Fast (ns) rise/fall light source

Ah, I think I've found the solution .. inexpensive VCSEL laser diodes with 50ps rise/fall.. only a few mW output power, but with a laser it should be usable. Everybody and his dog seems to make reasonably fast laser driver chips.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Yeah, LEDs are slow.

I was just about to suggest that. A small VCSEL will make a few mw at

850 nm, and will have rise/fall times in the 200 ps range if you drive it with an EclipsLite, or even a fast CMOS gate.

This

uses an Appointech or LaserMate fiber-coupled VCSEL, depending on wavelength, with an EL89 driver in current-steering mode. Avoid the Optek stuff.

But just a fast pulse generator and a resistor can drive a VCSEL pretty fast; the optical edges are faster then the electrical input.

Mail me if you need some lasers, or want to borrow a J724.

(This is The Brat's new web site design. Better than the old one.)

John

I like it.

You should sell signed copies of this:

Worked for Steve Wozniak, right?

-- funny how there was never a "Jobs" edition... elthough I wouldn't be surprised if someday there's a Jobs' edition of the Jesus Ph-, I mean, iPhone.)

More seriously, from the looks of things, I suspect you're just about at the point where some bigger company like Berkeley Nucleonics is going to offer to buy you out...

---Joel

That's exactly the idea! Like those Eddy Bauer SUVs. Maybe we'll include a copy of his book with every purchase.

I know the BNC people. No way, to be polite.

John

Yet for the same wavelength and similar Vf I can find LEDs with ~1us/100ns/10ns/1ns specs. Three orders of magnitude. Very strange.

Nice product. Thanks for the offer, I may take you up on it.

Clean, no spinning logos or other junk.. very good.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

"The Journey is the reward"

eff.com

RF modulate the current of a laser diode?

There was a company that bought a sonoluminesence from us as a fast light source. Lot's more than $100 though

George H.

How does $5 strike you? You would have to go dig the laser out of it, but that's about what I've paid for used SFPs on eb*y, which have a diode and a laser in them....you can get new ones for $50 or less (or you can pay several hundred for certain brands, but you don't need to.)

The really high power ones will be longer wavelengths than you want, but the wavelength you want is also available. Get gigabit SFPs, they all have diode lasers or VCSELs.

--
Cats, coffee, chocolate...vices to live by

For LED drive, testing is commonly done with current-on/current-off biasing. If you instead reverse the drive polarity and pull current from the LED, it turns off faster. Probably not fast enough, though.

A Q-switched gas laser can do it under 1 ns, of course.

Curiosity got the better of me. At the moment, $3.99 will get you an

850nm 2GB SFP at the large watery "auction" site.

To get 2GB data rates, it _has_ to be turning on/off faster than 1 nS

If you're willing to rip it open, I doubt you can get cheaper (depending what you are doing, you may not even need to rip - just sort through the pinout to make it behave as an assembled unit.)

--
Cats, coffee, chocolate...vices to live by

Impressive! But telecom units are probably AC-coupled, and may not make light pulses. But you could hack them to get at the laser.

Maybe I'll get a few of these and play with them.

John