proper response?

Not with polarization and phase noise in VCSELs :-(

Some VCSELS are horrors. The Optek parts tend to be especially bizarre. Whose are you using?

Lately, we've been buying an ST connectorized LaserMate 850 VCSEL, which makes really beautiful, flat optical 1 mW fiber-coupled pulses with (with my gear) unmeasurable risetimes.

After working with electrical signals, fiber optics seems absurdly crude. There are, as Phil Hobbs says, "etalons everywhere". If you unmate and remate a fiber cable, or maybe wiggle the connector a bit, you're lucky to get back to 3 dB from where you were. 25% pulse-top wigwags start to look "flat."

John

Raycan. They are top notch WRT line width. Well, for a VCSEL, that is. If you come across better ones (IR range) let me know. DFBs are a lot better but too slow for us in terms of tuning speed.

For this project we have to stay in the IR range and that's where it's slim pickens.

Tell me about it. You move a fiber or sometimes just sneeze within 3ft of it and the noise changes. I wonder how sensitive VCSELs are when it comes to back reflection.

I've not seen it that gross. We always get to within 0.5dB. But we are very diligent WRT surfaces. The instant a fiber is unplugged the little protector goes on. There is always a box of them, for every kind of connector we are using, on every bench. After all, we are getting all the smog from you guys blown up to this area. Then before re-connecting there will be two wipes on the cleaning cloth box.

It was. The dumb donkey was the one who impaled the sheep (and likes it).

We're using some Perot-Fabry (facet cavity) lasers at 1310 and 1550, and they're pretty good in the time domain. VCSELS tend to be twichier and fairly sensitive to back reflecions, although you can get vcsels with attenuating coatings that reduce reflection sensitivity but throw away power. DFBs are good but expensive.

One of my customers Who I May Not Name bought a lot of fiberoptic stuff from us. They happen to be spending billions of bucks, and we did the timing system to fire it all. They also happen to own the world's largest clean room. They started sending modules back to us, working badly or not working at all. After some investigation, we found all sorts of hairs and dust bunnies inside our laser and photodiode connectors. Then they asked "Should we keep those caps on? We threw them all away. Where can we get some more?" followed by "How do you clean these things?"

John

We need fast tuning (kHz range) over a couple of nm so anything fixed-cavity is out. Except for the esoteric stuff with tuned cavities but that becomes prohibitively expensive.

Ouch! I really would have thought those guys would know. That world's largest clean room kind of gave it away ... and hey, they've got a web site that actually works.

You're not really getting the sense of the sentence here. "I'll check it _as if it were_ designed by an idiot."[emphasis mine]. He's not calling anyone an idiot, just experssing that that's the kind of care he takes when checking something.

Hope This Helps! Rich

Hear, Hear!

The lazy man's way is always better, because it accomplishes the goal with the least possible expenditure of resources. ;-)

Cheers! Rich

I realize that. Still don't like that style. If my wife asked me to check her car because of a weird noise and I'd say "Ok, I'll check it bumper to bumper as if a moron had taken it over a curb" there would be no margaritas for the next several weeks.

You wife only withholds Margaritas ?:-)

...Jim Thompson

But not the ice :-)

That's different - she didn't _design_ the car. In a case like that, you'd go over it with a fine-toothed comb. ;-)

Cheers! Rich

Nope. After mounting a new muffler plus a new oil pan, and adjusting the suspension you wash your hands, bandage the bleeding knuckles and tell her "Oh, it was nothing, honey. Noise is gone."

(BTW, my wife has never done that, but somebody else's did)

Late at night, by candle light, John Larkin penned this immortal opus:

Most solutions are "obvious" once you've stumbled upon them.

- YD.

Which is why some amount of stumbling is advisable. Too many engineers are uncomfortable or impatient with the first, fuzzy part of a project when everything is unresolved. So some engineers glom onto the first architecture that seems feasible, and then settle into the implementation phase where they're more comfortable. So they may wind up implementing a monstrosity.

John

Late at night, by candle light, John Larkin penned this immortal opus:

What works for me is not thinking too hard about it at this stage. Just sort of mentally set up the various bits and pieces that'll go into it in a blocky kind of way. Make doodles and sketches, look around for similar circuitry to steal ideas. Then go for a long walk, look at the sky and sniff the flowers. When I get back look over what I have and start thinking "WTH am I doing it like this?" If there's a good answer let it stay, otherwise take another walk. Then go into the detailed design and calculations. Stuff just sort of falls into place at this stage. If it's not an urgent jobby the "setting up" stage may account for a large part of the total time.

- YD.

Then there is the more adventurous bunch of engineers. They love that first phase but once it goes on to implementation they begin to feel bored, wanting the next challenge instead.

Isn't that all of us though? I do the layout, assembly drawing, manuals thing--you have to--but that's the discipline. Making impossible stuff that works is the most fun, and that's usually in the design. (Could be in electronic, thermal, and/or mechanical design though.)

Cheers, James Arthur

Yep. Just completed a prototype debug, six boards. All 0603 and smaller, such as SC-75. #3 glasses plus magnifier. Oh man. Now off to Fedex.