TEC's as heaters @ 60 - 70 C

Does this laser have an internal TEC element? If not, use a resistor!

No, I'm stuck with my previous design. It works fine. It's not so much the method of getting to 60C. But the optical/ mechanical stability... there's one tweaky lens position and if that moves around more with thermal cycling it's ... not good.

Right now they mostly tune up within a few degree's of 25 C.

George H.

I don't think it's stupid, because student lab gear isn't exactly 24/7 stuff. Even if the lifetime were reduced by 10x, it would probably still be working in the 22nd Century.

Cheers

Phil Hobbs

Damn, you get to play with all the neat toys.

Woah, I only hope it lasts 10-20 years. It's not so much the lifetime that I'm worried about. The diode laser is a technically hard experiment. (At least for some fraction of our customers.*) I'm maybe adding another layer of hard... That's great for some people and poison for others.

On the other hand 795 looks to be a harder wavelength diode laser-wise, and there are some users who might want it.

George H.

• one story from many. I got an email, diode laser wasn't working, lots of mode hops. I went through the steps to align the grating. (two knobs on a mirror mount, not hard at all for an optics person.) They got it aligned, thanked me, and told me they had now put a sign on it .. "Do not touch!" Sigh.. to me the idea is to touch things... break it and make it work again.

Thanks Tom, there are lots of neat things out there. (I can think of several lifetimes worth of stuff.. but I'm slower than some.)

George H.

whole thing? The TECs have a max temperature although I don't recall if

Here is an alternate solution... If you are cooling/warming with forced air, just make sure the air doesn't go below the dew point. As long as the air volume is high enough that the *air* temperature doesn't drop to the dew point there will be no condensation. The temperature of the device is irrelevant. Still air will drop to the temperature of the device, moving air won't if it is moving fast enough.

A TEC heater is marginally more efficient, if much more expensive.

r head.

It's not the temperature of the air that matters, but the temperature of th e surface - and the air in the boundary layer up against the surface, that is in laminar flow, and not moving very fast at all.

A better idea is to get your cooling with dry nitrogen boiled off liquid ni trogen. That is very dry indeed, and traditionally very popular for cooling optical components below the local dew point, for just that reason.

Did you read what I wrote? There won't be laminar flow and there will be no condensation if the air is moving fast enough. But I'm not sure he is using any forced air movement. Looking at the link he provided I'm guessing he just lets radiation and convection do the job.

..

ser > >>> head.

the

if

ed

to

f the surface - and the air in the boundary layer up against the surface, t hat is in laminar flow, and not moving very fast at all.

Do you know what you are talking about? There's always a boundary layer aro und any turbulent flow; the flow within the - thin - boundary layer is lami nar. You can't create an infinitely steep velocity gradient in a real fluid (or at least not without supersonic flow rates). Even a supersonic shock w ave has finite thickness - roughly the mean free path of a molecule in the fluid, some 200nm in air.

nk

b.

TEC's don't work too well unless coupled to low thermal resistance heat sin ks, and big heat sinks need respectable air-flow rates to do anything usefu l.

None of this is going to let you avoid condensation by keeping the bulk of the air-flow turbulent - even ignoring the fact that in optical systems, tu rbulence equals vibration and optical paths that move around.

And the condensation happens on solid surfaces, and it's their temperature (and the local dew-point) which determines whether you have condensation or not. Which is the point I was trying to make, and you don't seem to have t aken on board.

How about using a trick from the cold camera days of a slab of Perspex around the device making an airtight seal and dry nitrogen fill.

Beam exit side double glazed and at the Brewster angle if you have to avoid ghosts. ISTR 1" thick Perspex was a pretty good insulator.

Might be useful in a practical lab so that they can look at both lines.

I'm not going to argue this with you ad nauseum. There may be some microscopic layer that has little movement, but it is irrelevant to this issue. The moisture in that layer is very little and by definition if it is stationary more moisture won't reach the surface.

The bottom line is that you can have a surface below the dew point and not have condensation if the air moves sufficiently so that it never cools to the dew point.

Ok, enjoy.

ar...

u'd

g the

l if

rced

as

p to

of the surface - and the air in the boundary layer up against the surface, that is in laminar flow, and not moving very fast at all.

around any turbulent flow; the flow within the - thin - boundary layer is laminar. You can't create an infinitely steep velocity gradient in a real f luid (or at least not without supersonic flow rates). Even a supersonic sho ck wave has finite thickness - roughly the mean free path of a molecule in the fluid, some 200nm in air.

You aren't arguing - you are just stating your position, which happens to b e under-informed.

by

Twaddle. Even if the air were stationary - which it isn't - moisture could still reach the surface by diffusion.

Which can't - in fact - happen.

sinks, and big heat sinks need respectable air-flow rates to do anything u seful.

of the air-flow turbulent - even ignoring the fact that in optical systems , turbulence equals vibration and optical paths that move around.

ure (and the local dew-point) which determines whether you have condensatio n or not. Which is the point I was trying to make, and you don't seem to ha ve taken on board.

r head.

Hi Rick, Well I've got ~500 laser diodes with a range of room temperature wavelengths. (from 777 nm to 785 nm) (The diodes are now obsolete and this was a "lifetime buy" from a few years ago.) Most will tune up nicely to 7

The physics at 795 and 780 is similar, but different too. At 795 nm there should be a bigger CPT (Coherent population trapping) signal. And that cou ld lead to more sales. (I still have to check this..)

Re: blowing air... In theory that sounds nice. In practice, unless I keep a very gentle air flow I'm afraid it would lead to wavelength "wobble" in t he laser. (laser is used to observe ~10 MHz line width adsorptions, with a laser freq of ~3x10^14 Hz.)

I guess my biggest concern is how the mechanical stuff changes with thermal cycling... I was thinking I'd run one up and down today (0C to 60 C and ba ck) and see what happens.

George H.

r head.

Thanks Martin, I guess if I had to run cold I'd try something like that. (maybe a bag of silica gel to absorb the water?) It's already has a plexig lass cover (with a hole for the light.) I have taped a microscope cover sl ip over the hole, but as you say one then has to be aware of etalon fringy things... and reflections back into the laser.

George H.

Poor optical quality, though, if the light actually goes through that much plastic. I'd vote for a plastic tube with a microscope slide glued on at Brewster's angle.

As long as you're using a TEC that's assembled with hard solder, e.g. Marlow, Tellurex, or Ferrotec, but not Laird (Melcor), it'll take a lot of temperature cycling, so there's no reason not to do both lines. To leading order you aren't doing anything worse to it than running it the other way round, with a heat sink at 60 C and a cold plate at room temperature.

TECs make pretty good heaters, no problem there.

From the picture, it looks like you'd have space to shave a little wedge angle onto the end of the mounting tube and cement a thin flat across it. With dry air inside the tube, it should work fine. If you can sneak a bit of desiccant onto the cold plate, it'll pump whatever water vapour leaks in over time.

Cheers

Phil Hobbs

If the TEC is oriented so that its thermal expansion changes moves the laser axially, that might be a bit difficult to temperature compensate. (I have a bunch of $1300 TO-3 packaged, TE cooled diode lasers in my drawer that are essentially useless for that reason.)

If you can sneak a quartz tube onto the laser can (TO-18, I'm guessing?) and cement the lens and laser to that, it'll do pretty well over temperature.

Cheers

Phil Hobbs

Hmm, Well I'm afraid I'm using cheap Laird (Melcor) TEC's. I've never had any problem with 'em though. Typically not much thermal cycling.

George H.