vcsel array

formatting link

This is a monolithic array of about 350 VCSEL lasers. We're playing with drivers for it.

The issue is how to measure the light pulse. We have a NuFocus 40 GHz o/e converter with multimode fiber input. We ordered an Ebay micromanipulator and maybe we can just hover a fiber over one of the VCSELs and get enough light for the detector. The NF detector is un-amplified and not very sensitive, 0.2 a/w or something like that.

We also ordered a fiber-coupled GRIN lens which we could hover over the laser and maybe collect light from a bunch of the emitters.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin
Loading thread data ...

How fast? I bought an ET-2030 severla years ago...

formatting link
mini circuit amp fit perfectly on top. (A few hundred MHz was fast enough for me.)
formatting link

George H.

Reply to
George Herold

What's the peak power and wavelength?

An SD-48 and SD-32 in one of your 11801Cs would get my vote, if the rep rate is reasonable (at least a few hundred hertz) and the wavelength is longer than 800 nm. I used that approach for my picosecond YAG/optical parametric generator setup a dozen or so years ago.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

The Tek samplers are great, but they are noisy. We want to measure jitter too, so we can't signal average for that. As George suggests, one of those boxed MiniCircuits type mmic amps would help. May as well go for a fast one, 20 GHz at least. I have an SD32, 50 GHz sampling head.

We're also playing with some gain-switched lasers, which can make light pulses in the 10s of ps, so bandwidth is good.

Amazon has nice optical benches and plates and stuff.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Check out Thor Labs. They're my main supplier of optomechanics these days. Their post bases with the claw-type clamp are the bomb, and all their stuff works together.

The 30-mm rod cage system is compatible with Microbench (I have a lot of MB in the drawer). Don't use their screw-together rods for anything precise--they're _way_ too floppy. Standard 6-mm centreless-ground stainless rods from McMaster-Carr are the ticket.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

It feels like a job for an integrating sphere otherwise you are going to get miss-match between the individual vcsel divergence and the acceptance angle of the collecting optics . Put your fibre at the exit port. Integrating spheres have low throughput of course and your detector is not very sensitive , but you'll be collecting the output of the whole array. You need to know the fibre acceptance angle and the fibre core diameter. The A-Omega product of a fibre is pretty small. A photo diode is better as it collects over pi sr.

It may be possible to do it by placing the fibre end close to the array, but you are not going to be able tell how many cells are contributing to the output

This sort of measurement isn't easy.

Brian

--
Brian
Reply to
Brian Howie

I suspect the the coupled power would be very low from an integrating sphere. We'll need at least a milliwatt into the detector. The entire array is good for a few watts, but maybe more if we pulse it hard.

Preaching the the choir!

The laser array has a cover window of some sort, so we can't get very close to the individual emitters. It will be interesting to crank the manipulator around and look at the power and see if we can resolve individual vcsels.

The detector is fiber coupled internally, so we can't blast its photodiode directly.

We also have a ThorLabs optical bench thing with a dual element lens and a laser mount and a fiber mount. We've used that to focus a single, fairly high power TO-can laser into a fiber. It might be able to focus one vcsel, or possibly a few, into a fiber.

Darn, a 60 u fiber isn't much area to hit.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

You can't measure short pulses with an integrating sphere, because the multiple scattering smears them out. Even a 1-inch sphere will give you about a 30-ns output pulse. I often recommend spheres to ultrafast laser folks for that reason--they prevent the peak power from driving the photodiode into ridiculous nonlinearity.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Good stuff.

Very roughly, the area of the vcsel array is 1e-6 sq meters. The area of a 60u fiber core is about 3e-9. The ratio is around 350:1, which is really not bad. The array might emit a few watts.

So just dangling the fiber end above the array might couple a few milliwatts.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

to

f

the

y,

to

I've always used these collimation lenses (on a tip tilt mirror mount) to couple into fiber. like these,

formatting link

-B

George H. (optics are spendy)

Reply to
George Herold

We just got one like that!

formatting link

It has an FC fiber connector on the back end.

I wanted to get a GRIN lens, but when we tried to order it they said that thay haven't made any since 2015. Must not be popular.

That thing above should be just as good.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

Which should be quite a lot for an SD-48. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Say, optimistically, we can couple 5 mW. The fast detector is about

0.2 A/W. So 1 mA into 50 ohms is 50 mV. Even with a noisy sampling head, we should see something.
--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

Well, you could use an SD-22 for the jitter measurement.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Light rattles around inside multimode fiber too. I've seen numbers like a few ns per kilometer time dispersion, which isn't bad for a meter of fiber, but it's one reason that telecom uses singlemode.

If we manage to make a, say, 20 ps optical pulse (probably not from a VCSEL) we'd have to be careful that the optical path doesn't smear it out.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

--
 Thanks, 
    - Win
Reply to
Winfield Hill

Run one at a time, no need for special optics.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

Just wondering if you have problems with back reflections in these lasers. I guess if it's pulsed fast enough.... I always get nasty reflections from fiber back into my diode laser. They are older, edge launched.

George H.

Reply to
George Herold

Run one vcsel? That are all in parallel.

We can't get very close to the array, so the light from one vcsel probably won't couple well into a fiber. But the light from a bunch of the dots might overlap at the fiber and give us some power.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

The application is free-space. If we pulse it fast enough, the drive is gone by the time any reflection arrives, so we shouldn't get mode jumps.

That's one cool thing about very fast pulses: fire and forget.

We have bought some sugar-cube type lasers that have integral isolators, which eat up back reflections. They're not expensive.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.