New Laser MPE Limits published

[norty303]

Is the significant difference with projectors that the light is very much non-coherent by the time it gets to the aperture?

[Laser Electronics Ltd]

Hi Norty
No the light from most current commercial LVP's maintains its coherence, this is why you dont have to focus the newer video projectors very much.

[JStewart]

Most readers will agree that if you take one of the 24 laser diodes from a Casio laser illuminated projector (LIP), add the typical optics people use on this forum, and aim it at a piece of card several metres away, you'll soon end up with a hole appearing and a burning smell. We all know that if you aim a fully functioning Casio LIP at the card from the same distance, there is no resulting damage.

Why is this so? The MEMS/DLP element in the projector acts as a diffuser, significantly altering the properties from the blue laser sources before they reach the output aperture. This is the reason that the retinal thermal hazard of all the combined embedded sources is substantially less than that of single unmodified source.

The exposure limits in HS(G)95, (which is guidance, not a regulation), are not really suitable for evaluating sources of this type; exposure limits expressed in radiance is the better way to go. There is an IEC committee currently working on a safety standard for LIP devices. The limits in the standard are expressed in radiance for the retinal hazard regions for this reason.

I haven't tested a Casio device, but have the radiometric data from a 5000 lumen LIP, and the retinal thermal exposure puts it into a RG2 grouping. i.e. direct exposure does not pose a hazard due to the aversion response. The same could not be said for viewing one of the single unmodified laser diode sources!

I made the Laserscope association, because in post #9, the concern being expressed about Casio was the fact they had introduced to market a product that was being used for harvesting laser diodes.

If Casio were making flying spot laser projectors, I agree that it could be useful for them to get involved, but hopefully you can now appreciate that, although the figures used for the embedded lasers appear quite alarming, the output beyond the aperture is not. And can therefore see why the likes of Casio would probably have little interest in getting involved in a problem that is not relevant to the normal use of their product.

James

[The_Doctor]

Sorry to bump this again (as if mentioning it twice in two days isn't enough..), but I think I have a good enough reason to post directly to it..

In a couple of threads in the same two days, there is talk of damage to CCD's from lasers during shows. I understand that it may be the newer, more expensive low-light sensitive devices that are most prone, and perhaps also those with large lenses. Anyway, whatever the cause, isn't this worth exploring for an MPE meter? It will take a lot of research that is far beyond my means, but I have some notions worth following up:

If a quantified exposure is known to produce some aberration, regardless of sensor response speed, it may be that taking a 'before' and 'after' image pair can be used to detect pulses above some threshold for energy, regardless of power, and that lateral extent may possibly allow inference of beam width.

Even if beam width can't be inferred that way, (I imagine a laser beam would focus down to a spot size based on wavelength rather than beam width if focus was optimal), there could be another way to do it in combination. If some replaceable element were designed, then damage could be collected and examined later, perhaps milliseconds later, by analysis of two image frames, and the element would be replaced before it became exposed to the point of giving poor readings.

All the above is assuming that some cameras are damaged even by eye-safe laser light, which according to observations in other threads, some of them may be.

EDIT: Instead of 'inferring' anything about beam width, just stick a lens cap on the camera. As 7mm seems to be the standard pupil size for MPE measurements, it is cheapest to drill a 7mm hole in a lens cap! (And some finer holes to detect those beams that are inadequately diverged).

[mixedgas]

I wish it was so easy with CCDs. When you use one for beam profiling, it takes skill, and manual adjustment to set the gain and gamma correctly to get a good reading. Their dynamic range is huge, but when you digitize with one close to saturation, it takes quite a bit of adjusting and fiddling with either the laser power or the camera settings. There are also issues with the readout timing, most of them do not do progressive scanning for a full frame, but do techniques with individual odd and even fields. Then you have fun with electrons spilling out of the wells and into the readout circuits on chip, usually resulting in saturated bars in the image. Most cameras do not have a global sample and hold, either. usually we set the profile thresholds for 90-10 or 84-16 edges.

With a non-Gaussian beam, the photodiode makes short work of this, compared to a CCD.

Without care and manual adjustment, its not something I'd trust my eyes to. Certainly not with a 8 bit video digitizer, or even with a 10 bit digitizer. 10 is about as good as it gets with a prosumer camera.


I'm the proud owner of a Spiricon Beam Profiler, which uses generic stock CCD cameras. When I try to use it with pulsed/scanning lasers, its hopeless without sync pulses supplied from the laser. The camera, if free running, samples in a manor analogous to lissajous patterns, but in intensity, not space. I still have three sliders on the Spiricon to set video digitization settings, so my subjective judgment can alter the beam profile.

There are some techniques for extracting edge threshold from a repetitive pulsed wave across a photodiode, ie constant fraction techniques. However you still need variable gain on how you process the photocurrent, and therein lies the issue. This would not deal with a single shot or single flash effect.

I wish it was that simple, really I do, but 10 bits does not work.

The simplest analog MPE meter ever made worked with a single wavelength, and stored the differentiated pulse in a carefully designed inductor that mimicked the eye response curve. A slow microprocessor then grabbed many pulses, measured the timing, and did the math. It then dumped the inductor as needed.

Steve

[The_Doctor]

Ok. Silicon IS complex, no doubt about it.. But how about this:

Some simple chemical film that is adequately sensitive to gentle destruction, and simple to characterise for MPE measures, then have some CCD watching that in some small chamber? If you can take bafore/after images that way the CCD need not take direct stress, but only take images of that film the way 'nature' intended.. While I don't doubt that this is no way to replace observations of the innards of the average eyeball, might it not be consistent enough to read for MPE given simple aperture limits on entry to the chamber?

Edit: Bit more thought on this one... If a fast photodiode detected the pulse, an otherwise shuttered-off CCD (with imaging lens) could have its shutter triggered open to take an image of the sensing film. It would have to be a very fast shutter in case a new beam arrived early enough to threaten the CCD, but the idea is that it might keep a CCD safe from the incoming light more often than not, and each capture prompted by the fast photodiode would be the 'after' image, and the 'before' for whatever came next. Software comparison of the images might be fast and maybe also easy if the sensing film was useful enough to do this. There may even be retinal cells made in labs that can help with that film, but I bet there are higher priorities for those right now..

[Laser Electronics Ltd]

A good rule of thumb here is that if damage occurs to ccd (charge coupled devices ) then it is not eye safe.
Unlike earlier laser devices the point source light isn't dissipated, the reason more issues are now occurring is because short cavity laser diodes produce higher levels of Ir and Uv between electron emissions of E1-E2 during the decay stage that causes phase conjugation from e1 to e2 during the decay rate for photon emission that isn't being controlled.

In Ion lasers you would use a magnetic field, We cross modulate and phase modulate the pump source at substrate level in all our systems, the only drawback is for light show use, we have to make the beam slightly bigger.

[Laser Electronics Ltd]

Hi James
In answer to the Why is this so? I'm surprised you asked that.
The generated light source is expanded and then re-collimated after the DLP.
A DLP device doesn't act as a diffuser but a multi segregated mirror deflection device.
Therefore the coherence maintains and this is why if multiple mirrors are engaged within the DLP that allow light emission, the MPE is exceeded.

[JStewart]

It was a rhetorical question Clive.

I’m afraid your assumptions and statements are wrong; the retinal thermal emissions from such devices are nowhere near the levels to cause the harm that even a low power laser direct exposure could result in.

Just this week there has been a five-day meeting of the IEC Optical Radiation Safety and Laser Equipment group in Frankfurt. Experts on this subject, from the world over, convened to discuss a whole range of issues. The topic of LIPs was covered in several sub-meetings. There are two very relevant facts from these meetings:

1) In the soon to be published new version of IEC 60825-1 (Ed.3), there is a radiance exemption clause. This exemption comes about because the harmful aspects of the original laser light source(s) is destroyed in the process that these LIPs devices, (which includes the Casio type device), use to create and project an image.

2) The other thing to come out of the meeting was further progress in the mature draft for a new safety standard for such projectors, which among other things, serves to differentiate these image projectors from more hazardous flying spot projectors that have more similarities with traditional laser light show projectors. The optical output characteristics of these LIPs devices are so far removed from traditional laser sources, that the new standard is not even part of the ‘60825 (“Laser”) series of standards and technical reports.

If what you were saying was correct, then there would of course be no need for the IEC to define the exemption or the completely new safety standard describing LIPs methods of projection.

James

[Laser Electronics Ltd]

Hi James

I didn't make any assumption, my statements were based on work done.

The retina absorbs coherent emissions and as such a point source light that is polarised in one axis H or V .

I'm aware of the meeting and as such your also aware of the arguments if you attended.

The lips (laser that illuminated a micro display area or laser illuminated projection system) also uses a DLP (digital micro mirror device or digital laser projector) these systems when viewed are way above the MPE levels in their current state.

As I've said before give me a call to discuss this.

Thanks

Clive.