| |
 |
|
| Science Forum Index » Optics Forum » VCSEL vs DFB lasers frequency stability... |
|
Page 2 of 2 Goto page Previous 1, 2 |
|
| Author |
Message |
| Hazem... |
 Posted: Thu Oct 08, 2009 3:59 am |
|
|
|
Guest
|
On Oct 7, 7:11 pm, Jürgen Appel <jap... at (no spam) linux01.gwdg.de> wrote:
[quote:e4747a6196]Phil Hobbs wrote:
First the OP has to make the laser single mode, because the modes are
much more than 0.001 nm apart. That's what makes the problem as stated
poorly defined. 1 part in 10**6 feedback is enough to make a FP laser
go bananas, and most VCSELs jump between several spatial modes during
turn on, and often afterwards as well. Some have horrible polarization
instability due to being very nearly rotationally symmetric.
A fancy temperature controller isn't a complete solution to the problem
by a long way--not that the OP has actually told us the problem.
100% agreed. And if his budget really is not limited, in my opinion looking
for a simple diode laser is looking the wrong way anyway.
For someone who has no experience at all in laser stabilization I would
recommend a 633nm HeNe-Laser to start with. They have a very narrow gain
bandwidth and so even in the presence of mode jumps the frequency will only
change at most by a few GHz.
For applications where frequency stability really matters, VCSEL lasers are
almost never used. This is mostly because their inherent line width
(typically >20 MHz) is much worse than what you can stabilize other diode
lasers to easily. For only a little tighter requirements this rules out DFB
lasers (mostly >1 MHz) for the same reason.
For the highest stability requirements (what time span are we talking about
actually?) irrespective of costs high-finesse sapphire resonators are used,
which are locked to atomic time standards via a frequency comb...
Cheers,
Jürgen
[/quote:e4747a6196]
OK, many thanks to all if you. I should qualify my requirements
further...
The system I am going to use should be either an HeNe laser, or some
kind of diode laser. I don't think any expensive exotic lasers will be
approved. I just want the best wavelength and spectral stability
available from those choices, with integrated or added on Bragg
stabilizers, TECs, optics, etc. The reason I said my budget is
unlimited was so as to not put an artificial ceiling on discussion. I
wanted to learn something about leaser stability from you. I just
didn't think we'd be talking about anything but single-mode fiber-
coupled VCSEL or DFBs.
Jurgen mentioned HeNe lasers and I have tested a number of them that
were available to me, based on the same reasoning. I observed mode
hops which made them unsuitable, however, the laser was not fiber-
coupled. I had much better performance from a fiber-coupled diode
laser, specifically the Lasiris PTL (red). Do you think that coupling
a single-mode fiber to an HeNe laser would eliminate the mode hops at
the output? Are these spacial or frequency hops or both? my testing
equipment cannot distinguish the two.
Thanks,
Hazem |
|
|
| Back to top |
|
|
|
| Hazem... |
| Posted: Thu Oct 08, 2009 7:16 am |
|
|
|
Guest
|
On Oct 8, 12:05 pm, AES <sieg... at (no spam) stanford.edu> wrote:
[quote:bc904cad25]In article
0618a0cc-7108-4a88-9e0c-0b0c9f2cc... at (no spam) e34g2000vbm.googlegroups.com>,
Hazem <hazem.biqa... at (no spam) gmail.com> wrote:
Jurgen mentioned HeNe lasers and I have tested a number of them that
were available to me, based on the same reasoning. I observed mode
hops which made them unsuitable, however, the laser was not fiber-
coupled. I had much better performance from a fiber-coupled diode
laser, specifically the Lasiris PTL (red). Do you think that coupling
a single-mode fiber to an HeNe laser would eliminate the mode hops at
the output? Are these spacial or frequency hops or both? my testing
equipment cannot distinguish the two.
Thanks,
Hazem
Hazem, I hope this won't come across as a nasty response, because it's
not at all intended that way.
Lasers are in one sense individually simple gadgets, but in another
sense complicated gadgets with a complex set of multiple properties --
and then, there are a large variety of different types of laser which
differ greatly from each other in their individual values of those
multiple properties.
My impression is, you're being tasked well beyond your experience in all
these complexities; and trying to get educated about a topic this
complex on a newsgroup may or may not be successful. I'd suggest
spending some of those funds on a competent consultant right from the
start could be a wise choice and a much more efficient approach. (It
might not take a lot of funds or time to do that.)
[/quote:bc904cad25]
No offence taken, and you're absolutely right. Laser physics is not my
area of expertise (although it is becoming one!) and my background in
lasers is limited. However, I find myself in a situation where I am
expected to show results in a project where much has been invested
already, and although we did hire a consultant who was very helpful,
the company I work for let him go as part of measures taken to reduce
overall expenditures in response to the economic crisis. To some
extent, it is my fault for not anticipating the extent of the
complexities I ran into, but on the other hand, that is hard to do,
and we did end up hiring a consultant for previous iterations, and I
could not have anticipated that he would be let go.
Knowing how to ask the right questions means that I have to educate
myself first, and I think I've come a long way. I believe I am close
to building a system that would satisfy our requirements. The company
has made an expert available to me recently, but the problem is that
his time is taken up by other projects so I still have to do alot of
the work on my own. On the specific subject of DFB vs VCSEL achievable
stability it seems there are varying opinions among the experts.
One way or the other i am going to find a suitable laser. Your
assistance will help shorten my search, and will be duely
acknowledged.
Hazem |
|
|
| Back to top |
|
|
|
| Phil Hobbs... |
| Posted: Thu Oct 08, 2009 8:54 am |
|
|
|
Guest
|
Hazem wrote:
[quote:7ea92626b7]On Oct 7, 7:11 pm, Jürgen Appel <jap... at (no spam) linux01.gwdg.de> wrote:
Phil Hobbs wrote:
First the OP has to make the laser single mode, because the modes are
much more than 0.001 nm apart. That's what makes the problem as stated
poorly defined. 1 part in 10**6 feedback is enough to make a FP laser
go bananas, and most VCSELs jump between several spatial modes during
turn on, and often afterwards as well. Some have horrible polarization
instability due to being very nearly rotationally symmetric.
A fancy temperature controller isn't a complete solution to the problem
by a long way--not that the OP has actually told us the problem.
100% agreed. And if his budget really is not limited, in my opinion looking
for a simple diode laser is looking the wrong way anyway.
For someone who has no experience at all in laser stabilization I would
recommend a 633nm HeNe-Laser to start with. They have a very narrow gain
bandwidth and so even in the presence of mode jumps the frequency will only
change at most by a few GHz.
For applications where frequency stability really matters, VCSEL lasers are
almost never used. This is mostly because their inherent line width
(typically >20 MHz) is much worse than what you can stabilize other diode
lasers to easily. For only a little tighter requirements this rules out DFB
lasers (mostly >1 MHz) for the same reason.
For the highest stability requirements (what time span are we talking about
actually?) irrespective of costs high-finesse sapphire resonators are used,
which are locked to atomic time standards via a frequency comb...
Cheers,
Jürgen
OK, many thanks to all if you. I should qualify my requirements
further...
The system I am going to use should be either an HeNe laser, or some
kind of diode laser. I don't think any expensive exotic lasers will be
approved. I just want the best wavelength and spectral stability
available from those choices, with integrated or added on Bragg
stabilizers, TECs, optics, etc. The reason I said my budget is
unlimited was so as to not put an artificial ceiling on discussion. I
wanted to learn something about leaser stability from you. I just
didn't think we'd be talking about anything but single-mode fiber-
coupled VCSEL or DFBs.
Jurgen mentioned HeNe lasers and I have tested a number of them that
were available to me, based on the same reasoning. I observed mode
hops which made them unsuitable, however, the laser was not fiber-
coupled. I had much better performance from a fiber-coupled diode
laser, specifically the Lasiris PTL (red). Do you think that coupling
a single-mode fiber to an HeNe laser would eliminate the mode hops at
the output? Are these spacial or frequency hops or both? my testing
equipment cannot distinguish the two.
Thanks,
Hazem
[/quote:7ea92626b7]
Just about all He-Nes are single transverse mode. You can get
single-frequency (i.e. single longitudinal mode) units, which generally
keep the gain very low and adjust the cavity length with a heater or
piezo so that one cavity mode sits at the gain peak. They're discussed
in gory detail on Sam's Laser FAQ. Because the gain has to be low,
they're not very powerful--usually below 1 mW, iirc. I suppose you
could use an intracavity etalon, as with argons, but I've never seen one
in a He-Ne.
The mode spacing on a 20 cm long gas laser is 750 MHz, which
coincidentally is just about exactly 0.001 nm. The longitudinal modes
tend to jump in and out of mode-locking, which leads to fairly strong
low frequency spurs (100 kHz-1 MHz or thereabouts) that wander around.
These are fourth-order mixing products, caused by the beating of one
mode with a third-order intermodulation product of two others. (If
there are more than three modes, they can all contribute.) Once the
beat frequency drops below 100 kHz or so, they seem to jump back into
lock, and the low frequency beats go away. It would probably be
possible to sense this and control the cavity length as in a
single-frequency laser.
Fibre coupling is a famous source of diode laser misbehaviour, due
largely to the huge facet reflection.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net |
|
|
| Back to top |
|
|
|
| AES... |
| Posted: Thu Oct 08, 2009 10:05 am |
|
|
|
Guest
|
In article
<0618a0cc-7108-4a88-9e0c-0b0c9f2cc0b6 at (no spam) e34g2000vbm.googlegroups.com>,
Hazem <hazem.biqaeen at (no spam) gmail.com> wrote:
[quote:3f53ddf540]Jurgen mentioned HeNe lasers and I have tested a number of them that
were available to me, based on the same reasoning. I observed mode
hops which made them unsuitable, however, the laser was not fiber-
coupled. I had much better performance from a fiber-coupled diode
laser, specifically the Lasiris PTL (red). Do you think that coupling
a single-mode fiber to an HeNe laser would eliminate the mode hops at
the output? Are these spacial or frequency hops or both? my testing
equipment cannot distinguish the two.
Thanks,
Hazem
[/quote:3f53ddf540]
Hazem, I hope this won't come across as a nasty response, because it's
not at all intended that way.
Lasers are in one sense individually simple gadgets, but in another
sense complicated gadgets with a complex set of multiple properties --
and then, there are a large variety of different types of laser which
differ greatly from each other in their individual values of those
multiple properties.
My impression is, you're being tasked well beyond your experience in all
these complexities; and trying to get educated about a topic this
complex on a newsgroup may or may not be successful. I'd suggest
spending some of those funds on a competent consultant right from the
start could be a wise choice and a much more efficient approach. (It
might not take a lot of funds or time to do that.) |
|
|
| Back to top |
|
|
|
| DavidM... |
| Posted: Fri Oct 09, 2009 9:18 am |
|
|
|
Guest
|
Phil Hobbs wrote:
[quote:b4b1e4caee]DavidM wrote:
Hazem wrote:
On Oct 6, 11:52 pm, "Bret Cannon" <nore... at (no spam) invalid.invalid> wrote:
"Hazem" <hazem.biqa... at (no spam) gmail.com> wrote in message
news:decc51bf-c0d1-47af-9e03-a4279893262b at (no spam) o21g2000vbl.googlegroups.com...
Please excuse me if this question has been asnwered before. The search
function does not work for older posts on Google Groups anymore.
I am trying to decide whether to acquire a VCSEL or a DFB laser for a
measurement application. The laser will be coupled to a single-mode
fiber, so many of the advantages of VCSEL over DFB diode lasers become
irrelevant. The one criterion I care most about is frequency/
wavelength stability. My understanding is that modern DFB or VCSELs
both have very good stability, but I'd like to know which one is the
best. Wavelength will be in the red/ near IR regions, and I have not
chosen it yet. Again, stability is more important and the choice of
wavelength will be secondary, as long as it is in the long visible or
short IR regions.
Your help is much appreciated.
Hazem
If you want good stability, there are diode lasers mounted on Peltier
coolers with thermistors inside the package. With such a diode
laser it is
straight forward to stabilize the diode laser temperature to 0.001°C
over
periods of hours. Since you don't say anything about your budget,
this may
be too expensive.
Bret Cannon- Hide quoted text -
- Show quoted text -
Yes, I am aware that thermal stabilization is critical, more so for
DFB than for VCSEL, although I doubt you can have 0.001C resolution!
what reason do you have to doubt it.
resolution is easy, stability a little harder but it can be realised.
Commercial drivers from Thorlabs or ILX can come close with careful
PID set up and taking time to design your own can hit 0.001C
stability. For greater stability you can put a box withiin a box and
control both of them.
Note a stability of only 0.01C can realise 0.001nm/C, given a typical
temperture tuning coefficient of 0.01nm/C
A single good driver can sit with 0.001nm stability with a 1550nm DFB.
If you want higher stability then you can always lock to an absorbtion
line.
Many DFBs ship with a packaged thermoelectric cooler and thermistor.
My question is: Out of the commercially available DFB or VCSEL
__SYSTEMS__, which have the higher wavelength stability for
wavelengths in the red/near IR region? My budget can be considered
unlimited.
Is linewidth important
can the laser be modulated or does it have to run constant power.
In fact if you have an unlimited budget then I'm pretty sure I can
certainly supply you with a system with that stability just let me
know what you want.
cheers
First the OP has to make the laser single mode, because the modes are
much more than 0.001 nm apart. That's what makes the problem as stated
poorly defined. 1 part in 10**6 feedback is enough to make a FP laser
go bananas, and most VCSELs jump between several spatial modes during
turn on, and often afterwards as well. Some have horrible polarization
instability due to being very nearly rotationally symmetric.
A fancy temperature controller isn't a complete solution to the problem
by a long way--not that the OP has actually told us the problem.
Cheers
Phil Hobbs
[/quote:b4b1e4caee]
Sure enough
I don't think a temperature controller on it's own is the solution but a
DFB with decent temp stabilisation can get within an order of magnitude
of what is required.
Getting a good DFB at the OP's wavelength of interest might be an issue.
However given the OPs reluctance to give a decent spec it is difficult
to nail down a solution.
I've only ever stabilised 1550nm DFBs or fibre lasers down at this level
and have never really looked at the near IR region the OP appears
interested in.
cheers
David |
|
|
| Back to top |
|
|
|
| Hazem... |
| Posted: Sat Oct 10, 2009 6:36 pm |
|
|
|
Guest
|
On Oct 9, 11:18 am, DavidM <no... at (no spam) nospam.com> wrote:
[quote:c1608ea331]Phil Hobbs wrote:
DavidM wrote:
Hazem wrote:
On Oct 6, 11:52 pm, "Bret Cannon" <nore... at (no spam) invalid.invalid> wrote:
"Hazem" <hazem.biqa... at (no spam) gmail.com> wrote in message
news:decc51bf-c0d1-47af-9e03-a4279893262b at (no spam) o21g2000vbl.googlegroups.com...
Please excuse me if this question has been asnwered before. The search
function does not work for older posts on Google Groups anymore.
I am trying to decide whether to acquire a VCSEL or a DFB laser for a
measurement application. The laser will be coupled to a single-mode
fiber, so many of the advantages of VCSEL over DFB diode lasers become
irrelevant. The one criterion I care most about is frequency/
wavelength stability. My understanding is that modern DFB or VCSELs
both have very good stability, but I'd like to know which one is the
best. Wavelength will be in the red/ near IR regions, and I have not
chosen it yet. Again, stability is more important and the choice of
wavelength will be secondary, as long as it is in the long visible or
short IR regions.
Your help is much appreciated.
Hazem
If you want good stability, there are diode lasers mounted on Peltier
coolers with thermistors inside the package. With such a diode
laser it is
straight forward to stabilize the diode laser temperature to 0.001°C
over
periods of hours. Since you don't say anything about your budget,
this may
be too expensive.
Bret Cannon- Hide quoted text -
- Show quoted text -
Yes, I am aware that thermal stabilization is critical, more so for
DFB than for VCSEL, although I doubt you can have 0.001C resolution!
what reason do you have to doubt it.
resolution is easy, stability a little harder but it can be realised.
Commercial drivers from Thorlabs or ILX can come close with careful
PID set up and taking time to design your own can hit 0.001C
stability. For greater stability you can put a box withiin a box and
control both of them.
Note a stability of only 0.01C can realise 0.001nm/C, given a typical
temperture tuning coefficient of 0.01nm/C
A single good driver can sit with 0.001nm stability with a 1550nm DFB.
If you want higher stability then you can always lock to an absorbtion
line.
Many DFBs ship with a packaged thermoelectric cooler and thermistor.
My question is: Out of the commercially available DFB or VCSEL
__SYSTEMS__, which have the higher wavelength stability for
wavelengths in the red/near IR region? My budget can be considered
unlimited.
Is linewidth important
can the laser be modulated or does it have to run constant power.
In fact if you have an unlimited budget then I'm pretty sure I can
certainly supply you with a system with that stability just let me
know what you want.
cheers
First the OP has to make the laser single mode, because the modes are
much more than 0.001 nm apart. That's what makes the problem as stated
poorly defined. 1 part in 10**6 feedback is enough to make a FP laser
go bananas, and most VCSELs jump between several spatial modes during
turn on, and often afterwards as well. Some have horrible polarization
instability due to being very nearly rotationally symmetric.
A fancy temperature controller isn't a complete solution to the problem
by a long way--not that the OP has actually told us the problem.
Cheers
Phil Hobbs
Sure enough
I don't think a temperature controller on it's own is the solution but a
DFB with decent temp stabilisation can get within an order of magnitude
of what is required.
Getting a good DFB at the OP's wavelength of interest might be an issue.
However given the OPs reluctance to give a decent spec it is difficult
to nail down a solution.
I've only ever stabilised 1550nm DFBs or fibre lasers down at this level
and have never really looked at the near IR region the OP appears
interested in.
cheers
David
[/quote:c1608ea331]
I can't give you a wavelength spec because I don't have one yet.
Wavelength stability is more important than wavelength, and the choice
of wavelength will depend in part on type of laser, based on
achievable stability performance.
Which brings me to a related question: Can we make broadly valid
comparisons between HeNe, DFB, and VCSEL lasers in terms of linewidth
and mode spacing?
thanks
Hazem |
|
|
| Back to top |
|
|
|
| Jürgen Appel... |
| Posted: Sun Oct 11, 2009 7:18 am |
|
|
|
Guest
|
Hazem wrote:
[quote:6a3e69e4e9]I can't give you a wavelength spec because I don't have one yet.
Wavelength stability is more important than wavelength, and the choice
of wavelength will depend in part on type of laser, based on
achievable stability performance.
[/quote:6a3e69e4e9]
Stability over what time scale, what power, and _again_ for what kind of
measurement? If you want this group's help, you will have to describe your
application better. The longer I read your postings the more I get the
feeling that you are heading the wrong way...
[quote:6a3e69e4e9]Which brings me to a related question: Can we make broadly valid
comparisons between HeNe, DFB, and VCSEL lasers in terms of linewidth
and mode spacing?
[/quote:6a3e69e4e9]
Any _elementary_ textbook of coherent optics should cover the relation of
these parameters to the laser resonator length. Also you can have a look at
the spec sheets of the suppliers. |
|
|
| Back to top |
|
|
|
| Hazem... |
| Posted: Sun Oct 11, 2009 2:15 pm |
|
|
|
Guest
|
On Oct 11, 9:18 am, Jürgen Appel <jap... at (no spam) linux01.gwdg.de> wrote:
[quote:fadfb4fbf3]Hazem wrote:
I can't give you a wavelength spec because I don't have one yet.
Wavelength stability is more important than wavelength, and the choice
of wavelength will depend in part on type of laser, based on
achievable stability performance.
Stability over what time scale, what power, and _again_ for what kind of
measurement? If you want this group's help, you will have to describe your
application better. The longer I read your postings the more I get the
feeling that you are heading the wrong way...
Which brings me to a related question: Can we make broadly valid
comparisons between HeNe, DFB, and VCSEL lasers in terms of linewidth
and mode spacing?
Any _elementary_ textbook of coherent optics should cover the relation of
these parameters to the laser resonator length. Also you can have a look at
the spec sheets of the suppliers.
[/quote:fadfb4fbf3]
Stability time scale: 30 min
Optical power ~1mW
Measurement: Cannot say due to a non-disclosure agreement, but you'd
be interested to know that it is related to minimizing dispersion
variation. Better wavelength stability ==> more predictable
dispersion.
I'm new to the field so it is not always obvious to me where to look
first. If i did, I wouldn't have sought the help of a discussion group
for things I could have answered myself. You have to expect to see all
levels of specialization among members of a public discussion forum.
Thanks for the tips,
Hazem |
|
|
| Back to top |
|
|
|
|
|
All times are GMT - 5 Hours
The time now is Mon Nov 23, 2009 2:16 pm
|
|