Amber LED vs Red LED for night-vision torch?

[julianh72]

There's an interesting article "Is red light really best?" by Robert Dick in the July 2016 issue of "Australian Sky & Telescope". The author makes a good case for using a 1 to 3 lux amber light (rather than red) for best night-time acuity (e.g. reading star charts etc), while still minimising the impact on recovery time of our night vision.

I have always struggled with reading under dim red light, and I suspect that my issues are exacerbated by my red-green colour-blindness, as red is a dull colour for me even in bright daylight. However, Robert Dick argues that our acuity (i.e. ability to resolve fine detail) is poorer in red light than shorter wavelengths such as amber. He also argues that low-level amber light allows our eyes to recover their night-vision sensitivity as quickly as red light - and he apparently has the support of the International Dark-Sky Association.

I'm going to rig up a basic low-lux amber LED and give it a try, but my suburban backyard is by no means a "dark sky" site, so my testing may not be very conclusive. Just wondering whether anyone else has had any experience with using amber LED rather than red LED illumination?

[pfitzgerald]

Hi Julian

Not yet, but like yourself I've read the article and will look into giving it a try.

Paul

[Saturnine]

Having also read the article by Robert Dick in the latest S & T, I'm keen to try an amber led. It seems like at least a few others out there that we've found that red leds' unless quite bright, are not easy to read by and an alternative would be welcome. Time to do some trials.

Jeff

[Shano592]

I have just purchased some amber LEDs, in the 590-593um range. I will chop out one of my red lights when they arrive, and will report back.

They are coming from Melbourne to Sydney, so the light should be ready for August outing, weather permitting.

[AEAJR]

Hummm, I think I am going to follow this discussion. Interested to hear your findings.

[Shano592]

OK, I was at Bunnings earlier today on a futile quest. A full spectrum light bulb shouldn't be that hard to source, but Bunnings.

On my way out glanced at an LED cap light. We used to wholesale these years ago, but I don't have any to experiment with for this project, so I grabbed one of these. For $3.89, it is very little to lose.

This will be my guinea pig for this test.

It's also a good thing, because it means that I can do a side-by-side comparison of red vs amber, not having to decimate my goto light.

I have already disassembled and then reassembled it, and determined the LED polarities. If I remember, I will bring the solder sucker and soldering iron into work tomorrow, and this should be done quickly, with a minimum of cussing.

The attached image is the light, with the loose amber LEDS. The key is simply to show scale.

[Shano592]

Update:

I managed an early escape from work, so I got home in daylight (!), and was able to get the LEDs swapped over.

From the outside, nothing is different; it looks no different to what I unpacked. Internally, there is a possibility (with no admission) that I may have reversed the polarity of the LEDs. If - IF that was the case, then simply reversing the batteries in their bay may fix this. If that was the case...

Now for first light, pun intended. They are bright, and I wonder if - wavelength questions aside - they may simply be too powerful. Later this evening, I will dark adapt and see what effect it has on night vision.

One thing is definite. It is much easier to see and read things in the dark.

I have attached a couple of images. One just shows the torches being compared. The other two are side-by-side comparisons, both right at the wall, and about a foot away. Granted the red only has 2 LEDs, while the amber has 3. However, the difference is marked. Keeping in mind that the red and amber LEDs are both 5mm, and rated around the 2 to 2.2v mark.

The amber is powered by 6 volts (2x CR2032 lithium cells), while the red is powered by a standard 9 volt battery, which has likely been stepped down to 4 - 4.5 volts, to avoid blowing out the LEDs. So the voltage per LED would be about the same.

Just finishing this post, I fitted the light to a baseball cap, and have been lighting up a wall about 6 metres away from where I am sitting.

Did I mention it was bright?

[sharptrack2]

Shane,

The brightness can be controlled by adding more resistance on the ground (-) side. Typically LED's run well at about 15mA. The newer ones, probably lower, been a long time since I fussed with them. The current is typically controlled in small devices like this by a single resistor which sets the current based on how much voltage is left to drop after the LED drop (typically 0.7 Vdc). I usually see resistors in the 300-600 ohm range for battery powered devices.

By putting the resistor on the ground side, you keep the voltage constant but limit the current through the whole circuit. I would think the LED's are wired in parallel so they all draw the same current at the same voltage. If you can find the common connection on the negative side and put a potentiometer (variable resistor) between it and the negative terminal, you can set the brightness to your liking. Probably a 1K to 2K potentiometer would suffice.

[Shano592]

Hi Kevin.

I was just going to add a piece of neutral density film over the front of the leds. I'm not sure that there is room to add anything in this particular housing.

I could also try removing one battery and bridging the connection, to see what 3v gives. If they are in parallel, then it will just drain the single battery faster, but i know that one led works just fine, connected directly to a single 3v battery. I have thousands of them in the warehouse, so draining one isn't an issue. If that dims the leds, then I can show a comparison between 3v and 6v.

[sharptrack2]

Half the voltage should half the current. Should work as a comparison.

Can you beg a light meter off your partner? :scared3:

You might be able to quantize the difference with readings from it. :D If not I have a couple of very old analogue meters we might be able to adapt. Might need some old school mercury batteries though.

I'm interested in adapting a couple of old Maglite flashlight bodies to LED, will PM you for details on the amber ones you have.

[Shano592]

I shall dig out the light meter and do an empirical experiment.

Results to follow tomorrow.

[Shano592]

OK, I pulled out an old luxmeter from the old (and I mean 90's) camera bag. It is a Calcu Light-X, for those in the know. It has a sensor on the top, and you can select either the direct-feed lens, or the translucent bubble, to gather your reading. I chose the bubble for this reading.

With 6 volts, the reading came back as 33 (I'm assuming lux, being a luxmeter). Pulling out one battery, and therefore giving only 3 volts to the 3 LEDs, the luxmeter came back with a reading of 24.

While the LEDs are still bright to look straight at (duh!), the effect of removing one of the batteries makes things a lot easier on the dark-adapted eye.

Reading is still easy, but significantly easier than the red LED setup.

I could probably remove one of these LEDs, and get an even better experience out of it. Three is probably unnecessary, as they produce much more visible light than the red. Covering it, it brings a reading of about 21, but visually it is significantly less light output in a dark space.