Monday, August 8, 2011

Laser power testing


Several months ago I set out to measure the powers of all my lasers. What seemed like a simple process became a protracted hassle. As with my Dirod electrostatic generator I made a number of mistakes. First the DVM module that I had on hand wouldn't work so I had to buy another one. Next I discovered that the power sensor and DVM grounds have to be separate so I had to come up with another power supply for the DVM. Then I discovered that the IR output from the DPSS lasers was so high that I needed a filter in order to measure the visible output alone. Finally, I checked my meter's sensitivity using the supplied test circuit. That last thing didn't get done because I was futzing around with everything else. When I finally got around to it I discovered that I needed to use the power sensor's built-in shunt resistor to get accurate readings.

At this point I'm visiting this topic for the third time now. The second post was never finished and the first one has been deleted because all the readings I got were completely wrong. Let's start all over from the beginning...

Test Rig

The first thing I had to do was build a test rig. Here's the cradle before its paint job:

Out the anode
The next thing I acquired was a 200mW Calibrated Laser Power Meter Sensor from Lasersbee. You simply attach it to a Digital Voltage Meter and the reading in millivolts is the power of the laser. The power sensor is calibrated for 650nm lasers. For other wavelengths a conversion chart is provided. To make things look cool I bought a back-lit DVM module from Jameco. I found that the sensor and the DVM module need to have separate power supplies to work properly together. Right now I'm using batteries. For the sensor I got a lithium 9V from the local Batteries Plus store. It should last forever as far as this project is concerned. The DVM module is running off of AA batteries.

The sensor itself appears to be some sort of photovoltaic device. It's relatively small so it can be tricky getting the entire beam aimed at it. Unlike gas tube lasers, diode lasers have an output beam that's rectangular rather than circular. The beam has to be collimated to bring it to a point. My 405nm Blu-ray diode laser is so poorly collimated that it isn't possible to get the entire beam on the sensor even with it rotated 45° (making it go corner-to-corner)

Laser Pointers

Laser pointers are a very popular item these days. They're cheap and the green ones are very bright to the eye. I got mine from DealExtreme, one of those "buy direct from China and wait four weeks for it to arrive" places that have sprung up over recent years. Despite the delay everything I've bought from them has been a really good deal. Note that they do offer a lot of junk so caveat emptor. (I don't think I'll be buying any of their condoms.) There are also tons of laser pointers for sale on ebay but it's a crap shoot. Most of the sellers say "5mW" and then lead you to believe it's actually a lot higher. They often take a 5mW and crank the power up. It works—for a while. Take a look at their feedback before you buy. When it comes to diode lasers you typically get what you pay for.

In some countries you can't import lasers at all so check before ordering. Here in the US you can't import laser pointers >5mW so don't bother ordering anything rated higher. You can, however, buy domestically manufactured "laser parts" of virtually any power. See Mad Science For Sale for dealers.

Having said all that, we're about to see that the "≤5mW" rule isn't always followed...

Power Measurements

I used a fog machine to make the beams visible in the following tests. The density of the fog varies so the brightness of the beam doesn't correspond to the actual power. Also, these are 5 second exposures so the digits on the meter change and blur the readout sometimes.

First up is the 405nm pointer with a Blu-ray diode. Note that 405nm is near-ultraviolet so you really don't want to be staring into this one. Here's the 405nm in the test cradle:

405nm ~24mW
This works out to about 25.5mW give or take the error range. The correction factor for 405nm is so high that the error range is quite large, ± a couple of milliwatts. As it happens, a friend of mine tested this laser with his pro-quality meter (with a pyrometer) and got exactly 24mW so I'd say we're doing pretty well there considering the limited accuracy.  And 24mW is a bit higher than 5mW so I definitely need my googles on when working with this one. Shame on Dealextreme for offering 5mW and delivering 24mW.

DPSS Lasers

Next up are the diode-pumped solid-state or DPSS lasers. They consist of two IR lasers and a frequency-doubling crystal. An 808nm laser pumps a 1064nm laser and its output has its wavelength halved to 532nm. The major hazard of DPSS lasers is that their efficiency is less than 100% so there's plenty of invisible infrared light being emitted. At 5mW or less this is not considered a hazard, but at higher powers it is. Remember, you can't see it so you don't know how bright it is or where it is. Because of this, DPSS lasers over 5mW should have an IR filter on the output. None of these cheap pointers of Asian origin have any discernible IR filtering. See this video for a demonstration of the problem. Note that you can also find videos that show how to disassemble a pointer and increase the power, burn things, etc. You're on your own with that stuff.

In order to measure both the visible and IR outputs of my DPSS lasers I had to acquire a good IR filter. In a wild stroke of luck I found a good one on-line for $6. It has some impressive specs:
  • Wavelength—Transmission—Color
  • 470.0nm—97.14% (blue)
  • 532.0nm—97.15% (green)
  • 808.0nm—0.09% (IR/near-IR)
  • 1064.0nm—0.53% (IR)
  • Material: B270
  • Glass in accordance with MIL-G-174
  • Surface quality: F/F per MIL-C-48497A
  • AR coated both sides
  • Filter Requirements: Angle Of Incidence: 0 +/-5 degrees
  • In band transmission: >95% @520~ 550 nm
This filter is perfect. It passes 97% of the green and practically none of the IR. True, I had some trouble finding a way to mount it with the angle of incidence to within ±5°. But I'm going for ballpark measurements here.


It finally occurred to me that any assumptions about the absolute power of the IR emissions from these lasers are unfounded. The relative levels of the 808nm and 1064nm lasers are unknown to me. The Laserbee sensor has two different correction factors for the two IR wavelengths—one greater than one, the other less than one. If the ratio of the two were constant then the calculation could be made.

When I find out more I'll update this article again.

More Power Measurements

Up first is a pointer of Asian origin that I traded a 405nm for. The picture is a bit dim as I forgot to turn all the lights on at this point and the fog is rather thick:

532nm #1
Unfortunately, I didn't manage to get a decent picture with the IR filter in place but the reading was 1.1mV, which about 1.8mW. With a reading of 4.1mV with all three wavelengths combined, obviously there's a lot of IR emission here and the visible power is well below 5mW.

Next up is a Dealextreme 532nm. It doesn't look as bright because there's less fog here:

532nm #2
16.3mV combined.

532nm #2 + IR filter
About 11.2mV (18.25mW) visible. Lots of IR. Say, wasn't this supposed to be "≤5mW"? Heh.

Up next is my AixiZ 50mW OEM diode head:

532nm #3

532nm #3 with IR filter
About 60.6mW visible. Clearly this laser has some IR filtering but I can't tell how much. 50mW is very bright. I had to wear safety goggles to align the beam.

Fortunately, unlike the visible beam, the IR isn't well collimated because of the large difference in wavelength. It's spread out over a larger area which reduces the danger slightly. In this picture, I have the 50mW head shining through my googles, which have an O.D. of 7+ at 532nm and pass nothing but a faint, tiny yellow dot of light. They attenuate the IR as well by at least 40% but a lot still makes it through. Beyond the goggle lens is an IR detector. As you can see, the dot is quite large:

Detecting IR
Green gas tube laser

While I'm working with green lasers, here's a 543.5nm helium-neon gas tube laser:

543.5nm HeNe
That's about 1.8mW, which is nominal for this tube. Tubes use high voltages and glow like a neon sign screaming "Mad Scientist," which is their whole appeal. The power supply for this one is 2600V at 6.5mA. I know I'll never see another large green tube again (actually, a used head went really cheap on ebay the other day but anyway...). I'm still looking for orange and yellow...

Laser tubes have a very small, round (Gaussian curve) bean with a very low divergence. Here's the entire tube in all its glory:

Out the cathode—Click to enlarge
Now you can see what the rest of the test cradle is for. At 19" it's the longest tube in my collection. A red laser tube this size would probably be 20mW. Green is not very efficient. Note that this tube is emitting from the cathode. I have another green tube in my collection that emits from the anode (it's in the first picture at the beginning of this post). It's 0.1~0.2mW. That's tenths of a milliwatt. The beam is so dim you can stare straight into it—not that I'm recommending that...

Next time around—lasers of the visible red kind.

1 comment:

  1. You have a fantastic blog, great information and honestly I am amazed at the setup(s) you have for your experiments. I wish that I had the time and $$ to do similar work.

    - Otto


Comments now require approval by a sentient life form. Prompted by an increase in spam and only after careful and lengthy consideration did the management reluctantly implement this change in policy. We apologize for any inconvenience and appreciate your understanding.