445nm beam collimation/correction

[Solarfire]

I did some updates here (not finished yet), added a chart with power measurements for the tested setups. The same diode was used in the same mount at the same mount temperature (22°C) @ 1070mW for all measurements. It was surprising so far!I finally got around to testing the 405-G-2 collimators!

I still have to do all the cylinder setups, sorry drlava. Just haven´t got much time these days.

Daves mount was a big help here, just the fact that the diode can easily be removed or changed. I reverse mounted the diode witch made is easier to measure the raw diode power, you can get real close to the diode aperture.
Thanks Dave! I´ll be getting some more of those!

I also found out that it will be worth TECìng these. A 15mm x 15mm TEC fits nicely on top of Daves mounts. It was noticed during the power tests, that at =>700mW significant power drop was noticed, after a few minutes of operation, especially when approaching => 1000mW.
Hey Dave how about a 4mm bore direct halfway between diode bore and top of the mount to approximately the center of the mount in depth?

Big surprise in the power tests!

445nm diode beam collimation with Aixiz 405/455 full tread, O-Like 405/445, 405-G-2, mccarrot (Laserextreme) blue, Insaneware and correction with prism and cylinder lens pairs. I’m looking for an optimal lensing setup (for my personal needs) where beam dimension at the aperture is <3.5mm and the divergence <1.5mrad. All of these setups can probably be tweaked a little more but before I get into that the basics have to be given (minimal beam- aberration, artifacts).

The diodes used for these tests were selected according to having a clean and symmetrical raw emitter pattern with the highest uniformity in illumination. Prisms are from the p1t8bull GB and the cylinders from the drlava GB. Diodes were driven at 230mA and the beam was projected onto or thru a black measuring grid with 1mm increments (except centerlines which have extra 0.5mm increments).

Aixiz collimator 405/445nm, price 10$ (just the collimator):

Hover Mouse over picture for short description.

For the price the results are actually pretty good except for the fact that the beam size at the aperture exceeds 3.5mm on the fast axis. Best results was with the prism pair with a setup time <5 minutes.

Mccarrot collimator (Laserextreme) blue, price 65$ (includes 5.6mm mount 11mm):

Hover Mouse over picture for short description.

This collimator was actually very disappointing at 65$ a shot with a minimum order of 4. Here I expected the ultra collimator, instead it turned out they are absolutely useless for this diode. They would have been perfect as far as aperture beam size is concerned but the far field is catastrophic. Mccarrot are you sure these are the correct collimators you sent, the look slightly different than the ones in the pictures? Only the case and not the lens carrier are brass like in the pictures? You got any suggestions? Can I return these for a refund or the correct lens? Unfortunately I’ve got 2 selected diodes stuck in these with thermal epoxy. And yes the results were the same with both collimators.

I managed to clean up the mccarrot setup. It appears there were some issues with reflections inside the collimator housing. I took the mccarrrot lens and put it in Daves mount, major improvement. See the last picture above!

O-Like collimator 405/445nm, price 8.20$ (just the collimator):

Hover Mouse over picture for short description.

Now this turned out to be the biggest bang for the buck all around. I couldn’t believe this one, I redid the tests just to verify that nothing was changed on the initial setup. Highest power, lowest loss, best divergence, cleanest beam, best beam spot, lowest price! Nothing to add here but, thanks O-Like I’ll take another 10,

Thank you!

Take a look at the chart and pictures!

I will defiantly be using some of these in 1, 2 and 4 setups. To get a better clear aperture on the scanners, the diode and prism correction will be rotated 45° and hit the scanner aperture diagonally this will be like having a 3.5mm beam.


Clear aperture on scanners.

Insaneware collimator 405/445nm, price 20$ (just the collimator):

Hover Mouse over picture for short description.

This one has pretty good beam properties but highest loss which makes 20$ way too much for this one!

405-G-2 collimator 405/445nm, price 28€ (lens and Modul housing):

Hover Mouse over picture for short description.

As expected with a short FL lens, small beam at the aperture of 3.5mm x 0.8mm but with a higher divergence of 1.5mrad. Excellent low loss of 2% after the collimator and a total of 8.3% after the Prism correction. Needless to say this is the winner under the short FL lenses in price and quality. These short FL lenses will probably have better resultson the beam profil with the still pending cylinder lens setup!
I will be using these in my setups with small aperture requirements and the O-Likes for setups which are not so critical with aperture size requirements.



Power chart!

In general the prism setup is the easiest. I just set the focus of the diode so the beam profile at 11m was about 1m in the fast axis and then rotated the diode until the beam profile was vertically straight. I then refocused the collimator until I had a sharp horizontal line (seems to be slightly converging since the far field beam was smaller than at the aperture) and dropped the prisms on Zoofs template and viola a little tweak here and there and it fit!
The cylinders are somewhat more time consuming to set up and tend to induce slight aberration into the beam profil.

Any suggestions, coments?

[dave]

Power measurements are pretty important here, as longer focal length lenses ( like the o-like ) clip and waste part of the beam..

[Solarfire]

This is true Dave and that will be the next step, but you have to agree that a beam is useless if it has a half of meter of trash around the collimated beam profile. I’m sure the mccarrot collimator, with a very short focal length, captures the entire beam but as you can see the beam profile is useless. Aside that, my power meter has bitten the dust and I’m getting a new one. I’m also open for suggestions on good lenses to test.


Frank

[mccarrot]

Hi Solarfire thanks for the test!

I'm sorry you are not satisfied with my collimator.

But please make your final verdict after the power tests.

I also don't understand why you find 1.5 x 3.5mm x 1,32mRad absolutely useless???

While you are satisfied with a fat 3,5 x 5mm beam where a lot of power is waste in the collimator AND at your scanners!

Did you know when you have 6mm wide galvo mirrors and you bounce the beam off at 90 degrees the effective area is 4,4mm, when the galvo's start to move and you bounce off a beam at a 45 or 135 degree angle (widemoves) you only have a effective area of 2mm!!! with a 5mm beam you are wasting 60%! of your power at your galvo mirrors)

So in my opinion the 3,5mm x 5mm beam is less usable than the 1,5 x 3,5mm you get with my collimator. unless you are planning to use it as a static beam only. Also most green lasers got a beam of 3mm, so with a 5mm fat beam you get a blue beam with a white centre (just like we now have with 635nm red, red beam with white centre.) because the diameters do not match.

Now about the divergence try to get 532nm green laser with +500mW to match your 445nm with a divergence of 0,59mRad. most green DPSS lasers got a divergence arrounf the 1,3mRad, so a perfect match with my collimator.


So looking at your first test results I find my lens a much better choice for a lasershow projector.

[mccarrot]

This is true Dave and that will be the next step, but you have to agree that a beam is useless if it has a half of meter of trash around the collimated beam profile. I’m sure the mccarrot collimator, with a very short focal length, captures the entire beam but as you can see the beam profile is useless. Aside that, my power meter has bitten the dust and I’m getting a new one. I’m also open for suggestions on good lenses to test.


Frank

The "trash" arround the beam is cutoff by you galvo's and is so weak compared to the centre beam you not see it when scanning.

another important factor to keep the beam as this as possible is brightness.

a 1W 3,5mm beam is much brighter than a 1W 5mm beam.

here are some examples I have stolen from the arctos website:

Power beam dia. Divergence brightness
1W 1,5mm 0,8 0,70
1W 3,5mm 0,8 0,13

(source: http://www.arctoslaser.de/en/arctos_arguments.htm )

thats why those OPSL's looks so damn bright, small beam diameter.

[edison]

Hello solarfire

I did the same tests with twenty lenses on both prism pairs and cylindrical lenses. I have tested several lenses with different FL,s and NA,s. Here,s a list of the lenses i tested:

Kvant,
Mccarrot
O-like red and blue
aixiz
daedal
laser4u (red)
meredith
And other lenses with different FL and NA,s

The bottom line is that with a very short FL like 4mm the beam at aperture after the prisms is 1 by 3.5 mm. But the divergence is bigger. When using a FL like 7 or 8mm like the aixiz have then beam at aperture will be bigger but the divergence is lower.
I think it is best that the beamsize and divergence will match the greens chinese lasers
With a divergence of 1.2 til 1.5mrd its fine.

[Solarfire]

Hi Mccarrot,

As stated above beam size and power was actually not an issue at this point, there is absolutely no doubt that your lens delivers superior power. The beam size at the aperture and low loss was actually my main reasons for buying this lens. The Problem is the higher level of beam trash which is an issue with graphic projections. With beam shows this would not be an issue. Just guessing at this point (will be measured later) but I would say that probably >5% of the power gain with this lens lies in the trash around the beam. This stray radiated power will definitely give a blue haze to the graphic projection area and will subjectively reduce contrast of the overall picture. The clear aperture of the galvo mirror assembly will not do much to restrict the visibility of this stray radiation, on the contrary the stray radiation will project an outline of the clear galvo mirror aperture area (I will show pictures of this later). Viewing the Picture on the far right above (roughly corrected with prisms) the stray radiation has a very high visibility even on the black surface of the measuring grid.

As to the Aixis lens, I also stated above that aside the fact that the beam size exceeds 3.5mm the beam quality was actually pretty good. Which means it is also not suited for my purpose due to the beam size at the aperture exceeding 3.5mm. Once again power loss at this point was not an issue.

By the way.. By hovering over the pictures with the mouse cursor you can see a short description of what each picture is showing.

@edison

I searched quit intensively on this subject here in the forum (and others) but information to this subject was quit scarce. Did you post any of your findings with data and pictures? A short FL is not always a benefit when it costs beam quality due to capturing parts of the beam which are better left trashed, finding this balance is what I’m trying to do here.

Frank

[kitatit]

Hi guys,
I'm just tagging this thread so I can keep track of it.

Thanks for everyones input, I'm learning a lot from this conversation

Kit

[mccarrot]

I already have this setup running in my projector, and I can confirm you the stray is hardly to not visible during scanning, and you must be a real laserfreak to even find a little bit of stray on the projected graphics. I had to pause the graphics image and walk to the scrim to see a very very tiny bit of stray, which is absolutely not annoying and a normal person (customer) would never ever see this.

There are 2 reasons for this:

1. the galvo mirrors act as 3-4mm aperture window, clipping the rest.

2. Have you ever had dirty mirror, where when you project a static beam and you see some stray around the beam? as soon as you start scanning you will not notice any stray anymore.

So please put a module inside a projector, scan some graphics and then judge if the stray is unacceptable.

[gottaluvlasers]

Excellent tests and results! Solarfire- thank you for taking the time to do these extensive tests.

My question is *still* why nobody is using a pinhole to shoot their beams through?!? Am i not taking something into account with the useage of pinholes? does it add any sort of beam crap?

With the price of blue the way it is- WATTS of power for <$500 (Home brews) why are we not using pin holes to just circle the beam out? Who cares about losing a watt of power (just say 2W in to get 1W out of a pin hole).

Im sincerely wondering why this isnt being used. I am starting to believe that perhaps "pinholing" adds some sort of interference or something in the beam?

Any ideas?

-Marc