You are currently browsing the laser tag archives.
So yesterday morning, one of my webpage articles got liked from Make: Magazine’s online site.
Holy crap. It was like the world was ending.
My personal webpages are all saved and served from the servers at my work. Server space isn’t a problem, there’s dozens of gigs free, and I can always install another hard drive if I want. Our internet connection is a T1, which I never thought would be a problem either. Wow was I wrong!
It’s like waking up one morning to discover a major freeway has been rerouted to go to your office’s front door. Instead of the normal 20 or 30 visitors an hour, you’ve got 50 people a minute trying to come through the doors. Traffic jam! Now nobody can get in, and customers are complaining!
So, I had to shut my whole site down. Later I thought hey, I have free webspace from my ISP from home, I’ll use that. I set up a mirror of the one page everyone and his brother was trying to view, put a redirect in Apache, and restarted my site. Geocities pulled the plug in about 5 seconds. I don’t know what their rules are, but seriously, I checked the Geocities site was working, then I set the redirect, and when I went to look at the site, it was offline.
So then, I don’t know what the hell I was thinking, I thought hey set up a mirror at home here. I have a linux box with a web server program, and a dynamic ip tracker set up. So I mirrored the page, changed the redirect, and was nearly knocked off the internet by the incoming surge.
As soon as I did it, I was trying to un-do it. It took me 15 minutes to turn off the redirect, and by the time the traffic was gone, I’d logged over 2000 http requests on my home server.
I checked the logs at work and saw over 12,000 hits were attempted on the web page in question, during the few hours in the morning before the I suspended the site.
So last night, about 9pm, I figured the dust had settled enough and it was well after business-hours anyhow, so I re-activated the site. Things were ok over night, it logged another 2000+ hits, but that wasn’t too bad. Then this morning, pow, everyone was at it again. And the redirect wouldn’t work any more!
Yeah, you folks out there are too clever for that! I redirect the one web page, but Geocities crashes it, so you grab the text of the page off Google’s cache then start grabbing the images one by one off my server!
So I had to yank the whole site again. Today I’ve got it keeping count. From 9:30 this morning to 4:30 this afternoon, so far there’s been 10,400 load attempts. Combine that with the 14,000 from yesterday. That’s more hits, in two days (well, day and a half really) that I’ve had on my whole entire website since I started the counter, back in 2004!
Looking through the log, I see that it’s not just Make, there’s a bunch of other folks too, who saw the link in Make and have linked to it as well. Someplace in Spain. A bunch of peoples’ blogs.
I don’t know what to do! I want everyone to be able to see the page and enjoy it and all, but I can’t let it affect the company or our customers. And if the traffic is too high for something like Geocities, what then? I mean, you’re swamping a T1! I don’t have buckets of money to spend, so if anyone knows a high-volume free or cheap hosting solution, I’d love to hear about it.
The page is nothing, just over 1MB with all the pictures. It’s just the amount of traffic that’s the problem.
I’ve already discussed the dangers of cheap green high-power lasers in another article. Now, I’ve fully disassembled that same laser in order to learn more about how it works, what it does well, and what it doesn’t. I’ll also hopefully find ways to fix or remedy some of the problems.
To review, here’s a photo of the laser as it was when it first arrived:
This laser exhibits the following problems:
- Came with no IR filter – lots of IR leakage! (fixed as per my other article with an IR filter)
- Starts off with a lot of output power but quickly drops off very low.
- Output is very messy, at 10 feet, the speckle is at least 6 feet across – lots of power wasted in speckle instead of the main beam.
- Output of the laser has a large line through the dot, like dirt or a scratch somewhere inside.
The following image is one I took of this laser going from my livingroom to a bedroom. From the laser to the doorway, the distance is about 13 feet. Note, to the right of the visible beam, you can see green light along the wall and onto the edge of the sofa. Now look to the left of the beam, to the coat tree. Just to the left of the coat tree, just below the lower edge of the red jacket, you can see some more green light. That light is speckle, emitted from the laser. It’s not a TEM multi-mode, it’s just wasted energy, coming out at all angles and cannot be collimated. It is not dust on the lens.
When shining this laser at a blank white wall, the dot is surrounded by a ‘galaxy’ of little dots. It’s all technically output though, and if all that junk were part of the primary beam, the primary beam would be that much stronger. Instead, perhaps 15% of its power is wasted. It’s still ‘output’ though, along with the IR that you get without a proper filter.
And now, on with the dissection… I don’t have any photos of how I got the laser module and circuit board out of the shiny metal tube. Suffice to say, it was a destructive process and the shiny metal tube did not survive.
The Laser Guts
Here is the complete module. From left to right, the major components are: Green circuit board (with black pushbutton); Knurled brass diode holder; Brass lens barrel; Chrome aperture. The aperture part unscrews very easily.
You can see at the back of the circuit board, the stubs of red and black wires that I had soldered to the board when I had this laser module mounted inside a project box. You can also see the trim-pot that can be used to modify the output of the laser. Bear in mind that the output of this laser is not stable at any power level.
To unscrew the lens barrel from the rest of the laser, I gripped the knurled part tightly in pliers then gently but firmly unscrewed the barrel. If you are doing this yourself, start very slowly, like 1/10th of a turn, and verify that the diode and circuit board are not moving relative to the knurled ring or lens barrel.
Without the lens barrel, the laser still works fine, but the beam collimation is much worse – it starts off as a very tiny pin-prick but at 10 feet its about 2 or 3 inches across. The lens barrel contains two lenses, the first expands the beam even wider then the second lens collimates that…more or less.
Here you can see the laser working just fine without the lenses.
Once the lens barrel is off, you will see inside the end of the laser module a smaller threaded brass tube. This tube holds the MCA (multiple crystal assembly) in place infront of the pump diode. The MCA holder just unscrews and comes out.
From the back, the laser diode appears to be a standard 5.6mm can format. When assembled, the case of the diode and the brass assembly is at +3vdc.
I haven’t bothered to map out the circuit diagram, or to check the current the board draws. It is pretty straighforward though – a button, a trim pot, one resistor, two capacitors, one transistor and one component which looks like a voltage regulator.
From the front, the pump diode does not look at all like a 5.6mm can. In fact there’s no can at all. It’s basicaly just a bare diode in there. It is tiny, you can’t even see the diode in this picture. The diode sits on the lower side of the brass ‘rectangle’ bit, between the two pins. The square bit below all that is the photodiode, used for feedback.
Also note that there is no lens here to focus the pump diode – it is completely bare. The divergence from a bare IR diode like this is measured in degrees, not mrads. When activating the laser in this form, at about 3 inches the 808nm light is about 1/8 inch thick and about 3 inches long.
I tried to measure the raw 808nm output to see what it was producing. With 2.4vdc input (two NiMH AAA batteries) I was able to measure just over 300mW of IR on my LaserCheck meter. If I had a good 3 volts going in, there might be up to 500mW of unfocused, uncollimated IR available.
Here is a close-up of the input side of the MCA. The grey gunk is glue that holds the crystal in position. The purple of the crystal is, I think, the Vanadate responding to the flash in the camera. Note also that the purple is not consistent. You can clearly see lots of imperfections on the face. I am pretty sure these imperfections cause some of the serious speckle and poor beam quality that this laser exhibits.
The poor quality of this image is due to the magnification. This component is about 5mm in diameter at most. The MCA itself is perhaps 1.5mm x 1.5mm x 3mm long.
You can see the output coupler side of the MCA here. This is the KTP crystal. I think the green ‘glow’ is a result again of the crystal interacting with the flash in the camera. To the naked eye, the crystal looks slightly pink when looking through it. This side of the crystal looks a lot nicer than the input side.
This is the side of the lens barrel that faces into the MCA. The lens you can see is just cheap plastic, and it is held in place with glue. You can’t get a good idea of depth but the lens is about 8 or 10 mm down into the barrel.
What you don’t see here is the IR filter, because there isn’t one. If there was though, this is where I would have put it – after the MCA, before the lenses. That way nobody can remove it without complete disassembly.
The external lens is also cheap plastic, and also held in place with glue. This makes it really hard to adjust the focus / collimation. By really hard, I mean, you can’t. I was able to remove the output lens as the white glue was kind of crumbly, but you can’t focus it, you just kind of push the lens around and hope for the best. If it were threaded you could at least tweak it till it was good.
I had fun taking the laser apart and learned a lot more about it – what it does well, what it doesn’t.
As for the problems listed above, here’s what I’ve determined:
- I added an IR filter into the chrome aperture, as per my other article.
- I don’t know the cause or solution to the power drop-off yet. It’s either a diode problem or an MCA problem. More research will tell me the Why. I think it’s the diode though, shifting frequency as its temperature changes.
- The speckle is definately because the MCA is bunged up on the Vanadate side, and maybe bunged up inside as well. Basicaly the MCA is low quality, IMHO.
- The line across the output is because the pump diode is not shaped or focused! The laser outputs a line because the diode is feeding a line into the MCA.
Now, if I were to go and buy a new MCA that would clean up most or all of the speckle, and maybe add a lens to focus the pump better into the MCA, and then maybe devise a better set of lenses for shaping and collimating the beam… As for the power dropping off fast that probably would require a new pump diode anyhow. By this point I’d have built a whole new laser from scratch. Cheaper and easier to go buy a good laser in the first place.
That’s about it for now. I hope you’ve all learned some more about cheap green high-power lasers.
The Fine Print
Information provided on this page is based on my own personal experiences.
For no reason at all, here’s a picture of the three red lasers I made out of computer DVD burner drives.
More tinkering with lasers. I’ve pumped up my DVD laser even more. A better lens gives it higher output. From 180mW to 205mW, on week-old batteries.
Enough power, when focused, to cut through electrical tape in 6 seconds. Enough power even, to burn the wood-like-finish of my desk!
Stephanie likes lasers.
Here’s a great link, everything you want to know about how laser diodes work.
Laser diodes are super tiny but can be surprisingly powerful. Find out how, and why. Warning: Linked site may contain science, physics, and education.
Photo from Wiki