Pangolin ? laser not perform at maximum

[norty303]

@ Norty

I get an air of "conspiracy theory" from you. I even got this in our private email discussions.

Absolutely Bill I'll only press so far, and as I said, either you didn't pick up the specific point I was referring to, or you were avoiding it (either way might lead someone to think that way... ) - we both agree we don't quite 'get' each other's perspective - and to be clear, no bad feelings at all from me.

But as it was raised again here, I thought I'd add it to the 'information pot'. I'm glad that the original point about scanned output also being affected (maybe?) is also on the table. I felt the whole focus to date has been about targetted beams, which as far as I'd read it wasn't the original point raised way back when...

I think the confusion may have come about as you responded that you'd investigated the issue and fixed the beam system but didn't necessarily say you'd found nothing with scanned output - which lead me to the previous conclusions... as you say, non-face to face communication will always lead to misunderstandings, because I can't butt in and just say 'sorry Bill, to clarify, are you saying you have/have not looked at the scanned image issue raised?'

Wish I could afford to get across the pond for SELEM...

[Pangolin]

@ rfourt

This is pretty easy to believe that if you project a single circle it looks like one thing, but if you project four circles it has higher power. Steve already gave the answer to this and so did I, but, unless you really understand lasers and what goes on inside them, you would not catch, or understand, the answer.

The reason is "thermal modulation" which results from "light modulation".

And it should be understood by everyone in the world (or that is my wish) that this has nothing to do with the FB3 or Pangolin software.


@ Norty

The single most frustrating point for me is that almost nobody in PL can talk on my level. Nobody really understands these things the way I do. I try to point out the obvious -- I have numerous patents that cover this industry and nobody else here does, so it tends to indicate that I understand things that nobody else does. Steve already answered your question. So have I. But I think you are missing the answer, or thinking there must be more to it.

There is nothing going on here guys. People are basically acting on emotion and misunderstandings rather than cold hard facts and observations based on real testing with real test equipment.

But we'll prove that at SELEM. Power meters, scopes, and a zillion people standing around a projector won't lie.

Finally, I will make one point. Back in the days of the development of the DPSS lasers, people at CREOL (the Center for Research on Electro Optics and Lasers here at the University of Central Florida) said that it would be impossible to ever directly modulate the power of a DPSS laser, because the temperatures inside the laser are so critical. But here we are doing something that the scientists who invented these lasers believed could never be done.

Yes, it is being done. BUT THERE ARE SACRIFICES AND TRADEOFFS!!!! When you directly modulate the power of a DPSS laser, the power stability goes to hell.

Some DPSS are better than others. At the PALM show in Beijing, I saw an RGB laser module made by LaserWave that appeared to be VERY stable. I have also heard that UltraLasers are also stable (but I will find out for myself pretty soon). By the way, RGB Laser in Hungary recently figured out a way to completely solve this problem, using a different-style resonator design.

Anyway, it's like I wrote -- stop complaining about Pangolin software and start complaining about your lasers!! In the last, we discussed with LaserWave that they needed to improve their lasers -- AND THEY HAVE!!! The more people who speak up and complain to laser companies, the better the lasers will become. And I am serious -- when you complain to your laser companies, tell them to contact Pangolin and we'll discuss ways that they can improve their lasers.

Buffo, remind me to bring my real-time photodiode so we can sample the output of DPSS (and even direct-diode) lasers and compare it with the drive voltage. It will give another data point about how the behavior (or misbehavior) is similar no matter what software you use. Now THAT will be informative.

Bill

[DMills]

VI, The laser tuning is often set up to favor long trains of constant illumination, being off for long periods of time results in a dim laser for a few seconds until it reaches steady state again.

That can only be poor thermal management, several SECONDS is longer then anything except a thermal time constant.
I would suspect that a single TEC is cooling both the laser crystal and the diode (or may in fact be trying to tune the entire cavity to some kind of compromise).

Even with type II phase patching that is going to be a major performance compromise where eve a simple resistive heater on the doubler would largely fix it.
On fact come to think of it, one could pretty easily build a diode driver that ran at constant power irrespective of modulation level and dumped the excess into a power device mounted on the same tec as the diode.... Wouldn't help with the KTP/LBO but would remove most of the hunting from the diode tec control loop.

If you built a shunt modulation driver and put the modulation pass transistor on the diode mount and the current sense resistor on the YAG/Vanidate mount then by carefully selecting the sense resistor value such that at full blanking it was dissipating heat equal to that which the pump beam deposited into the laser crystal and the pass transistor was dissipating the rest then both the diode and the laser crystal cooling would see a heat load independent of the modulation level.

Yet I have also seen lasers that are nearly perfect. The 808 nm diodes can easily follow to 150-250 Khz or more. The doubling is instant. The laser material has a long upper state storage time, and this can be a problem.

You sure about that, the upper state lifetime for YAG is ~330us IIRC, but it is being actively depopulated by the laser action so the effective half life is going to be much shorter then that. If it is 180us you are only getting half the possible power as the other half will be non coherent decay, so I would expect the effective half life while the system is above threshold to be more like a few tens of microseconds.

I think the issue is far more about thermal effects then it is upper state lifetime, consider that going from steady state on to blank is quick, it is starting up again that is painful, and startup cannot be due to upper state decay time constants.

All of this needs hashed out to be a standard. It would somewhat help if control software/hardware had independent blanking delays for each color.

Agreed, threshold and gamma would be good numbers to specify, and the impedance thing is just someone cheaping out, there is no excuse for needing more then a mA or so into a control input (I normally design for 10K differential, 100K+ common mode).

Well the discussion has given me an interesting idea to play with (The constant thermal load driver), need to do some design now.

Regards, Dan.

[norty303]

15 chars...nothing to see here, move along....

[norty303]

Actually, after reading this, could the reason my latest Laserwave 532 exhibit strange fluctuations on CW be because they are now tuning them to be optimized when modulated?

[Hayden]

OK, what I have seen:

I. Bad thermal management in the laser. KTP and vandate/yag are temperature sensitive, LBO even more so. All are insulators with poor thermal conductivity, so cooling them is difficult. Non linear crystals change their phase matching angle with temperature. Really good lasers heat the KTP and cool the vandate/yag.
Laser output tends to depend on recent history in the poorer lasers.

II. Poor choice of the logic input structure on the TTL diode driver. Many of them, as so called "TTL" need enough current sunk that only 100 ohms or less to ground will start them to blank. Some have a TTL spec, some are CMOS with a pull up, some have a comparator, some wink if you couple 60 hertz in with your finger. Some have horrible hysteresis. I'm not sure about the current ILDA standard, but the input impedance in the old standard was 600 ohms. I've seen some at infinity, and I've been told about some at 50-100 ohms. I have one laser where the input device is a LED in a very slow optocoupler with 200 ohms in series. If there is one laser out there that is on at 2.4 volts or higher and off at .6 volts or lower per the TTL standard, I have yet to find it.

III Analog inputs, ditto on the impedance issues. Some lasers are biased internally so that any voltage above 0 volts adds to diode current set at just below lasing current. Some have the diode at zero current and threshold is a few volts up, cramming all the output into the last volt. Some are very good, fast, and linearized. Again, impedance is all over the place. I've seen lasers that need a analog opamp follower driving the input.

IV, There is no standard for linearity. There is no standard for threshold. There is no standard for rise and fall time. Last but NOT least there is no standard for how a laser is supposed to behave for a single blanked point in a train of illuminated points, nor a standard for how to respond to a single illuminated point in a train of blanked points.

VI, The laser tuning is often set up to favor long trains of constant illumination, being off for long periods of time results in a dim laser for a few seconds until it reaches steady state again.

Yet I have also seen lasers that are nearly perfect. The 808 nm diodes can easily follow to 150-250 Khz or more. The doubling is instant. The laser material has a long upper state storage time, and this can be a problem.

DPSS Blue is even worse then green, for a variety of other reasons. LBO in a 473 nm blue needs held to about +/- .1' C. 473 blue often has 5% low frequency noise riding on the cw beam

All of this needs hashed out to be a standard. It would somewhat help if control software/hardware had independent blanking delays for each color.

This is not new, there were complaints on early PCAOMs that they would not saturate at just 5 volts, and that they would not 100% extinguish, a leakage signal when off 30 dB down still being visible in a very dark room.

There is a dormant portion of the ILDA test pattern dealing with blanking. It is there for tuning galvo blanking, but how it was chosen is a mystery to me. There is someone I can ask, if he knows. I'll call him tomorrow.

Steve

DING DING DING!!!

We have a winner folks.

[mixedgas]

Actually, after reading this, could the reason my latest Laserwave 532 exhibit strange fluctuations on CW be because they are now tuning them to be optimized when modulated?

Ding Ding Ding! Another winner!

Yeah, I'm hearing they tune some now for 50% duty cycle, so when you run it cw...
In fact, one should ask/specify when one orders the laser.

Or just order a polarized laser, 25% more powerful then needed, and just get a AOM and get it over with.



Steve

[Doc]

I've had some good improvement using the latest build of LP and re-tuning the crystal and diode temps whilst modulating with the FB3; I lost some top end CW, but that is irrelevant for our purposes anyway.

Two concentric full power circles yeilded circa 200mW from my 1W CNI, now it is circa 700mW. Full power beams are now a around a Watt (down about 12-15% from optimal CW setting).

[mixedgas]

Um Dan, you should know, the pump diode can shift as much as .2 nm per % C or more and emits a wavelength profile that when plotted looks like a train of irregular spikes. The ND adsorption profile is also a train of spikes. The spikes can line up and overlap in variable amounts while shifting

The Nd: emission profile is 1.1 nm wide in most lasers, with a secondary offset "hump" driving the ktp batty with phase matching effects and the "green noise problem" Wide gain profiles in many lasers reduce power from competition of many different modes within the gain profile, that shows up as noise, as well.

So you have KTP temp, diode wavelength shift due to temp, and localized optical heating effects in the lasing crystal, plus adsorption shifting in the KTP. its a mess.

Nobody at PL has really tried light feedback except me, DZ and Heroic, and none of us have published anything.. Hint! I'd go after light but I don't have a dpss I can spare for hacking.

Steve

[mixedgas]

I've had some good improvement using the latest build of LP and re-tuning the crystal and diode temps whilst modulating with the FB3; I lost some top end CW, but that is irrelevant for our purposes anyway.

Two concentric full power circles yeilded circa 200mW from my 1W CNI, now it is circa 700mW. Full power beams are now a around a Watt (down about 12-15% from optimal CW setting).

This gives me some hope for a tuning standard for the two mirror cavities. How bad is the decay or gain if blanked for a few seconds and then returned to the pattern?

Steve