Two questions about the dynamics of a laser diode

1. Jan 31, 2012

_Shellback

I am using a laser diode to count aerosol particles flowing through a gas stream. The refracted light is being funneled to a photo-multiplier via fiber optics. When there is no gas flowing (and therefore no particles) we are seeing some strange results from the output signal of the photomultiplier. After troubleshooting this problem for more than a year, I am left to believe that there is something going on with the laser diode that I am not able to capture. Is it possible that the light being emitted from the diode is shifting around and causing slight changes in the focal point? What I'm seeing, electrically, is an occasional burst of very small square waves coming from the PMT (photomultiplier). Sometimes it'll last for less than a minute, sometimes it'll last for more than an hour. It starts as occasional spikes (either positive or negative) which get wider and more frequent until the square wave develops and then fades out to more spikes inverse to what it started as. This is my first foray into working with lasers after 20 years with electronics. A more optically experienced hand would be very appreciated.

I would also like to know if the Iop of a diode is posted at a given mA, is it advisable to use it at that current, or should it be operated somewhere between the posted operating current and the threshold current?

2. Jan 31, 2012

fzero

You might get better results posting this in the EE section: https://www.physicsforums.com/forumdisplay.php?f=102

As a theorist and nonexpert on practical semiconductor matters, I can only provide some basic guidance. First of all, the threshold current, $I_\mathrm{th}$, is the minimum current that the diode begins lasing. Below $I_\mathrm{th}$, the diode acts like an ordinary LED. Above the threshold, the power output of the diode is roughly linear with current. The operating current $I_\mathrm{op}$ is the estimated current needed to produce the rated power of the diode.

Because of manufacturing imperfections, there will be some difference between the actual and rated output at the specified current. If you needed a specific amount of power, you would need to test the output and tune your operating current accordingly. For your application, perhaps the actual value of the output power is not critical, but its stability is very important. The power fluctuations you have observed could easily be due to fluctuations in the applied current. So the first thing that you should do is try to measure the amount of noise in the supply for your laser. It might be necessary to modify your input (perhaps with a feedback loop) to stabilize the injected current.

3. Feb 1, 2012

_Shellback

Thank you for your reply and I will certainly re-submit the question in the EE section.

One thing I have tried is to add some of the thermal compound (similar to what is used to couple heat-sinks to computer processors) to the collars and housing that contains the diode. This does not seem to have any effect. It is possible that I was using the wrong kind a compound (it seemed rather generic and I have no idea how old it was), is there any that would be more preferable?

As far as monitoring the signal, I have done so using both an averaging value (to manage the amount of data) over a 24 hour period, and a high speed method (50ks/sec) observing carefully at the time of the episode. I have monitored current and voltage being output from the laser driver board, power input to the laser driver board, and the PMT bias voltage from the sbRIO-9631. All of them are rock steady and show no correlations to the episodes. This has been a year long process of troubleshooting every electrical possibility and cleaning and shielding and rearranging. Very frustrating. The only thing we have noticed, and keep coming back to, is that the problem gets worse or better (but never fully goes away) when we change the laser diode.

Thank you for your suggestions. I'm going to try a faster sampling rate and examine the frequency spectrum of the PMT output next. See if there's something hiding in what little noise is left on the PMT output that might give some indication as to what is going on.

4. Feb 2, 2012

maimonides

Yes.
Laser diodes are multimode and the modes can differ in beam geometry, especially with the very short cavity. There will be a lot of "modehopping".
Maybe you can borrow a single mode, stabilized laser (HeNe?) and test with it. I´ve never heard of single mode laser diodes, but don´t take my word on that.

5. Feb 2, 2012

_Shellback

The data sheet for the diode says:
"a high-power, high-efficient AlGaInP semiconductor laser which provides a stable, single transverse mode oscillation"
attached is an image stolen from Wikipedia which is what I saw after passing the beam through a 150um orifice. I'm guessing this is the "Multi-Mode" you're referring to? So are you saying that this beam will shift slightly over time?

File size:
93.9 KB
Views:
166
6. Feb 2, 2012

maimonides

I don´t think the pattern in the picture has anything to do with multimode.
Seems there are singlemode diode lasers now and I apparently i haven´t watched progress in the field close enough.
(One meaning of) modes are the wavelength´s the laser can emit (determined by cavity size, which may change due to thermal expansion).
The change from one wavelength to another will make a small discontinuity in the emitted light, but probably no change in beam geometry. You can use a Fabry-Perot interferometer to detect these modehops. (The cavity length of my kind of single mode lasers is actively controlled)
Thermal stabilization may mitigate your problems. (I assume you are using enough pressure and an appropriate heat sink.) Have you thought of a peltier cooler and are you monitoring the temperature of your laser diode? Is the diode output power stabilized?
Have you checked with a different PMT? (They have noise and other imperfections)

7. Feb 2, 2012

jim hardy

Back when i was interested in laser diodes i did some reading,,,

and learned that commercial ones have a thermal sensor in the same package that's intended for controlling the input current with intent of keeping it below some temperature where undesirable effects spring up.

if i remember whose application notes they were will post.

Does your laser have a third wire?

probably you're lightyears ahead of me, but thought i'd offer it anyway. No insult intended.

8. Feb 2, 2012

_Shellback

No insult taken, I understand.

The LD does have three pins. According to the schematics, pin 1 - Anode, pin 2 - NC, pin 3 - The cathode of the diode is connected to the package which is connected to pin 3. I'm working on a reply to a previous comment that will have pics and descriptions of what I have done in assembling this thing. Please keep in mind, this is my first laser project, I know plenty about electronics and I understand (for the most part) the theory behind LD's, but it's the real world specifics that I'm looking for in their function.

9. Feb 2, 2012

_Shellback

One of the issues we have is that there is no room for anything. The diode is mounted such that there's really no room for anything else near it. There's the diode mounted inside a collimating lens housing, which is mounted in a centering disk which holds the laser centered in a small tube segment which holds the whole sub-assembly in a z-axis adjustment apparatus, which is mounted onto a X/Y axis adjuster thing which mounts onto an X-bracket that mounts onto 4 rods that connect to the “Cage” where the flow-cell is. Mounted in-front of the collimating lens is a 150um orifice and an ND Filter. I don’t even know where to try and thread a thermocouple to take temperature measurements for the LD. I can say that when the LD is mounted, we squeeze it into the collimating housing with an aluminum heat sink collar and a threaded black ring that pushes the diode against the collar and the collar is then pressed against a small shoulder in the housing. In my mind, this should be sufficient for thermal dissipation. I will attach a series of images to show what I’m talking about with the waveform. All of these images are taken with no gas flowing, no particles, and both ends of the sampling tube had been sealed off.

BTW, questions are awesome. If I have the opportunity to solve my own problem because you asked the right question, that’s fantastic.

Attached Files:

File size:
155.7 KB
Views:
126
File size:
23.9 KB
Views:
131
• IMAG0023.jpg
File size:
15.3 KB
Views:
136
10. Feb 3, 2012

maimonides

First, to give things their proper name: lasers have transverse (sometimes called spatial) modes as well as longitudinal ones. Single mode applied to a laser diode commonly refers to transverse modes; there may still be multiple longitudinal modes. When I said single mode I meant (longitudinal) mode stabilized lasers (used in spectroscopy and metrology).
The "rock steady" output voltage is a bit surprising to me. Uf of a diode is notoriously temperature dependent; I´d expect at least some change in the first minutes of operation until reaching steady temperature. You might want to do your own measurement with some reasonable DVM; maybe you can get at least some idea of temperature changes by monitoring forward voltage.
If you´ve really pinned it down to the LD:
My pet hypothesis still is output fluctuations (because of longitudinal mode competition?) caused by thermal expansion of the cavity.
You can test this by (cautiously) heating (or cooling) your unit from the outside (hot air gun, peltier element) and checking wether there is any effect.
(The next step is to reproduce the effect with a "naked" LD out of your appliance and a different detector? - a photodiode may be good enough).
Some more labeling of the axes in the diagrams might be helpful. Are you recording your working signal on the same scale or is this amplified to show the interference more clearly? Time scale is the 50Ks/sec you mentioned?
From the picture the mechanics looks very good to me.
I´d suggest using the Iop recommended by the manufacturer.

11. Feb 3, 2012

_Shellback

First, to give things their proper name: lasers have transverse (sometimes called spatial) modes as well as longitudinal ones. Single mode applied to a laser diode commonly refers to transverse modes; there may still be multiple longitudinal modes. When I said single mode I meant (longitudinal) mode stabilized lasers (used in spectroscopy and metrology).

“The "rock steady" output voltage is a bit surprising to me. Uf of a diode is notoriously temperature dependent; I´d expect at least some change in the first minutes of operation until reaching steady temperature.”

When I said “rock steady”, I was referring to monitoring voltage and current supplied to the laser overnight. The driver board does a fantastic job of maintaining the current I set when it’s on, but there is a very small amount of fluctuation during power up. The current tends to be low at first, but as the driver board and the LD warm up (<1min) things level off and become remarkably stable.

“You might want to do your own measurement with some reasonable DVM; maybe you can get at least some idea of temperature changes by monitoring forward voltage.”

I did this with the same DAQ card I used to monitor the output of the PMT and found no correlating changes in current or voltage that would indicate that something on those wires was causing the problem. I time stamped everything and overlaid the charts and found nothing.

”If you´ve really pinned it down to the LD:
My pet hypothesis still is output fluctuations (because of longitudinal mode competition?) caused by thermal expansion of the cavity. You can test this by (cautiously) heating (or cooling) your unit from the outside (hot air gun, peltier element) and checking wether there is any effect.“

This sounds like a very good direction. I will begin working this way. I would guess, by “cautiously” you mean don’t just hold a lighter under it?

12. Feb 3, 2012

maimonides

I had hoped you can use the diode as its own makeshift temperature sensor. Any diode Uf has -2mV/K temperature dependence according to the Shockley diode equation, but it might not apply to laser diodes. The effect is small in any case, You´d have to check the resolution af your ADC.
And no, no lighter.
Sorry, this is my last post. I´m going on holidays and I`ll be offline for a week.
All the best for your troubleshooting

13. Feb 3, 2012

_Shellback

Sounds good. I appreciate your help. I've been told to put this on hold for a little while and handle some other projects for the next few weeks. Enjoy your holidays.

14. Feb 3, 2012

Bobbywhy

One off-the-wall comment: Have you considered that the signal you are detecting is not coming from the laser diode? A remote possibility: the photomultiplier circuit may be oscillatiing due to some random noise. One more: PM tubes are known to output big signals when bomarded by cosmic rays!

15. Feb 3, 2012

_Shellback

We suspected the PMT for a long time. We even got some of the R&D engineers here from Japan to look at our application and the problem we're having. They did have a few suggestions, but in the end what we kept coming back to is that the problem stays the same when we change PMT's, but it will get better or worse when we change diodes.

I was not aware of the cosmic rays issue, but I figured it would be pretty well protected in an explosion proof enclosure with 3/4" thick aluminum. I realize it's not as dense as lead, but at that thickness, the PMT (with all of it's added internal shielding) should be pretty stable.

16. Feb 3, 2012