LPM Request (Free diode!)

[Rifter]

While I would absolutely treat a 20mW beam as being dangerous, it is worth noting that the ANSI MPE values are based on a factor of safety of around 10. So in their studies, it actually must have taken something in the ballpark of 50mW to cause visible, measurable retinal burns from a brief exposure.

Regardless, 20mW is close enough to that number that it should be treated with a great deal of respect. It's just interesting to see how these power limits were initially established.

I have read the same info, im sure I have seen a study done by a US university that showed up to 40-50mw is still "blink reflex safe"

The 5mw # used is not taking into account the factor for safety the FDA and ANSI and all other standards associations use.

That said this is your eyesite we are talking about, DONT TAKE CHANCES. I use safety goggles anytime I am using anything class 3B or higher where its possible to catch a specular reflection off anything.

 

[Bionic-Badger]

I have read the same info, im sure I have seen a study done by a US university that showed up to 40-50mw is still "blink reflex safe"

I'd like to see the citations for that.

The paper "A Critical Consideration of the Blink Reflex as a Means for Laser Safety Regulations", H.-D. Reidenbach, et al finds that the 0.25 second "blink reflex time" is actually too optimistic when considering class 2 lasers of 1mW which are supposed to be blink reflex safe.

You can also read the derivation of power densities for low power lasers on Sam's Laser FAQ.

You have one pair of eyes. Don't be a fool and risk them just because of some bogus safety claims and anecdotes on a website forum.

 

[Insomniac]

From this paper:

"This is the ANSI “maximum permissible exposure” (MPE) safety level, set at approximately 1/10th of the laser exposure level expected to produce biologic effects. At a laser exposure level of one times MPE (1x MPE) absolute safety would be expected and retinal exposure to laser radiation at this level would be expected to have no biologic effect. Based on ANSI data, a 50% risk of suffering a barely visible (threshold) retinal burn is generally encountered at 10x MPE for conventional CW laser exposure. "

 

[Bionic-Badger]

That's just the definition of MPE, and would be the standard for a Class 1 laser that is considered safe under all circumstances that do not involve light-concentrating optics. This would include laser such as those used in check-out lanes at grocery stores, and also the standard that must be met for most commercial laser shows at the location that beams intersect the audience (if allowed to). It is well below that of Class 2 lasers, which are < 1mW, that rely on the "blink reflex." The 10x factor is also for a visible retinal burn.

 

[Insomniac]

Well, yes. I certainly wouldn't call anything over 5mW 'safe'. My point is that 5mW is not some magical number, below which lasers are safe and above which lasers are dangerous.

My understaning is that ANSI did a bunch of tests and found an exposure level at which they would expect to see a retinal burn about 50% of the time. They divided this by ten and called that level the MPE. Lasers higher than 5mW will exceed the MPE, even if you blink. However, the factor of safety means exceeding the MPE by a factor less than ten should be *unlikely* to cause a visible burn. Whether invisible damage will occur... Well, who knows?

Either way, the consequences are too great to rely on this factor of safety and >5mW should be treated as damaging.

 

[The Lightning Stalker]

Everything was going well and then the diode suddenly failed for no apparent reason. I
blame the poor packaging. It came in a polyethylene bag inside a polystyrene "box". I was
able to do a power test, but not a wavelength test, unfortunately.

Lens was a standard AixiZ style uncoated acrylic

Threshold Current: 29mA
Voltage Drop @ 44mA: 2.3V
Photodiode Voltage @ 5mW: 498mV

These are case positive, which is a bit nonstandard for red diodes.

I went ahead and bought 2 more of these diodes. Hopefully at least one will come
through the hell this guy subjects diodes to and still work.

 

[Insomniac]

Excellent, that was exactly what I wanted to see. Thanks. Shame your diode died, however. Mine was packed the same way and is running fine in a pointer. No issues with it dying and I've been using it for a few days now.

Unfortunately, I have realised that I'm not using mine to its full potential. It is mounted in a modified green module and much of the beam is being cut off before the lens. It appears slightly dimmer than my 650 nm pointer. On the plus side, my Aixiz housing arrived so I will be able to make a better one once my second diode arrives.

 

[The Lightning Stalker]

Through a bit of Google wizardry, I believe I may have found the source and part number of these diodes.
They are being sold on Aliexpress as LT-LD6305 and are available for $1.61/pc if you buy 50 pieces. I
don't know what I would do with 50 pieces, though.

 

[RayJay]

So mine came in yesterday, the same day as my new nospin ophir! I will post my results later tonight.

 

[Insomniac]

That's a nice coincidence! Cheers.

EDIT: @The Lightning Stalker

That diode does indeed seem to match. I also have another two diodes on the way, one for use in a pointer and one spare. Are you using yours for anything in particular, or just trying to get better data?



I spent today playing around with a breadboard, trying to make a driver to replace my simple resistor setup. I'm trying to achieve regulated ~35mA output with a supply voltage of 2.4 - 3.2 V. However I'm having trouble with it. I can achieve decent regulation (with a couple of 2N3904 transistors), however it does not start regulating properly until 3+ V. I'm gonna try again tomorrow after buying a low voltage op-amp. My reason is that I want to use it in my 2*AAA pen and be able to use NIMH or alkaline batteries without a large change in output power. I realise I can always jump over to lithium cells or a different battery configuration, but I'm trying to avoid that.

The circuit I've been playing around with is the one HERE. I've fiddled with it a bit and simulated a few different configurations in PSPICE, but I just can't get regulation at such a low supply voltage.

Does anyone here have a trick to get around this issue, or is this ultimately futile given how ludicrously small the drop-out voltage of my driver will need to be?

 

[The Lightning Stalker]

Does anyone here have a trick to get around this issue, or is this ultimately futile given how ludicrously small the drop-out voltage of my driver will need to be?

It's amazing just how well this type of setup can regulate with how simple it is. Here is a
modification that which might lower the dropout somewhat. For the FET, you can use a
2N7000 or almost anything really since the current is so small. I'll just post the image since
something is going on again with my crummy hosting.

There is a more complete description at
users.silenceisdefeat.net/~lgtngstk/Sites/Circuits/High_Power_LED_Driver/High_Power_LED_Driver.html
whenever it decides to come back up. I know it says LED driver, but it can drive a laser
diode. Just put a 10µF capacitor in parallel with the diode.

edit: You can lower the drop even more by using a germanium transistor (about 0.3V). You
will also have to change the resistors to compensate for that and the gain.

 

[Insomniac]

Hmm. I'd seen the version with a FET, but I don't think it will work for me due to the relatively high threshold voltage of FETs that I've seen. The 2N7000 you suggest, for instance, can have a threshold voltage anywhere from 0.8 to 3V according to the datasheet. Maybe some samples will work, but not all. I did try with an IRF540 I had laying around, but it didn't work as well as the BJT version.

Using germanium transistors seems like it might be the best bet. Might have to grab some on ebay to test with, thanks for the suggestion.

 

[The Lightning Stalker]

Yeah, that is typical of the IRF540. Check out the DMG4800LK3-13. Paired with a germanium BJT, it
might be just the thing.

I've been doing some research on Ge transistors. They can have a high leakage. That means it needs
less resistance on the pull-up to compensate. Also, this circuit needs to be played with because no
two transistors are the same. Either get a resistor assortment or use a pot on the pull-up and swap
out sense resistors since a pot is no good as a sense resistor. Also make sure the resistor is foil or
carbon and not wirewound, or else it might cause some problems. I have used wirewound, but haven't
tried scoping it yet to see what is going on.

 

[Insomniac]

Just gave it a shot with a VN10KM MOSFET with a Vgs(th) of 1.2 V.

Unfortunately, performance was almost the same as with a BJT, with the difference being I could use a much larger pull-up resistor on the gate (as no current needs to flow through the gate).

Either I got unlucky and my MOSFET's threshold current is at the upper end of the allowed range, or this modification simply doesn't address the problem. I'll hold on to it for when I get a germanium resistor.

You're suggesting wirewound resistors will act as inductors?

 

[The Lightning Stalker]

You're suggesting wirewound resistors will act as inductors?

Most wirewound resistors ARE inductors. They are inductors of an unknown inductance!
There are also "non-inductive" wirewound, which I think are just a straight piece of wire, but
they are uncommon.

 

[Bionic-Badger]

The only problem with the transistor circuit above is that it can really heat up because of the voltage dropped by the transistor and the resistor. Otherwise, it's a nice circuit and you can also use it for analog control of your output.