Which Collimation Lens to Choose? A Set of Observations for Your Review

[dkumpula]

Hey Folks,

Last month, LaserCo pointed me to an LPF thread about a couple new collimation lenses that had recently become available and it got me thinking. Most of the collimation lens information detailed in both PL and LPF tends to be one-off comparisons of one diode with two lenses. I decided to try to take some of the lenses I've accumulated along with DTR's popular lower power diodes and start a rough analysis of which lenses pair best with which diodes. What follows below is the information I accumulated over a couple days of testing.

To be fair, this analysis could certainly have some flaws such as the diode not being centered perfectly, the collimation lens being a poor example, minor temperature variations affecting power output, etc. etc. That said, I tried multiple of each type of collimation lens when I had more than one available (all cases except G2LFL), I kept the temperature, warm up times before reading, modulation voltage and other such things relatively constant between readings. I also went back to the G2 lens that I started with as a baseline at the end of testing each diode to ensure that things hadn't changed output-wise. No consideration was given to long-term durability of the lens (e.g. an acrylic lens with 500mW of 638 going through it) or temperature variations affect on the lenses themselves. In any case, I'd take this as a set of imperfect observations that should be indicative of their performance, although your mileage may vary somewhat.

Also, an apology for the table that follows. I tried uploading 5 proper table types (all of which were refused) and using the table function here in PL. Further that tabs don't exist here, spaces are removed from posts and you end up with the awful looking table below. Hopefully, you'll be able to figure it out just the same.

<edit: here's a screen cap if that helps>




*Note, % is a baseline comparison against the G2 lens*
*Note: Size measurements are all at 20M*
*Note: All beam notes are far-field notes except where noted as near-field"

Beam Pictures
Sorry these pics aren't as clear as they could be. Hopefully you can get the general idea along with my comments above.

Osram PLP520 520nm, 120mW


Oclaro HL63133DG 638nm, 170mW


Mitsubishi ML520G71 638nm, 300mW


Opnext HL63603TG 638nm, 120mw


As I hope you can see from the data above, the G2 and G2LFL lenses are (not surprisingly) high performers, but the surprise was the new AixiZ lens. This is an $8 lens. Compare power and beam results against the G2 at $30 and that should get you thinking. . . The acrylic and other glass lenses shined in a couple areas as well, despite their relatively low power.

Anyway, I hope to further this post with lower power blues and additional lenses that just arrived in the near future. In the meantime, I invite you more knowledgeable folks to point out other considerations people should be aware of as they pair diodes with collimation lenses.

-David

[planters]

Wow, that's an impressive presentation. This will be very useful for a long time. Thanks.

I am not familiar with some of these lenses and another critical specification is the near field beam dimensions. This should be related to the FL, but not always if a particular lens is vignetting. This is important to determine the projection optics required, the number of beams that will fit when knife edged and stacked and how much beam expansion can be accommodated.

[absolom7691]

Thank you for this! Quite a good way to fine tune the projectors to get a white beam and not a white beam with some multicolored fringe. VERY usful, indeed!

[logsquared]

Very nice! Thanks for taking the time to post.

Very cool about the 3 element and the g71.

For some reason I thought the G2 LFL was a 8mm Fl lens. Guess I was wrong about that.

[andy_con]

ummm think there is a bit of a floor in your testing.

most of the multi glass lens like the olike type are 6-8fl, where as the g2 lens is more like a 3-4fl. a shorter FL will give much small beam size at aperture and bigger at distance, where as a longer FL will give much bigger at aperture and smaller at distance.

so that information you provide isn't that helpful as it doesn't have true meaning with aperture measurements.

sorry to bring a downer to the thread

also some of the beam spots you show are really really poor for single mode diodes, I don't think the diode was straight in its holder

[planters]

so that information you provide isn't that helpful as it doesn't have true meaning with aperture measurements.

Sheesh! I've already mentioned that the beam at aperture would be a useful additional piece of information, But, to say what has been provided isn't that helpful is not true.

[dkumpula]

Thanks for the kind words, guys. As I said, these are just some indicative observations that would certainly benefit from further measurements. The focus for me (pun intended) was, in declining order of importance: power, beam quality in the far field, and lens cost.

I primarily use single mode diodes in my builds (and PBS paired, not knife-edged), so several of the variables that should be considered for correcting multi-modes, knife-edging or fitting such beams on galvo mirrors wasn't really forefront in my mind as I did the measurements. I also use the G2 lens almost exclusively, so I don't have a lot of experience with the downside of other lenses. That's the pleasure of working with the broad experience base that exists in PL as compared to the more single-minded (pun also intended) laser-pointers builders. I sincerely appreciate the additional factors that should be considered when pairing a collimation lens with a particular diode.

Regarding, near-field, measurements, I totally, hear you Andy and Eric! I didn't think to measure it at the time, but I do understand the need. I noted on a couple occasions where the near-field spot was noticeably large at the short distance that I was doing power measurements (which was after I had adjusted for what I thought was an optimal beam at 20 meters). In the cases where there are no such notes, the beam fit in the 0.5x0.5cm square for my power meter at a distance of ~16cm from the lens. Hopefully that's good enough for getting beams onto galvo mirrors in most projectors, but not considerate of knife-edging requirements, I suppose.

Regarding the picture quality, I agree that they suck. I didn't look at the pics full size until after I finished my measurements. In addition, they were taken with the diodes running at their peak output, so even with the single modes, that's too bright for most cameras without making adjustments that I failed to do. The spots looked much better in real life and I am glad I made notes. That said, the diodes may have been slightly askew in their holders in some cases, but most of those were pressed by DTRs into casings, so I didn't have control over that aspect. (I expect that DTR can do a better job at mounting diodes in those cases, than most folks, however.) I did mention it specifically as a caveat.

I'll measure the near-field size going forward, but I hope the other variables are somewhat useful to others in narrowing down their pairing options for at least a subset of the potential builders out there. I'm also very happy to add to the spreadsheet any observations people wish to contribute; higher power diodes, other lenses, or anything else.

-David

[tribble]

This is the awesome resource I was searching for a few months ago when I was asking about collimators. Sure, there's always another piece of data that would be useful, but this is great stuff, Dave! Thanks for posting!

[edison]

Although the test is not super accurate it shows the difference between the lenses. You see can see that triple lenses will give allot of losses with mitsubishi reds because it clips of the beam since its a longer FL and have triple the losses from a single aspheric lens. The longer FL from DTR is interesting too.

Any info is welcome and the effort is highly appreciated. Thanks dave!!

[frostypaw]

Ah fab work! I think this thread has just saved me asking a question - I've been chasing some mild reflections/halos which seem to be those aberrations outside the spot from the G2 lenses. It seems far worse with the 445nm PL450b - the beam is almost like -o- but at least I can stop hunting the mystery bit of dirt now

It would be interesting to see parts of this experiment repeated with the more adjustable diode mounts to ensure things are bang-on centre, see how much difference it can make.