanical computers that have newsgroup readers, so I
assume you are using an electronic computer

In 2005 there were 61,269,000 electronic fluorescent lamp
ballasts sold in the US. See Current Industrial Reports -
Fluorescent Lamp Ballasts: 2001, available at
http://www.census.gov/cir/www/335/mq335c.html.

With this many ballast sold, we would have a lot of very
angry customers if the failure rate was not very low.

--
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.

 

[Paul M. Eldridge]

Just a short sidebar to this conversation. Our firm recently upgraded
the office and lab lighting for a local client -- approximately 650
fixtures were converted in this phase of the project.

Most of these fixtures are 35 to 40 years old and contained four 40 or
34-watt cool-white T12s driven by two magnetic ballasts. Nameplate
wattage came in at a whopping 186-watts. Each fixture is now equipped
with two 28-watt T8s driven by a single electronic ballast and power
consumption has fallen to just 52-watts. We selected Osram Sylvania
850/XP/ECO lamps for this particular project.

What's amazing to us is the extremely positive feedback we've received
from management and staff. Everyone is thrilled with the results and
although actual light levels are lower, many folks now joke they need
to wear sunglasses to do their work. Undoubtedly, the move from a
4,100K/62 CRI light source to a 5,000K/85 CRI lamp is largely
responsible for this (by comparison, the old lamps looked dull and
yellow and the environment took on a somewhat dingy appearance).

In addition to significant energy savings, the T8/electronic ballast
combination offers other benefits, including quieter operation,
elimination of lamp flicker (an important consideration for CRT use),
reduced air conditioning and transformer loading, extended lamp life,
better lumen maintenance, lower maintenance costs/fewer disruptions to
operations, etc. It also permitted the removal and proper disposal of
older (and badly leaking) ballasts, many of which contained PCBs.

Cheers,
Paul

 

[Victor Roberts]

In 2005 there were 61,269,000 electronic fluorescent lamp
ballasts sold in the US. See Current Industrial Reports -
Fluorescent Lamp Ballasts: 2001, available at
http://www.census.gov/cir/www/335/mq335c.html.

This last should have read "2005" not "2001".

--
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.

 

[Don Klipstein]

Replace a 40w tube with a 36w T8 and the power use also falls, though
not as much as that, due to mag ballast losses.

However, that is not the real world comparison. The real world energy
comparison also involves:
- energy used in transport to go get a new fitting

To toss some numbers, suppose 100 units transported 25 miles each way in
a truck that gets 8 MPG. (Really long trips would more likely involve a
whole truckload of units.)
One unit's share of the truck mileage is .6 mile, and at 8 MPG that's
..075 gallon of gasoline. If I calculated right, that's roughly 2.6
kilowatt hours. A 25 watt power consumption only requires about 104
operating hours to make that up.

- energy used in manufacture of new fittings
- energy used in parts/materials of new fitting

Somehow I suspect the factory's price for decent ones in quantities of
100,000's has to be less than the retail price of a cheap garbage grade
dual-4-foot shoplight with a garbage grade magnetic ballast - and I have
heard of about $10 for those.
So, I don't think I'm badly out of the ballpark to pull out of a hat $8
for a 2-lamp 4-foot unit, FOB the factory's loading dock in truckload
quantities. Just for the sake of argument, suppose 100% of that cost was
for energy to obtain materials and to manufacture the unit - starting from
dirt/rocks/air/water that is. Suppose 8 cents per kilowatt hour. That's
1,000 KWH, which a 25 watt savings will make up in 40,000 operating hours,
which most 4-foot fixtures will accumulate in a decade or two.

Somehow I think that figure is high by at least an order of magnitude.

- energy used in disposal of old fitting

Weighing .005-.01 ton? How much diesel fuel to truck round-trip 20 tons
of trash to a landfill 100 miles away - 40 gallons or less? .01-.02
gallon per unit - requiring maybe 15-30 operating hours for a 25 watt
power consumption decrease to make up?

- energy used in manufacturing, supplying and applying a coat of paint
to the ceiling when end user notices the new fitting is not identically
sized to the old, and the resulting paint appearance is bad.

Usually not the case - sizes of fluoresceint lighting units are
well-standardized. This is especially true for the ones to fit into the
drop ceilings that most offices have nowadays, where a fixture replacement
will not require repainting - especially not of acoustic tile!

Meanwhile, the share ceiling space associated with one luminaire
requires how much paint? Something like half a dollar to a dollar's
worth? Even if 100% of the cost of the paint is from energy consumption,
that would be in the ballpark of half a dollar to a dollar's worth of
electricity. At 8 cents per KWH, a power reduction of 25 watts accounts
for half a dollar to a dollar in 250-500 operating hours.
As for applying the paint in the unlikely event that is required - burn
50 calories or 12 cents worth of potato chips?

If you do a real world energy comparison, it is more than hard to
justify replacing the fitting on energy saving grounds.

Don Klipstein ([email protected])

 

[Victor Roberts]

On 3 Dec 2006 04:51:54 -0800, [email protected] wrote:

[snip]

If you do a real world energy comparison, it is more than hard to
justify replacing the fitting on energy saving grounds.

No product can cost less than the cost of the energy used to
produce that product and all it component parts, including
all the energy needed to extract and transport the raw
materials. This is basic economics.

The value of the energy saved is far greater than the cost
of the new lamps, ballasts and luminaires. (If it were not,
people would never make the switch.) Even if the user of
the luminaire pays a higher rate for energy than the
industrial users who built and transported the device, the
cost savings would tend to rule out your assertion.

--
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.

 

[Victor Roberts]

mag ballasts with 20 connections?? I've never seen anything like that
here, most have 2, some have 3. Sounds like you have some very exotic
ballasts over there.

Internal connections, not external. That is one measure of
mean time to failure.

UK mag ballasts have mean lives of over a century. AFAIK it has never
dropped below this.

How do you know this since fluorescent lamps have not been
around for 100 years? Even if this is correct, and I guess
it might be for a simple choke ballast, how many remain in
ceilings for 100 years?

When Philips first showed the QL at Hannover in 1992 or
perhaps 1993 they gave an estimated life of 60,000 hours
(at that time) and said it was the "life of the building"
since the space would be renovated before the lamp had
operated for 60K hours. I'm not sure I agree with 60K hours
life for a building space, but it is certainly not 100
years.

What I said was that their reliability does not compare to mag
ballasts. This is true. Thus reliability is not a reason to replace the
fitting.



lets keep it sensible now


I'm not sure why you're having difficulty accepting that one is more
reliable than the other. They just dont compare when it comes to
reliabilty. None of the above has anything to do with it.

I agreed with this in my opening remarks.

We have precisely that situation with a whole range of consumer goods
here. Many cheap end goods have appalling lifetimes, eg power tools
that die after 10 hrs use. Sellers count on replacing a percentage,
users buy them mostly in ignorance, and less often when they know but
are in no position to take the tool with them. There are large
quantities of angry customers, but far more that are willing to buy on
price alone, so business continues.

So, are you comparing the reliability of electronic ballasts
to cheap electric tools that fail in 10 hours? What is your
point in this comparison?

Its probably best policy to establish the facts before being
condescending.

How true.

--
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.

 

[Victor Roberts]

mag ballasts normally have 2 internal connections, one at each end of
the winding.

This is only true for simple reactor ballasts. 2-lamp EM
fluorescent lamp ballasts used in the US have multiple
internal windings and two capacitors.

this is basic stuff. If for example 50% had died after 50 years, we'd
know mean life was apx 50 years. If 10% died after 50 years we'd know
mean life was nearer 500, etc. Lots of items are life rated at beyond
the time theyve existed on this basis, standard industry practice.

You must know this is only true for a linear mortality
curve. Please take a look at a typical mortality curve and
you will see that it is far from linear. The failure rate
for times much earlier than mean life cannot predict the
mean life point.

--
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.

 

[Andrew Gabriel]

This is only true for simple reactor ballasts.

That's what they mostly are on 230V supplies.

2-lamp EM
fluorescent lamp ballasts used in the US have multiple
internal windings and two capacitors.

2-lamp EM fluorescent lamp ballasts commonly used in
offices here are just one ballast and two tubes in series,
so that drops to just a single internal connection per tube ;-)
(It also makes the ballast more than twice as efficient
as a pair of separate ballasts driving separate tubes.)

This is partly what I was getting at in an earlier post where
I said the economics just seem quite different between US/120V
and EU/230V control gear systems.

You must know this is only true for a linear mortality
curve. Please take a look at a typical mortality curve and
you will see that it is far from linear. The failure rate
for times much earlier than mean life cannot predict the
mean life point.

I've never come across any simple series ballasts failing in
UK fluorescent fittings (except in one case where water came
through the ceiling). At this point, I think we can say their
life is longer than they've been around.

We have also had some more complex EM control gear designs,
but I've not come across enough to comment on their
reliability as they're relatively rare.

 

[Andrew Gabriel]

I recently noticed that there are a few of these still in use where I
work; two of them right outside my office. Most of them have had the
ofiginal lampholders replaced, but it least one, in the sparks'
workshop, still has the original ones, with bipin adaptors. In a
little-used storeroom there are two that still have the original covers
over the lampholders and clips. I'd like to acquire one of these when
they are taken down. The whole building is being gutted and
refurbished over the next few years, so they won't be there much
longer. There was one leaning against the wall by the boiler house
door recently, together with the box that its replacement came in. I
looked at it, but the ballast was burned out. I had a word with the
sparks, and he let me have the two large four-pin starters out of it.

Unless it's a very old thermal four-pin starter, you'll find
that inside is just a regular glow starter across the locking
pins, and the small pins are simply shorted together. It's
most unlikely an original thermal starter would still be
working unless it was in some situation where it wasn't used
for 50 years.

I did notice "The Shop on the Bridge" in Reading had some
bayonet cap fluorescent tubes on sale when I was last in there
about 2 years ago. Knowing them, they might well be 1950's
stock;-) Bayonet cap to bi-pin adaptors were very common 40
years ago when the bayonet cap tubes were phased out, but it's
probably easier to buy a flying bi-pin socket nowadays and swap
out the bayonet cap lampholder, if you want to run those lamps.

Of course, you could fit a modern ballast if you just want the
fitting for 1950's styling.

 

[Andrew Gabriel]

Where is this shop; I'd like to get hold of a bayonet tube, or two?

http://www.streetmap.co.uk/newmap.srf?x=468588&y=174158

I suggest you phone before travelling, to make sure they
have one in the right size (supposing they still have any).

I'd like to get gold of one of the ones which still have the bayonet
holders if possible, or failing that, one of the converted ones, and
convert it back. I really want one with the original control gear. I
did find one bipin adapter on the floor in one of the lift machine
rooms, but I couldn't find another one. It has a small 'tab' on one

I can see the tab in a picture. The Thorn part number is GB1515.

side, which I assume is intended to make contact with the metlic
stripe, on tubes that had it, MCFA I think was the designation. The

Not sure, as you won't need MCFA tubes in these metal fittings with
switchstart control gear. They were for Quickstart and Rapid start
control gear where the tube was not near earthed metalwork. I don't
think these starterless control gear types appeared until after BC
fittings had gone, but I might be wrong.

lifts were replaced about 18 months ago; the original ones lasted about
50 years; I would be surprised if the new ones manage 20, Two of the
old ones had beed modernised some years ago, but the third still had
its dc drive and motor-generator set until it was replaced. The
machine room has modern fluorescent fittings, probably installed at the
time the lifts were replaced.

By the way, the history of the bayonet cap fluorescent tubes is
interesting. The first fluorescent tube installation in the UK
was in one of the London Underground stations around 1937 IIRC.
During WWII, there was a move to switch over to more fluorescent
lighting, but the lampholders which came from the US were no
longer available (I don't know if they were bi-pin at that point).
Ever resourceful, they decided that if they were swapping out
filament lamps, they could reuse the lampholders if the fluorescents
used the same type, so fluorescents switched over to BC lampholders.
The 4' tube (which was the original size used in the US at that point)
was abandoned in the UK during WWII, and 5', 6' and 8' tubes produced
because installations of these used fewer raw materials. The 4' tube
didn't reappear in the UK for some time after the war (I guess mid
1950's, but I don't know for sure) and had bi-pin lampholders.
The other tube sizes then changed to bi-pin too.