|
| Home | Guest - Not logged in | |||
|
» Register
» Login
Gear Reviews
Documents
Tools
» Contact
|
Reviews > Lighting > Headlamps - LED > Princeton Tec Quad > Roger Caffin > Field Report
| |||
Reviewer Details
| Reviewer: | Roger Caffin |
| Age: | 61 |
| Gender: | M |
| Weight: | 63 kg (139 lb) |
| Height: | 167 cm (67") |
| Email address: | r dot [surname] at acm dot org |
| Home: | Sydney, Australia |
I started bushwalking at 14 and took up rock climbing at University with the girl who became my wife and my permanent walking partner. Ski touring and canyoning followed. Winter and summer, we prefer long hard trips by ourselves: about a week in Australia, up to two months in Europe/UK. We prefer fast and light in unfrequented trackless country. We would be out walking and skiing for at least three months a year. We have now moved to lightweight gear, much to our backs' relief. I designed and made much of our lightweight gear myself.
I am also the maintainer of the Australian aus.bushwalking FAQ web site www.bushwalking.org.au/FAQ/.
Product Information
| Manufacturer: | Princeton Tec | |
| Manufacturer URL: | www.princetontec.com/ | |
| Year of manufacture: | assumed 2006 | |
| Country of manufacture: | China | |
| Colour: | Blue & Yellow housing, black and grey headband | |
| Batteries: | 3 AAA, any sort | |
| Listed weight: | 96 g (3.39 oz) | |
| Actual weight: | 96 g (3.39 oz) | |
| MSRP: | na |
Product Claims
Background
I am a consultant research scientist, and have spent many years working in the area of computer image analysis. One of the most fundamental issues in image analysis is getting the lighting right, so I have spent a lot of time working on designing and testing lighting systems. In addition I have made my own small lightweight Light Emitting Diode (LED) headlamps using miniature switch-mode power supplies which I have designed and built. So one might say I have a bit of an interest in the technical side of headlamps. Finally, I do have a fairly good research laboratory of my own.
Laboratory Testing
It seemed to me that this is the sort of thing which can be tested in the laboratory as well as in the field, so I decided to investigate the power supply and the light regulation on the bench. Specifically, I wanted to find out what sort of current drain the headlamp puts on the batteries at the different power levels. This would help me estimate the likely battery life, given that I have data on battery performance. Along the way I hoped to find out a bit more about what sort of regulation system is used in this headlamp, and how efficient it is.
As batteries get used up the voltage they deliver decays. When the output voltage gets low enough one says the battery has died. The higher the current drain, the faster this happens. So the obvious test is to monitor the current into the headlamp as the voltage is varied. It might be expected that as the voltage decreases the output power or light level would decrease, so this should be monitored at the same time. Ideally it would be nice to know how much power is being delivered to the LEDs as well, but I have not yet found a way into the 'insides' of the unit yet, so this has not been possible. However, the light coming out does give some guidance to this.
I considered mapping out the brightness field, but really I could not see that this would be of all that much use in practice. The headlamp uses Nichia super-bright white LEDs (according to information from Princeton Tec obtained elsewhere), and the field from the headlamp is dictated by those LEDs. Instead I will comment on the usefulness of the light field in the Field Testing section.
Experimental Setup
I opened up the battery case by undoing the single screw at the back and then mounted the headlamp on an optical bench, as shown to the right. I supplied power to the battery case from an external variable power supply, through a current meter, and I monitored the voltage at the battery case connections as well. Some distance away I positioned the sensor for a Lux Meter. This is essentially a light meter, but with a filter designed to match the response of the human eye.
At this stage I should point out that while the measurements of volts, milliamps and power are entirely reproducible by someone else, the measurements of light intensity (or Lux) are not. This is because even a small change in the alignment of the headlamp to the sensor or in the distance between the two will cause a change in the 'scale factor'. So someone else doing the same measurements would get different numbers for the Lux readings - but they would get the same sort of curves. My discussions will therefore focus on the behaviour of the light output, not on the numerical values. This will become clearer as I go.
Measurements
The basic results I obtained are shown here. The blue lines show the current as a function of the voltage at the battery case terminals. The red lines show the light output as a function of voltage at the battery case terminals. The curves come in three pairs: 'H' for high power, 'M' for medium power, and 'L' for low power. These are the three power levels the headlamp provides. I did not try measuring the flashing mode.
The headlamp uses three AAA cells: Duracell alkaline ones were provided. These should provide slightly more than 1.5 volts each when new, or a fraction over 4.6 volts total. Typically the 'end of life' for alkaline batteries is defined as 0.9 volts, making 2.7 volts total. So my measurements were made over the range from 4.6 volts to 2.8 volts, with a 'zero' reading at 2.2 volts.
I should also explain that the nominal voltage required to drive a white LED at the nominal current is 3.6 V. However, the voltage/current curve for a white LED is not sharp: it actually looks very much like the smooth curve seen here between 2.5 and 3.8 V for the H case. Technically this is an exponential curve, but no matter. The nominal current rating for a standard LED is 20 mA, and LEDs have something like a 50,000 hour life at this current. I believe the High setting puts about 75 mA through each LED, which is going to stress the heat dissipation a bit and reduce the life of the LEDs somewhat. Well, even if it reduces the life to 5,000 hours (down to 1/10th of the nominal value), that is probably longer than most users will ever need.
What these results tell me about the headlamp is as follows. Provided the batteries deliver more than the required voltage, the internal regulator limits the current supplied to the LEDs to a roughly constant value. Well - 'roughly constant'. In the H case the regulation is not very good, so the current to the LEDs varies a bit as the voltage falls, but in the M and L cases the LED current is maintained at a fairly constant level. Investigation with an oscilloscope showed that in the H case above 4.2 V there is a high-frequency ripple on the supply, but this ripple does not appear under any other conditions. This means there are actually three different operating regimes for the power supply inside the headlamp, as follows.
These results make sense out of the Princeton Tec specifications for burn time, which talk about a 'sophisticated current-regulating circuit that maintains initial brightness as long as the batteries have sufficient voltage'. The specifications go on to mention both 'Regulated LEDs' and 'Unregulated LEDs'. The regulated regimes, which by their specifications have a rather short period, correspond to the first two regimes above.
Finally, I tested the little red 'Battery Power Meter LED' found under the white LEDs. The specifications state that 'If the headlamp is turned off and the battery voltage is low, the red ... LED ... will start blinking'. This does indeed happen when the total battery voltage falls below 2.7 V. This corresponds to a single battery voltage of 0.9 V, which is a very common 'end-of-life' value. I found I had to remove power from the headlamp to stop the red LED from blinking.
Battery Characteristics
It seemed to me that it might be useful to put the above measurements into context by looking at battery characteristics. I have put typical cell ratings for the common Eveready Alkaline AAA cell and the new Eveready Lithium AAA cell to the right here. Several things stand out from these curves.
The above details help explain why the Princeton Tec specifications claim the headlamp will run in regulated mode at High power for 4 hours on Lithium cells but only for 1 hour on Alkaline cells. The details also explain why the headlamp will run for a lot longer in UNregulated mode (with a decaying output brightness) with Alkaline cells than with Lithium cells: the Lithium cells hold their voltage up right to the end (thereby staying in the regulated region), but the Alkaline cells don't.
The curves also point out a very slight problem with the low battery threshold of 0.9 V. An Alkaline cell still has a small amount of power left at 0.9 V - not much, to be sure. But a Lithium cell at 0.9 V has almost no power left at all! This means that I might have some minutes left to fossick around in my gear to find replacement batteries if I am running on Alkalines, but maybe not if I am running on Lithiums. In reality I don't think the red LED has much value in the field. It isn't hard to see when the headlamp is losing power: everything starts to get a bit dimmer, and changing between power settings has no effect. And that is before I turn the headlamp off! If I can't tell that the batteries are dying some considerable time before the red LED starts blinking, then I might as well give up.
Battery Case
The battery case has what looks like a good O-ring seal, as shown here. I have not pressure-tested this, but my experience with O-ring seals on other equipment leaves me happy with the design of this one.
The battery connections seem to be designed so that putting a battery in reversed should not work. There is a recess at each positive terminal in the battery case ('Recessed +ve terminal') protected by surrounding plastic. The recess should only be spanned by the protruding lump on the positive end of the AAA cell ('+ve end with lump'). However, I found that the negative end of the AAA cells provided with the headlamp could make an unwanted contact when inserted reversed ('Unwanted connection'). This is illustrated by the middle cell, which has been inserted into its slot back to front.
This is a distinct pity as a very slight modification to the metal tabs at the positive terminals would prevent this. All that is needed is for the dimple on the tab to be reduced in height by about 1 mm (0.040") and the gap in the plastic reduced in width by about the same amount. This would mean that the shoulder of the battery at the wide negative end of the cell hits the plastic around the recessed terminal first. With this slight modification it would still be possible to insert the cells in the wrong direction, but they would not make electrical connection and there would be no risk to the electronics or the LEDs.
In the meantime I can still load the batteries by touch in the dark, making sure I put the flatter end of each cell (Flat -ve end) against the spring (Spring -ve terminal) in each cell slot.
Summary from Laboratory Testing
Now I know what is inside the headlamp. There is a regulator of sorts, but it is a rather simple one which does not really make the best use of the power available in the cells in my (professional) opinion. For better performance a genuine switch-mode regulator would be required. Against that I have to balance the typical retail price for the Quad of between US$26 and US$36: it costs me almost that much just for the components to make a genuine switch-mode regulator for an application such as this! Princeton Tec has achieved a fair bit for the low price.
I must make it clear that the current-regulator used in the Quad does work, and it does limit the voltage applied to the LEDs so they can have a long life. Pushing about 75 mA through each white LED at the High setting is stressing them a bit, but the other settings are no problem. I should point out that it is very likely that with alkaline cells more time will be spent in the unregulated mode than in the regulated mode. With lithium cells the regulated time would be much longer relative to the unregulated time.
I am going to be a trifle pedantic here and say that calling this a 'fully regulated' headlamp would not be correct. However, Princeton Tec does not claim this; the company only claims that 'The Quad uses a sophisticated current-regulating circuit that maintains initial brightness as long as the batteries have sufficient voltage.' In my opinion this is a correct description of what the Quad offers. It's just that only the lithium cells can maintain that 'sufficient voltage' for very long.
I note that another well-known manufacturer of headlamps has come out (on their web site) against the use of lithium cells, despite their obvious advantages of capacity and low temperature operation. It seems that the current-limiting mechanism used by this other brand cannot handle the higher power of the lithium cells and the headlamps can overheat and be damaged. Well, Princeton Tec not only 'allows' the use of lithium cells, but the company gives performance figures for them. I applaud this.
All three settings are bright. I can see the change in light output when I switch between the three settings. However, I also know that humans have some difficulty seeing anything much less than a factor of two in a change in brightness. The Medium and Low settings are barely this much different. A Low setting considerably dimmer than what is provided would be of greater value. I know I normally work with a much dimmer light around camp when I am using the headlamps I have made myself.
Field Testing
I have taken the Princeton Tec Quad headlamp on a number of long trips during the Field Test period. These have been at altitudes from sea level to about 1000 m (3,300'). Temperatures have ranged from about freezing to about 30 C (86 F). Fortunately, none of these trips have required that I travel by night very far, but I have tried wandering around the campsite at night with the headlamp. The scrub around many of our campsites has been fairly thick rather than open grassland. Somewhat fortuitously, most of these trips have had no moon in the evening, so the headlamp really has been needed. These outside uses have required the headlamp to illuminate at a moderate distance of a few metres (yards).
I have also used the headlamp to cook dinner by inside the tent on those many occasions when our days have ended a bit late in the evening. There has been enough light to get the tent up, but not enough by the time I had got the stove out. These uses have required short-range illumination of about half a metre (yard).
In both outside and inside use I found that the lowest brightness was quite sufficient for me. In fact I would have been quite happy with an even lower brightness than the headlamp can offer. A lower brightness would mean a longer battery life of course.
One thing I did find during use inside the tent was that it was very easy for me to dazzle my wife accidentally, even at the lowest brightness. I was able to reduce this problem by adjusting the tilt of the headlamp to be very much downwards rather than forwards. The stepped adjustment mechanism the headlamp offers proved to be quick, convenient and stable. The broad headstrap or strip of elastic held the headlamp on my head quite stably as well, so the 'pointing' was stable in general. The downward tilt is shown in the picture here. Of course, the flash on the camera has washed out the beam.
Very late in the evening after dinner when getting ready to go to sleep I found it was more convenient to hang the headlamp from a 'skyhook' at the roof of my tent, pointing downwards, although this required some ingenuity with the broad headstrap. It does not offer a simple method of doing this, so I had to tie an overhand knot with the band. It worked fine.
I have experimented a bit with the switch itself. It can actually be quite hard to sense because it has virtually no protrusion sticking up. This is presumably designed to minimise the chances of accidental activation in a pack. I dare say that over time and with more practice finding the switch might become easier, but it was a bit tricky at first. Eventually I found that I could find it not by position but by first locating the side with the battery case screw, then rubbing my finger across the top surface. The compliant rubber yellow seal has much more friction than the plastic case, so it feels different and can be located even with gloves on.
I do have to report that the idea of the almost-recessed on-switch seems good but has failed at least once. That is, I did find a glow from the bottom of my pack once: the switch had been activated inside my pack and the headlight had been glowing away there for some time. This could be very embarrassing if the batteries end up flat just when I need the headlight most. For this reason I always put an interlock on the headlights I make: something like a slip of plastic inserted between a battery terminal and the contact to prevent activation. However, I can't see any good way of doing this here without substantial modification to the headlamp shell. That would remove the waterproof feature. I could of course remove a battery each morning before packing the headlamp away - a lot more work which often does not get done.
A Small Annoyance
While overall experience with the headlamp has been fairly good, it does have one feature which is very annoying. When I switch it on, it goes to the brightest setting - and this is too bright for me. Every time I had to give the switch not one click but three, with my hand over the front of the headlamp to avoid dazzling myself and my wife. It would be ever so simple to reverse the settings in the design of the regulator chip so the first click gives the lowest setting. This is not the only headlamp with this fault.
A change in the order of settings would also prevent me from over-running the three clicks to four, thereby getting that infernal flashing bright setting. Frankly, I hate that flashing! I would dearly love it to be replaced by a much dimmer setting again. A nice idea on the designer's desk perhaps, but a menace in the field.
Summary so far
Things I have been looking at include the following:
Recommendations for Improvement
