Help needed with Divergence measurement

[clansley]

AFAICT I use the full lit area for full angle divergence, and the inner bright area for half angle divergence. But there seems to be three levels of brightness in my spot and there's a strange effect below the spot should that be included in the measurement or ignored?

Spot at 9 meters:
full angle = 7mm, half angle = 3.5mm or 2.5mm if using the middle

Spot at 16 meters:
full angle = 12mm, half angle = 5.5mm or 3mm if using the middle

That would give divergence of:
full angle = 0.71mRad
half angle = 0.28mRad (or 0.07mRad using middle)

All comments/advice welcomed.

 

[Cyparagon]

Where are you getting you half-angle from? Half-angle is just half of whatever the full angle is. Don't use half-angle. It's misleading and pointless.

Where is the edge of the beam? The most common way to define it is Full width at half maximum, so you might have something closer to <.5mRad.

 

[clansley]

Where are you getting you half-angle from? Half-angle is just half of whatever the full angle is. Don't use half-angle. It's misleading and pointless.

Where is the edge of the beam? The most common way to define it is Full width at half maximum, so you might have something closer to <.5mRad.

Hi Cyparagon - I'm clueless so I was using this laser world page as a reference. It suggests using all the beam including any "blur" - whilst other references talk about taking a point between center and the edge (such as you've done). Thanks for your help, Chris.

 

[clansley]

Not sure if this is a valid way of doing FWHM but I took the pixel value of the center (which I was careful not to over expose) and then measured using those pixels greater than half the value. I.e. if the center was 255 green, then I measured those pixels with value 128 green and above.

Using this I got 5mm for the 9meter dia, and 10mm for the 16meter dia. This gave 0.71mRad.

 

[Sigurthr]

Yep, use FWHM (though I've always called it "the 3dB bandwidth" - same thing), which for noncritical measurements is good enough to eyeball as the "center spot". In your 9M photo that would be the 3.5mm measurement.

I wouldn't go off of pixel intensity because you're almost always going to see saturation of the "core" where nonlinearity enters because of the camera. This, of course, isn't true to life. Laser beams exhibit gaussian curves when looking at the intensity plot.

Btw I've been using http://www.pseudonomen.com/lasers/calculators/mRadCalculator.html for divergence calculations, and looking over your posted images I would use the following values:
9M: 3mm
16M: 6.5mm
yielding 0.49mRad, which coincides with Cyparagon's assessment.

 

[clansley]

I'm beginning to doubt that photographs are the right way to go as below is the same spot (2cm from laser opening) but with different camera settings:

camera in auto mode (causes center to be overexposed):


camera in manual mode, adjusted until the center is not over exposed:


AFAICT each image would result in a different measurement.

 

[clansley]

Sorry guys - I'm still not understanding how to do it by eye...

Ok - the first image is over exposed so isn't useful for working out FWHM, but the second should be as the center isn't over exposed.

I've plotted the green pixel value and lined it up to the image, I'm guessing as my beam is a rectangle it's no surprise it's flat at the top (it shouldn't be over exposed), and we do seem to have half Gaussian curves at either side (ignoring the drops due to the lines on the ruler).

So FWHM for the above image would be 6mm?

 

[Sigurthr]

See the flat topping of the waveform there? Even though it isn't overexposed it is saturated, and that's the problem. You can adjust the settings all you want but without using ND filtration you can't desaturate the sensor.

Don't use a camera, use your eyes. Just look for the brightest central region and use that. If your eyes are saturated you'll have the same problem, so use eye friendly ND filtration: laser goggles!

Also, if the module is focused so that the beam is convergent with a positive focal length you'll see a wider beam at the aperture or near it than at a distance that's beyond the focal point.

 

[clansley]

See the flat topping of the waveform there? Even though it isn't overexposed it is saturated, and that's the problem. You can adjust the settings all you want but without using ND filtration you can't desaturate the sensor.

I don't really understand why wouldn't increasing the shutter speed avoid saturation but I'm happy to accept it for now :thinking:

Don't use a camera, use your eyes. Just look for the brightest central region and use that. If your eyes are saturated you'll have the same problem, so use eye friendly ND filtration: laser goggles!

Will do, I will also try the camera through the glasses to see how it compares to what I see. I fear that my OD4 glasses may filter out too much.

Thanks for your input.

 

[Sigurthr]

To be completely accurate it is possible to eliminate the sensor saturation with settings, it just isn't worth it in this application because there are so many variables. You have to account for the sensitivity of the particular sensor to the particular wavelength and take in a ton of other variables (such as target photometrics) to do it on the fly. It's a lot easier but more labor intensive to keep analyzing shots and going back and adjusting until there's no more nonlinearity present.

In the end you have to ask yourself how much precision and accuracy do you really need for the job. Are you binning laser systems or doing research optics alignment? Chances are that's not the case.

 

[clansley]

To be completely accurate it is possible to eliminate the sensor saturation with settings, it just isn't worth it in this application because there are so many variables. You have to account for the sensitivity of the particular sensor to the particular wavelength and take in a ton of other variables (such as target photometrics) to do it on the fly. It's a lot easier but more labor intensive to keep analyzing shots and going back and adjusting until there's no more nonlinearity present.

In the end you have to ask yourself how much precision and accuracy do you really need for the job. Are you binning laser systems or doing research optics alignment? Chances are that's not the case.

Sigurthr thanks for your input on this thread - I'm finally able to rep people so +rep :yh:

I agree that usually it's not worth the effort but I was trying to do my best as a DIY experiment. I hope to do more here when I get some spare time.

 

[steve001]

AFAICT I use the full lit area for full angle divergence, and the inner bright area for half angle divergence. But there seems to be three levels of brightness in my spot and there's a strange effect below the spot should that be included in the measurement or ignored?

Spot at 9 meters:
full angle = 7mm, half angle = 3.5mm or 2.5mm if using the middle

Spot at 16 meters:
full angle = 12mm, half angle = 5.5mm or 3mm if using the middle

That would give divergence of:
full angle = 0.71mRad
half angle = 0.28mRad (or 0.07mRad using middle)

All comments/advice welcomed.

Measuring divergence is arbitrary. Meaning don't worry about full width angle or half full width. It's easiest to measure the entire diameter because divergence will be the same.
Typically I use a piece of picture frame glass and on it I mark +, much bigger of course. I then use a metric rule and mark lines every millimeter on the X,Y axis. I also do this in a well lit room or outside to eliminate blooming. I don't use a camera just my eyes. A white unlined index card can be use too. Typically make the first measurement at 10 meters

 

[lazeristasUVISIR]

Typically make the first measurement at 10 meters

It depends on the beam size. The smaller beam will reach its "far-field" faster.

In principle it is possible to start from the laser and to do spot-size measurements at several (better many) distances and to process with math assuming the beam is Gaussian.