Apologies for the lack of response. No one was able to help me so I had long discarded this as a resource for information. However, CPF has helped me in the past so I feel obligated to help the next people in line. I hope this response will be a catch-all to those wanting to make a 365nm flashlight.
Wow. How far off was their reference spectra from what you measured?
Unfortunately I do not have a spectrometer to test these things out with. My Samsung phone has a debug mode that gave me readings from the light sensor though. This is the best information I could get:
LED ENGIN LZ1
- 7000K ~6000 lux held right against sensor driven at 1A (the max according to LED Engin's spec sheet. I noticed the same green tint at 500ma but did not do any testing at that range.)
Chinese no-name 365nm LED. Here's a picture of a 3x LED version. I've seen them in 1x 3W,2x 5W, or 3x 7W offerings:
http://imgur.com/7zcLbZo . They have a max of 1.2A per LED according to the eBay seller I contacted. Take that with a grain of salt. I tested at 1A safely though.
- 7000K ~400 lux held right against sensor driven at 500ma
Not quite apples and oranges with the different draws but hopefully that illustrates just how much visible light is coming out of the LED Engin LZ1
Other members here have had great luck with the LED ENGIN LZ1. Maybe you did something wrong with your build?
Perhaps. It was only a sample size of 2. I don't discount the possibility that this particular batch was bad or my 2 samples were not representative. I can only speak from my experience. I'm not sure where one could go wrong though. It's literally an LED and driver soldered together for P60 dropins. Sure color temperatures and outputs can change with temperature increases but the problems consistently happened whether it was a cold start or after 5 minutes of operation.
Conclusion. What I did in the end:
Even though I berated the LZ1 I stuck with it. Why? Because it was the most efficient at 1A. To compensate for the visible light I used a cutoff filter. They can go under different names:
- Woods filter
- LW filter (LW = Longwave). SW filter should be the same as my recommended glass.
- Hoya U-325C My recommended glass. The other 2 only transfer about 60% UVA. This one does 80%. Don't quote me on those exact numbers. Just know that this one transfers the most UV. Also the most expensive
If you just want a solution you can use out of the box the no name 365nm LEDs that Tank, Ultrafire, etc. use. They're all the same LED I bet. If you are willing to cut glass or use adhesive to stick the filter onto the end of your flashlight go with the LED Engin.
Examples:
Here are some photos to help illustrate my point. Focus on the materials fluorescing as opposed to the visible light. That's the point of a UV light after all. I tried to make the brightnesses equal at the same EV so I don't see the need to emphasize the settings I used. Also keep in mind the no name Chinese is running at 500ma.
In this first set of imagesimages we can see that all of them will do fine for discovering stains. I'm not sure what these stains are from exactly as they were left by the previous owner but I'd bet on chemical. I'm told that urine stains are more subtle so maybe that's what the blue is?
http://imgur.com/a/xXYja
The first image is the no name Chinese 365nm LED. You can easily see the yellow staining and the blue stain pops out. Much better with the visible eye. Remember 500ma vs 1A! In the second the same is true for the LED Engin LZ1. It might be a bit harder to see due to less contrast with the visible light but it's evident if you look hard enough. The third image definitely won't make sense unless you've used a proper UV light. It's the filtered 365nm. What's going on here is that the bleach that the carpet makers used is weakly fluorescing but doing so nonetheless. The fact that this can pick up on that makes this a top class 365nm light in my book. Also notice how much more blue staining you can see.
http://imgur.com/a/bV7XA
I didn't feel that that illustrated my point well enough though so I took another set. It hopefully illustrates the detriments visible light does to subtle fluorescence The subject is a mangano calcite from Peru. The first photo is a natural light image. The second is the filtered 365nm. This is what our LEDs should strive for ignoring the blue lint. The third is the no name 365nm LED. It clearly shows fluorescence but the color seems slightly altered. Likely due to the slight bit of purple light coming out. The last one is the LED Engin LZ1. This one shows why you want little visible light out of your UV lights. You can barely see the fluorescence at all if you compare it to the visible light shot.
