Last week I finished the
required drainage and stormwater handling of the new workshop. With the council
inspection completed, it was time to move on to equipping the new workshop with
adequate light.
Natural Lighting
During construction of the workshop, the 14 x 6
metre roof was equipped with two translucent panels. These panels are the same
size and shape as the normal roofing sheets, so their installation is
straightforward.
In most weather conditions, each panel provides a
very large amount of light. So why only two panels?
In sunny south-east Queensland (apparently we have
300 days of sunshine a year!) I was concerned that using more panels would cause
the workshop to get hot. A smaller shed that I previously had in South Australia
was at times uncomfortably hot when standing beneath either of the two
translucent panels. So in this larger workshop I figured that using (say) four
panels would be too many – they’d provide lots of light but also plenty of heat.
The two translucent panels were positioned on the
northern aspect of the roof and towards the end of the shed in which the machine
tools and workbench will be located. These are at the opposite end of the
workshop to the twin roller doors, so with the roller doors up, on overcast days
the natural light through the shed can be fairly even. However, on sunny days
the natural illumination is skewed much more to the ‘work’ end of the shed.
Artificial Lighting
One of the most important aspects in a workshop is
to have good artificial lighting. A workshop with good lighting can be used at
night or in overcast conditions with little decrease in effectiveness. That
might seem obvious, but I’d suggest that the majority of workshops are way
under-illuminated – the lighting level apparently set for normal visibility
rather than precision work.
Different types of artificial lighting have
different characteristics. But before they can be described, you need to know
what those characteristics are.
Lighting Terms
Luminous intensity is measured in Candela (cd).
Luminous intensity is used to describe the amount
of light emitted in selected directions from lamps and fittings.
Luminous flux is measured in lumens, abbreviated
to ‘lm’.
Just as there is an electrical power input
measured in watts, there is a “light power” output measured in lumens. The
reason that “light power” is not measured in watts is because the response of
the eye to different colours needs to be taken into account.
To obtain a measure of the luminous flux of a
light, the radiant flux (measured in watts) is weighted by the frequency
response curve of the eye. Thus, a high power output light at a wavelength to
which the eye is less sensitive results in a lower luminous flux value.
Conversely, if the light emits a great deal of radiation at 555nm (where the eye
is most sensitive), its lumen rating will be high.
Luminous flux measurements are widely used in
lighting. A typical application is in expressing luminous efficacy. This is a
measurement of how much light output there is for a given electrical power
input. It is expressed in lumens/watt, abbreviated to lm/W.
Illuminance is measured in lux, abbreviated to lx.
Illuminance is a measurement of how many lumens
there are per square metre. There are recommended values of maintained
illuminance for various activities, with the table below showing some
International Commission on Illumination (CIE) suggestions.
Normal handheld light meters measure in lux.
Location |
Illuminance
(lux) |
Toilets |
100 |
Instrument assembly |
1500 |
Garment manufacture - sewing |
750 |
School classrooms |
500 |
Cinema auditorium |
50 |
Kitchen work areas |
500 |
Hospital ward at night |
1 |
Operating theatre (local lighting) |
100,000 |
Supermarket |
750 |
On a flat surface where there are few reflections,
illuminance can be easily plotted using lines of equal illuminance. These lines
are called isolux contours. Basically, it is a diagram of the “pool of light”
found beneath outside street lights - the one so beloved of writers of detective
fiction!
Such a diagram is useful when designing the
lighting system of a car park, or in working out the spacing of lights to
provide even illumination in a large building.
An object at any temperature will emit radiation.
At low temperatures the wavelengths of the radiation are mostly in the infra-red
region and so cannot be seen. However, if the temperature is increased, the
object (for example, a piece of steel) will start to glow. It is then emitting
radiation that can be seen. The temperature of the object can be measured in
Kelvin (K), which is its temperature in degrees Celsius plus 273.15.
The radiation properties of a hypothetical
so-called black body radiator mean that at 1000K it will be red, at near 3000K
it will be yellow, at near 5000K white, blue-ish white near 10,000K and pale
blue near 30,000K. This means that the colour of a light source can be specified
in terms of its colour temperature - the colour that a blackbody radiator would
be if heated to the specified temperature.
Electric lights have widely varying colour
temperatures, but because your eyes are very tolerant of differing colour
temperatures, the perceived colour of different light sources varies relatively
little. Daylight has a colour temperature of about 5500K, while an incandescent
light bulb is around 2800K. Fluorescent tubes are available with colour
temperatures from 2900-6500K.
Colour rendering refers to the appearance of an
object when it is illuminated by the light source under consideration. Light
sources of similar colour temperature can have completely different wavelength
compositions and so can provide great differences in colour rendering.
A low pressure sodium lamp produces light at just
a single wavelength and so the lamp reveals only that colour. An incandescent
lamp has an output that covers all wavelengths fairly evenly - although there is
an emphasis on red. A high pressure mercury vapour lamp has a mixture of some
‘lines’ (high outputs at specific wavelengths) mixed with a continuous
background spectrum and a band of energy at the red end.
Of these light sources, the incandescent lamp
gives the best colour rendering, followed by the high pressure mercury lamp and
then the low pressure sodium lamp.
Colour rendering is measured on a colour rendering
index (expressed as Ra) scale of 1-100, where 100 provides the best colour
rendering. The Ra scale for a lamp is based on the illuminated appearance of
fourteen different colour chips.
The colour rendering of incandescent lights is
very good at 99Ra, while fluorescent lights vary from 85-90Ra.
When assessing how items will look under different
lighting, colour rendering is much more important than colour temperature.
|
Selecting Lights
If you’ve absorbed what’s been described above,
you’ll know that what you want are:
-
lights with a high luminous efficacy (ie lots of
lumens output per electric watt input) and...
-
good colour rendering (so there’s no colour
cast)...
-
mounted in a luminaire that provides high luminous
intensity (ie directs the light appropriately)...
-
the upshot being an achieved illuminance of around
(say) 500 lux at bench level
Let’s start off with efficacy – how many
lumens/watt are produced. Remember, this figure takes into account the
sensitivity of the eye to different wavelengths, so it’s a ‘real world’ figure.
This is the key figure when assessing electrical running costs – higher efficacy
equals brighter lights for the same power cost.
Here are the luminous efficacies for a range of
lighting sources:
Light Source |
Efficacy (lm/w) |
Incandescent lamps |
10-15 |
Halogen incandescent lamps |
15-30 |
Fluorescent lamps* |
40-55 |
Mercury lamps* |
35-105 |
Metal halide lamps* |
60-100 |
High-pressure sodium lamps* |
100-140 |
* The above values for discharge lamps do not
include the effect of the ballasts
that must be used with those lamps. Taking ballast losses into account reduces
total efficacies typically by 10-20%, depending upon the type of ballast used.
From this list a few points can be immediately
made. Except for machine-specific illumination (eg a work light on a lathe),
incandescent (conventional light bulbs) and halogen incandescent (eg 12V MR16)
lights are out.
That leaves fluorescent, mercury, metal halide and
high pressure sodium. But what are the colour rendering properties of these
light sources?
Low pressure sodium (not shown in the above table)
has very poor colour rendering – the reason people look so weird under yellow
sodium street lights. High pressure sodium is better than low pressure sodium.
Metal halide and fluorescent lights have a colour rendering index that varies,
depending on the bulb in question, from accurate (Ra 90 -100) to moderate (Ra 60
– 80).
The most widely available of these light sources
is fluorescent. Able to be bought in both compact types (the ballast and starter
are built in) and conventional linear types (long tubes that are inserted in a
fitting that contains the ballast and starter), fluorescent lights and fittings
can be easily sourced.
Fluorescent sounds good, but in the case of my 14
x 6 metre workshop, I figured something like 14 double battens would be needed –
perhaps $500 for the cheapest new fittings, and $750 for ‘brand name’ lights and
fittings. Sometimes fluorescent battens come up secondhand, but usually in twos
and threes – not fourteen fittings of the same type. I was also a little
unconvinced as to how bright the resulting lighting system would be.
But what about compact fluorescent lights? These
are now available in up to 48 watt ratings, this wattage claimed to be
equivalent of a 240W incandescent light. With a price of about $27 each, at the
(guessed) requirement for about six lights, that would be only about $175,
including the standard bayonet sockets. Hmmm.
Finally, discharge lights like sodium and metal
halide are easy to find secondhand – people seem worried by the complexity and
tend to get rid of them without asking for a lot of money. This means it’s not
at all hard to acquire these lights – in fact, I had previously bought 15 of
them for just $30!
Testing
Here’s where the theory hits the reality – and it
starts to get really interesting. When assessing any lighting system, you must
use a lux meter. That might seem like an extravagance, but these digital meters
are now available very cheaply on eBay (AUD$25 including postage).
A lux meter shows quickly and easily the actual
illuminance that is occurring. When assessing the evenness and intensity of
illumination, it is vastly better than using your eyes alone.
My first test was with a high power compact
fluorescent. A 48W unit was purchased and hung in the workshop at an appropriate
height. Directly under it, the light meter recorded a disappointing 72 lux.
A large metal reflector was then placed around the
light. At the position at which the reflector could be easily mounted (ie the
bayonet socket screwed to it), the reading rose to 180 lux.
Then, with the reflector positioned for optimal
output (a position that would need a special bracket), the measured lux reading
rose to 260 lux. (This 3.6 times increase in output shows how important the
reflector is!)
However, even this peak reading was a lot less
than my desired 500 lux.
(All illuminance measurements made with the light
suspended 3 metres above the ground and the lux reading taken at bench
level.)
The next test was of one of the 15 second-hand
high intensity discharge lights that I’d bought as a job lot. Using a clear 150W
metal halide bulb and the standard reflector, the measured illuminance beneath
the light was an excellent 650 lux.
But here’s where it gets difficult. Of the 15 high
intensity discharge secondhand lights I had bought, only two proved to have
these metal halide bulbs. The others had what I assume to be high pressure
sodium (the control gear says it can handle 100 – 150W high pressure sodium and
70 – 150W metal halide) and these lights proved to be much dimmer – in fact, a
measured reading of only about 250 lux. Even with three of the lights
temporarily suspended, the overall illumination was disappointing.
Clearly, for best brightness, the metal halides
worked very well. So why not replace the high pressure sodium bulbs with metal
halides? That way, I’d get an estimated 650 lux right through the workshop –
good illumination indeed! But metal halide bulbs are expensive – typically
around AUD$80 each! However, some searching found them around $50 a bulb when
purchased by the dozen. But that’s still a high overall cost...
OK, time to take a step back.
High and Low Power Lighting
To get the illumination I wanted (at least 500 lux
throughout the workshop) I’d need the equivalent of twelve 150W metal halide
bulbs mounted in the high discharge light fittings I’d bought secondhand. That
required an expenditure of around $600 for the bulbs.
Having all the high power lights switched on would
take about 2.8kW – a load the equivalent to just under three one-bar electric
radiators.
But how often would I need the full 14 metre x 6
metre workshop so brightly illuminated? Not very often.
If I switched them in two groups of six, I could
selectively operate these lights. For example, when working down the end of the
workshop with the bench and machine tools, I could have six of the high power
lights running. If I do panel work on a car, I could have the six lights at the
other end of the workshop running.
OK, so viewing these lights as special ‘working’
lights reduces the effective running costs because they would normally be used
six at a time.
But if you take this approach, you then need some
other lights to provide general illumination! These other lights are also
useful in that high pressure discharge lights (like metal halide) take a while
to reach full brilliance, and if there’s a momentary power cut, they will not
re-light until they cool. Without the other lights, you could be left in the
darkness.
OK, so what could I use as this ‘low power’
lighting? I did some more testing, comparing the 48W compact fluorescent against
some double (36W x 2) traditional fluoro light fittings I’d picked up
secondhand.
Certainly, this comparison was unfair in that the
secondhand tubes may have been pretty old, but the 48W compact fluoro had a far
better output than the double tube fitting. And not only was the intensity
better underneath, the compact fluoro also had much better coverage.
So running three 48W compact fluoros would give
adequate lighting for normal shed use – eg moving a car into the shed,
collecting something from storage, etc. Furthermore, the total load of these
lights would be only about 150W, so this lighting would be very economical to
run. That was the theory – but how well would they work? I temporarily powered
the three 48W compact fluoros from extension cords and mounted them in place.
The results were fine.
Evaluation
OK, let’s take a look at what did and
didn’t work.
The secondhand double fluoro battens were way
inadequate for any workshop lighting. Even with their low power consumption,
they were useless. (This was a fascinating result, because when plugged into a
power point inside my house, they looked really good – helped no doubt by the
much smaller room with white-painted walls and ceiling.)
The secondhand high pressure discharge lights,
complete with reflectors, were inadequate when equipped with high pressure
sodium bulbs, providing an illumination of 250 lux. Considering their high power
consumption, these lights were also of little use.
The 48W compact fluorescent didn’t produce enough
light to be the main, powerful lighting source – too many of them would have
been needed. (At an educated guess – about 24!) However, three would provide
adequate low power illumination – and at very low running costs.
Finally, the secondhand high pressure discharge
lights with reflectors and 150W metal halide globes provide excellent
illumination. With a spacing and mounting height both of about 3 metres, the
illuminance looked like it could exceed 500 lux – and with better colour
rendering than achieved by the high pressure sodium.
The use of a switch panel with three switches
would allow simultaneous switching of the three compact fluorescent lights, and
the metal halide lights in two groups of six.
Yet
Another Light
At
the last minute I decided to add another(!) light to the mix. This is placed on
the outside front of the shed and aimed down and forward.
Where
I live there are no streetlights and so if you’re loading a car out the front
when it is dark, it has always been a bit problematic seeing what you are doing.
|
Mounting the Lights
The metal halide lights are an old-fashioned
design where the control gear is mounted in aluminium housing, with the light
socket and reflector suspended beneath. A short length of cable terminated in a
three-pin plug connects to the control gear.
The first step was shorten the cable between the
control gear and the light socket – this was about five times too long for the
application! Note: this cable carries much higher than mains voltage so
appropriate cable must be used – you can’t replace it with normal power
cable.
After the cable was shortened, the lights could be
hung. I used a short section of chain pop-riveted across the ‘top hat’ section
purlin, with the chain connected to the in-built hook of the light by a
‘joining’ chain link.
A box of 20 non-switched power outlets was bought
(eBay - $40). These used click-in bases and the bases were pop-riveted to the
purlins. In the pic a base is arrowed.
For best results, the compact fluorescent light
had to hang about 15cm down from the rafters. This was achieved by mounting a
‘ceiling rose’ fitting on a folded sheet metal bracket that was bolted to the
rafters. A short length of cable from the ceiling rose to a bayonet light socket
fitting allowed the bulb to be mounted at the correct height.
Because only three were needed (ie three ceiling
roses, three bayonet sockets and three 48W compact fluorescents), these were all
purchased at a hardware store at fairly exorbitant prices – about $110 for the
lot.
Electrician
So that’s the lighting fittings all up and ready
to go. Now, what about an electrician?
In addition to wiring the lights, the electrician
was also required to install a workshop switchboard complete with circuit
breakers and RCD (safety switch) devices, wire-in 13 power points, and run the
cable from the house switchboard. All the lights and power points were already
mounted.
But when I started getting quotes on the cost of
this work, I was massively taken aback. So much so, that we’ll leave the story
of the electrical work until next week...
Conclusion
Obtaining adequate illumination in the workshop
proved to be much harder than it first appeared. However, using a light meter
and conducting some simple tests soon showed what would and would not work. The
final system provides both adequate low-running-cost illumination and also
excellent high intensity illumination for when work is being carried out.
Next week: getting the electrical work
done
Go here for the next in this series.
Interested in home workshop projects and techniques? You�re sure then to be interested in the Home Workshop Sourcebook, available now.
|