Open can and blu-ray DVD diodes…

I have acquired some more laser diodes for testing their suitability for holography.

The first thing to try out was to measure the mode spectrum of one of those high power “open can” DVD burner diodes, which can do several hundered milliwatts when tortured. Well, the news is –not unexpected– that those diodes are totally unsuitable for holography. see here. Moreover I found the the familiar Sony SLD1236VL and SLD1239JL-54 diodes a little bit better, but not too much.

Summarizing, for all DVD diodes I tested the rule seems to be that they can run single mode at up to 70-90mA, which yields approx 20-30mW. There may be exceptional stable spots higher up, eg. I got approx 50mW with one Rohm diode, for example. The best diode so far was the Mitsubishi ML101J27 which got up to approx 80mW (with some weak other modes; true single mode up to approx 50 mW). Thorlabs sells it as single mode diode, probably for good reason, however there is no mention of this in the data sheet.

There is a number of declared single longitudinal mode diodes, esp from Opnext/Hitachi, some with powers exceeding 100mW. Most likely they would be the way to go, I will try to get a handle on them and do some measurements; unfortunately the more powerful ones are quite pricey. But it seems there is no easy way around that if one aims for more than 30-50mW.

Note added: I now checked a GH04P21A2GE/PHR-803T “blu-ray” diode and the results are here. In a nutshell: totally unsuitable for holopgrapy!

Advertisements

16 Responses to “Open can and blu-ray DVD diodes…”

  1. D Says:

    Do you think the behavior is similar for all diodes of a certain type and brand, or can it vary? For example, could it be that one Sony SLD1239JL-54 is unsuitable for holography, but another one would work? If one wants to try this, would it be best to go for the diodes you have found having some stable zones with SLM?

  2. rxlaser Says:

    Hello D,

    I’d think the behavoir would be similar for diodes of the same type, but I really don’t know. It may well be that one diode is better than the other; I have tried several different Rohm diodes and I found only minor differences, but those where from the same production batch.

    Of course the precise location of the stable zones will vary from diode to diode, and will also depend on how it is mounted. I would say that for the Sony’s stay below 70mA and for the Rohms below 90mA as a first rule of thumb. The one Mitsubishi I tried was definitely better but it has died … I am about re-running the open can ones in addition.

    I wonder whether the Opnext/Hitachi ones would be better. Some are claimed to have single mode at 90mW or more; however they are quite expensive and as far as I can see, the affordable ones don’t have much more power than the Rohm’s & Co.

  3. D Says:

    Yes, it would be interesting to find out whether Opnext´s 642 nm diodes are suitable or not. But as you say, they are expensive. There are also short open can red diodes in the SF-AW210 and GGW-H20L sleds that may be interesting to try.

  4. D Says:

    Dear W.

    Can you please explain the mode analysis spectrum on the diode mode analysis page? You say, under the Spectrum analysis header, that blue color indicates single mode behavior, and that grey or black represents noise. But then you say that the results for the Sony SLD1239JL-54 show that it is not suitable over 70 mA, though there are lots of blue in the chart for currents above 70 mA up to 130 mA . Is this due to noise, or why isn´t the blue ok in this case? Or is it light blue that represents single mode here?

    Regards
    D

  5. rxlaser Says:

    Dear D,

    the colors are not consistent between different pictures, they are just a means to visualize the variation of the “combined line width”. For the two Sony diodes green and cyan represent single mode regions. The noise is superimposed as shades of grey.

    However one should keep in mind that the boundary between colors is to some extent arbitrary, for example in the green region the power is simply lower as compared to the cyan region, and thus the peak in the spectrum analyzer becomes smaller and thus thinner, so it just looks as if the line width would be smaller.

    In fact, especially at higher powers, the variation of line width is almost continuous, because new modes often emerge with low power and so don’t necessarily add a large contribution to the total line width; which means that the latter increases only by a small amout. See the two plots for the long open can diode, you see that when more and more modes emerge, the total line width increases gradually.

    Thanks for pointing out this potential confusion, I will try to make this point more clear on that page.

    Best,
    W

  6. D Says:

    Ok, thank you for the clarification.

    D

  7. D Says:

    W, there will eventually be a group buy for the 150 mW 642 nm Opnext for $100 that may interest you:

    http://www.photonlexicon.com/forums/showthread.php?t=5455&page=3

    http://www.photonlexicon.com/forums/showthread.php?t=7931

  8. rxlaser Says:

    D, thanks for telling, I signed up as well! Hopefully this deal will go through. These diodes are supposed to run single longitudinal mode, and even if they could to 100mW only in this way, this would be tremendously interesting – also because 642nm is much better for color holography than 660nm.

    I am about finishing a couple for ready-to-go turnkey laser heads and was about putting in those Mitsubishi ML101J27 which do close to 50mW, but 100mW 642nm diodes would make much more worth the effort.

  9. D Says:

    I´m excited to hear that you have almost finished laser heads. Sounds promising.

  10. D Says:

    W, according to the data sheet the HL6385DG is supposed to be single longitudinal mode, but how certain is this, and under what circumstances does this apply? Do you know the specs for the diodes you have tested so far? What mode are they supposed to have?

  11. rxlaser Says:

    Hi D, that’s exactly the question – what does single longitudinal mode mean precisely. It is almost certain that thist does not apply to the full operating range, but it is unclear to what range. Probably this even depends on the particular diode.

    If that is of any significance, I tested two of those Mitsubishi diodes that Thorlabs lists under single longitudinal mode, despite there is no statement in their data sheets. And indeed these were substantially better than any other diodes of the same power, notably the Sony ones. It boiled down that a diode spec’d at 130mW CW runs single mode up to 50mW approx. To complicate matters, there is some kind of “almost single mode” up to higher powers, like 100mW or so. Means that the spectrum is predominantly single mode but there are weak but numerous extra modes. So again, what does single mode then mean…

    Things may be similar for the Opnext diodes, but I don’t see any way of knowing that before trying out. At least these diodes have single longitudinal mode mentioned as a feature in the data sheet, and this is not so for other Opnext diodes.

    I really hope that this deal would work out, but I am afraid it will be hard to find people for 50 diodes. Well, I’d think they would sell on ebay like gold nuggets…

  12. D Says:

    W,

    I´d like to discuss your Optical Spectrum Analyzer. I think this looks interesting. Have I got this correct that it is basically a grating that reflects the laser light, and depending on the wavelength the angle of the reflected light will be different. You can then use a CCD to capture the light, and by using a long throw you can discriminate between the narrow modes from a diode laser. So, basically this is a more refined version of Hans´ DVD method, when he just looks at the visual output by eye to search for single- or multimode behavior?

    You write:
    “Probably a device with similar properties can be built simply by using a DVD as grating, a parabolic telescope mirror and two razor blades forming the slit.”

    Is it necessary to use a slit before the grating? Why is a parabolic mirror necessary? Isn´t a flat mirror possible to use, and then tilt the grating to get the light on the CCD? I´m uncertain about what is necessary to get enough light and resolution on the CCD.

    Do you believe it is expensive to build a similar device? It would be great to have such a Spectrum Analyzer coupled to a diode laser to dial in single mode operation.

    Is it possible to see noise or is it only the larger modes that are visible? How fast changes are visible?

    Regards
    D

    • rxlaser Says:

      Hi “D”,

      “….So, basically this is a more refined version of Hans´ DVD method, when he just looks at the visual output by eye to search for single- or multimode behavior?

      Well that’s simply the way grating based spetrum analyzers work..

      “Is it necessary to use a slit before the grating?
      Why is a parabolic mirror necessary? Isn´t a flat mirror possible to use, and then tilt the grating to get the light on the CCD? I´m uncertain about what is necessary to get enough light and resolution on the CCD.

      Slit and parabolic mirror greatly enhance the resolution. The mirror focuses an image of the slit onto the CCD, and if the latter has, say 10um pixel width, then you need some good imaging optics to make the image that small in order to get full resolution. Without slit the image will be a few mm across, and the resolution goes down correspondingly, by a factor of 100’s say.

      Since one can see mode broadening by the naked eye off a DVD without any slit, one obviously should be able to see something with a CCD as well, and one may be tempted to do away with the slit etc. However I believe the following: when a diode runs strongly multi-mode, there can be like 10 or 20 modes at the same time, and perhaps this very broad spectrum is then what one can see with the naked eye. I doubt that one can separate two lines, say, by the naked eye. But it’s a simple matter to compute given the known formulas, to be sure.

      “Do you believe it is expensive to build a similar device? It would be great to have such a Spectrum Analyzer coupled to a diode laser to dial in single mode operation.

      Well this depends where you get the grating and mirror from; I think I paid for mine like $80, which was a lucky deal on ebay, actually this was a spectrometer with a scale in the 3um range and thus probably was considered useless for most. But the grating and the mirror were just right.
      Then there were costs for the CCD, other components, PCB and so on, for perhaps again the same amount, but I also used IC’s I had anyway. I had given a copy of the board to a friend and he managed to get the same spectrometer model from ebay as well, so there are two versions running 😉

      A scanning interferometer should be easier to build, though, and has better resolution. On the other hand it doesn’t present the spectrum of a diode so nicely. I think I still have a bunch of the necessary mirrors.

      “Is it possible to see noise or is it only the larger modes that are visible? How fast changes are visible?

      Noise indicates a change of mode structure (due to a minute change of amplitude) and this can be very fast. So one does not see noise but rather a quick jump from one frequency to another, or to several other ones. That goes really quick, the scan rate is like a few thousand times per second or so, thus it looks instantaneous. It is instructive to check the noise in parallel, then one clearly see that with any change of mode profile there is a brief noise burst, but there can also be permanent noise when the spectrum is chaotic (very many lines).

      I think by just monitoring the noise goes already a long way; while I know that situations can exist where you have eg. 2 lines lasing and zero noise, monitoring the noise is still a good indicator for finding unstable regimes, and with a good “single mode” diode the useable regions are quite broad and one should be able to find them but just observing the noise. As far as I can see, the following procedure should be sufficient for eg the Mitsubishi diodes: cool down to 16 degrees or so, then rise the current until you get in a noise transition region. Then go down with the current by 10% and then there is a good change you are in a stable single mode region.

      BTW just tommorrow (strictly speaking, later today), I am supposed to receice those Opnext 150mW 642nm single mode diodes and first thing to do after work will be to check their spectrum 😉

  13. D Says:

    Thank you for your elaborate answer. I now understand the idea with the slit and the parabolic mirror focusing the slit reflected from the grating onto the ccd, and how this enhances the resolution.

    W wrote:
    “It is instructive to check the noise in parallel, then one clearly see that with any change of mode profile there is a brief noise burst, but there can also be permanent noise when the spectrum is chaotic (very many lines).”

    How do you monitor the noise?

    W wrote:
    “BTW just tommorrow (strictly speaking, later today), I am supposed to receice those Opnext 150mW 642nm single mode diodes and first thing to do after work will be to check their spectrum”

    That was fast. I´m excited to read about your findings.

  14. rxlaser Says:

    H D,

    “How do you monitor the noise?”

    Shine some attenuated part of the beam on a photodiode, coupled to a sensitive AC amplifier. Some people even use an audio amplifier and listen to the noise by a loudspeaker. I haven’t tried that, I use simply a scope for real-time measurements and some signal processing circuits when doing automated measurements via computer, but that’s kind of an overkill.

  15. D Says:

    Ok, thanks.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: