COMPARISON OF WHITE LED CAVING LAMPS by Stuart France
White LED lighting has moved on very fast since I designed a 24-led lamp six or so years ago. These units are used by myself and several others in CSS, about 40 being sold as kits of parts or assembled by me for friends. As far as I know, none of them has failed in use. The leds are 5mm pieces soldered into an array of 24 that fits inside an Oldham or CEAG head. More recently, powerful single die leds have appeared, such as Luxeon. These single-die LEDs are used by Speleotechnics in their futuristic Nova lamp and the Mine Explorer lamp (www.mineexplorer.org.uk). I went to try out the led lamps I had access to in OFD.
ULTRALEDS. The 2006 Stuart France design with 24 so-called 18,000mcd leds from Ultraleds. co.uk. These were set up in 12 chains of 2 to give best power efficiency using 3.6 volt batteries, i.e. a fully helmet-mounted setup.
NICHIA. The 2001 Stuart France design using 24 x 8,000mcd leds purchased from Nichia. These leds are top rank for colour and second rank for brightness. These were wired in 8 chains of 3 to support an FX5.
MINE EXPLORER. This is a new lamp design using two 1 watt Luxeons on a PCB that drops into an Oldham/CEAG head. It is said to operate from 1.3 to 7.0 volts, so possibly not OK for a well-charged FX5.
TEST PHOTOS IN OFD
The photos above were taken of Anna Northover, a prospective CSS member, in OFD using an Olympus digital camera with only the LED lamp on test for lighting each shot. Click here to see the colour photo montage.
My general impression is that the Nichia leds I bought years ago have the best colour balance with good browns and skin tones, but they are not the brightest leds now. They do create a bright centre, but not a particularly well-focused one, though the leds are supposed to be 20-degree beam angle. This provides pleasant views of the whole of the passage, but a spot beam. The Ultraleds array was more sharply focused and much brighter than the Nichia one (per watt) but with a blue cast. The light output was similar to the Mine Explorer on its highest setting but at a substantial saving on battery power. Soon I shall convert my own caving lamp to use these leds.
The Mine Explorer lamp on its brightest settings, no.3 or no.4, is very good. It does have a fairly bright spot with useful side scatter too, but there is a green cast typical of Luxeon leds. The unit I tested was on an Oldham battery. Its 4 brightness settings are selected in turn by rapidly turning the caplamp switch off and back on again.
EFFICIENCY
The greatest conversion rate of electrical power into light is the goal, i.e. the highest Lumens per Watt, provided that the light is in useful amounts and the battery does not go flat too quickly or is too bulky or heavy to carry. All the LED lamps here were better than than the standard Oldham halogen bulb both in terms of light output (measured on an Oldham lumen meter) and the efficiency of electrical power converted into light. The Ultraled version of the Stuart France lamp produced 7% less light than the MineExplorer on no.4 but is 30% more efficient at making light.
|
MAIN BEAM
|
Nichia 24x5mm leds on high setting |
Mine Explorer 2 x 1W Luxeon on highest setting |
Ultraleds 24x5mm leds on high setting |
Oldham 4W halogen bulb |
Lumens |
41 |
48 |
45 |
38 |
Battery Volts |
3.8 |
4.1 |
4.1 |
4.1 |
Battery Amps |
0.7 |
0.7 |
0.5 |
1.0 |
Battery Watts |
2.8 |
2.8 |
2.1 |
4.1 |
Lum/W (more is better) |
14.5 |
16.8 |
22.0 |
9.2 |
PILOT SETTING
|
Nichia 24x5mm leds on low setting |
Mine Explorer 2 x 1W Luxeon on lowest setting |
Ultraleds 24x5mm leds on low setting |
Oldham pilot bulb
|
Lumens |
10 |
13 |
7 |
6 |
Battery Volts |
4.03 |
4.2 |
4.43 |
4.2 |
Battery Amps |
0.11 |
0.1 |
0.04 |
0.49 |
Battery Watts |
0.4 |
0.4 |
0.2 |
2.1 |
Lum/W (more is better) |
22.6 |
31.2 |
39.5 |
2.9 |
Probably any caver would be pleased to own any of these LED lamps as they were all wonderful underground compared to the traditional 4 watt halogen bulb with its parabolic reflector giving a tunnel vision effect.
Whether it is a good idea to put microprocessor software into a caving lamp and the long-term reliability of the Mine Explorer product remain to be seen as it is so new. There have been reports in the BCA CREG journal of software crashes due to radio interference from cave rescue transmitters. Still, turning it off and then on should reboot the lamp software and restore normal operation: nuisance if it happened on a pitch. Certain LED lamp switch-mode power supplies (but not mine) put significant radio interference on to cave radio receivers.
OTHER LED LAMPS FOR CAVING
The Speleotechnics Nova lamp has, it seems, gone through various design alterations since it appeared which affect brightness, the design of the on-off switch, and presumably software. I did not have a Nova to test in OFD, so if any CSS member has one then I would interested in borrowing it to do the comparative tests.
Oldham have brought out an LED lamp but it is not an upgrade to their usual caplamp product: it is a total replacement of it. Their unit is made by First National Battery in South Africa. It uses a 4V 4AH sealed lead acid battery in a small Oldham-style battery case carried on a belt. The headpiece is the usual FNB unit which is bigger than Oldham's and only has a single on-off setting, i.e. no pilot. The light output was only 16 Lumens on the one I tested which makes it more like a “bright pilot bulb” than a main beam when compared to the others I tested. However, its relative dimness is rescued by its tight spot beam of only 75cm at a range of 5 metres. There is no sophisticated power supply or brightness control and it uses just one resistor to control the current. It had the best efficiency of 48 Lumens/Watt amongst the units I tested, but it fails my “utility” test because it is too bulky, too dim, and too tightly focused for comfortable caving, and no brightness control.
24-LED LAMP DESIGN 2006 UPDATE
Other than wiring the LED array as 12x2 rather than 8x3, using the same PCB as in 2001, which has the effect of selecting a different bank of 8 LEDs for the pilot setting, the only other change I have made is to drop the 20K feedback resistor to 15K when running the 12x2 array off a 3.6 volt 4AH helmet-mounted battery pack.
Main Beam 2001 Pilot 2006 Pilot
The original circuit for this lamp is published in the BCRA CREG Journal No. 49 (2002) pp 26-29.
The 12x2 array runs at about 6.6 volts by stepping up the voltage of a helmet-mounted battery . This means it cannot be run with an FX5/Kidney pack because those types are well over 7V fully charged.
