Coating Linearity (White Uncoated TECs)

[lasersbee]

I thought I'd share this finding...

We were doing some Coating Linearity tests yesterday
using the new Digital Adjustable Laser Power Supply
I had built for the shop in my other Thread.

I decided to do a quick test of some Uncoated White
TECs to see if they were Wavelength Broadband and
Power linear.

Here are some basic results I got at 1000 mW :-

White Uncoated TEC

808nm - 1000mW
520nm - 1062mw
450nm - 1184mW

As a comparison here is our AX3 coating at 1000mW :-

LaserBee AX3 Coated TEC

808nm - 1000mW
520nm - 1008mW
450nm - 1010mW


Jerry

 

[Alaskan]

Sure does a great job!

 

[paul1598419]

Were one of the uncoated TECs the one astralist uses in his 20 watt setup? Just curious.

 

[lasersbee]

Sure does a great job!

Not quite sure what you mean....
The uncoated TEC skews the 450nm reading
by 18.4%...

Were one of the uncoated TECs the one astralist uses in his 20 watt setup? Just curious.

I would assume it may be the same...
I used one of our 15mm X 15mm TECs
we use with some of our LaserBee
products.


Jerry

 

[Alaskan]

I meant the AX3 coating does a great job.

 

[lasersbee]

I meant the AX3 coating does a great job.

Got it...
Yeah the AX3 coating is quite Broadband linear
and it can handle a 3mm beam at over 5 Watts
without smoking or damage.

We are working on a new coating that will handle
a 3mm diameter beam at over 10 watts. It is not
yet ready for the market.

[EDIT]

Were one of the uncoated TECs the one astralist uses in his 20 watt setup? Just curious.

From one of Astralist's Posts....

Technical specification:

36 mm² high speed sensor (default)
100mm² slow sensor.

No coating at all, yielding broadband rejector instead of broadband absorber with the highest damage threshold among the three (damage threshold of alumina ceramic Al²O³). Therefore producing higher measurable power albeit at worse resolution (13 uW).

It looks like he only uses 2 different TECs... A 6mm X 6mm
TEC and/or a 10mm X 10mm TEC.

I don't understand the Broadband Rejector reference...:undecided:
Couldn't find anything on Google...
If the TEC sensor doesn't absorb the Laser's photons it
won't detect the Laser's beam power.


Jerry

 

[paul1598419]

I believe he means rejection in that it is white and, therefore, won't absorb as much of the heat as a black one will. I could be wrong and there maybe some other property of the TEC that I don't know yet. Maybe something in the coating itself, IDK.

 

[Benm]

I thought I'd share this finding...

Here are some basic results I got at 1000 mW :-

White Uncoated TEC

808nm - 1000mW
520nm - 1062mw
450nm - 1184mW

I presume you used the 808 measurement as the reference?

I'm actually surprised that the results are not -that- far off without any coating on the TECs. It would be nice to have a figure for red (635 or 650 nm) as well, but it's not that awful to begin with.

Sure, 450 nm reads an unacceptably high number being 18% off, but it's not like you're getting double or half the actual power either.

Since the coating sort of limits the power limit for meters i can see reasons to do without it - you'd have to enter the wavelength on the meter somehow, but beyond that it seems perfectly acceptable.

Problems come with mixed wavelengths obviously, like dpss greens, where some (optical) sensor systems show very different sensitivity between say 1064, 808 and 532 nm which all come from the same laser.


That said the coating yielding uniform results to within 1% is impressively good.

 

[lasersbee]

I would have done a test in the RED wavelength
range if I had a high power RED Laser in the shop.
My Oclaro doesn't go higher than 560mW with a
G2 lens. Need to build a > 1Watt Red...

Jerry

 

[Benm]

I suppose you could do the 500 mW test and extrapolate the results from that? Or you two lasers to get the combined power of 1000 mW onto the sensor if you have them - i doubt it would matter much if you did so.

I think it's interesting to look into this a bit further since uncoated tec's can usually handle a lot more power density than coated ones. Obviously the very linear response from the coating is nice if you don't want to input the wavelength somewhere or don't really know what it is (or even if it's mixed color like a projector).

On the other hand getting the wavelength response curve for the uncoated version is interesting for higher power measurements that would just burn the coating off otherwise.

The mixed output of dpss used to throw off wavelength sensitive meters like the coherent lasercheck, but with all these direct diode ones we get now it could be interesting to get extra power handling at the cost of having to somehow specify the wavelength so the meter can compensate for it.

 

[paul1598419]

If I need to make concessions by using a chart for different wavelengths on the uncoated TEC I don't have a problem with that. It might even be easier to have it in the firmware, or at least in the software, making it even easier to plug in the values.

 

[lasersbee]

I think it's interesting to look into this a bit further since uncoated tec's can usually handle a lot more power density than coated ones. Obviously the very linear response from the coating is nice if you don't want to input the wavelength somewhere or don't really know what it is (or even if it's mixed color like a projector).

Once I find time to assemble the new 1W Red LD
we bought recently I'll test that Wavelength.
BTW... I'm assuming that white uncoated TECs
all use the same ceramic materials for the plates.
I could be wrong...

In the mean time... you are welcome to do these
tests on your own and post some results....:beer:

If I need to make concessions by using a chart for different wavelengths on the uncoated TEC I don't have a problem with that. It might even be easier to have it in the firmware, or at least in the software, making it even easier to plug in the values.

The problem that I see is that since most of the
Laser's beam is being reflected due to the white
surface therefore the Differential signal between
the plates is greatly reduced. This increases the
signal to noise ratio... for lower powers this can
be a problem without amplification and a lot of
complicated filtering to keep the reading clean.

Jerry

 

[Benm]

The coated ones clearly are much better for lower power lasers, and the linearity of the coating adds the convenience of not having to enter the wavelength into the meter (assuming you even know it, and it is a single wavelength).

The uncoated ones have the advantage of higher power handling, but still come with some uncertainties too: Like if the optical response of the coating would change over time - perhaps it tarnishes very slowly in the air, or perhaps it gets altered by exposure to extreme light levels over time.

Deterioration of the coating could be tested by exposing it for a long time to intense laser light (say do a month long full blast test and see if the results are off after that), but that would only reveil deterioration due to light (and heat to some degree), not chemical processes that may happen over longer periods regardless of light.

 

[paul1598419]

Clearly, the uncoated TECs would be used for higher power lasers....or at least that is my intent. There are some filtering in the software, but I have not looked into this at all yet. The white, uncoated TEC was an after thought when I decided to purchase the high power coated one. It may take a table of wavelengths to use with the uncoated one to get a more accurate reading, but that would be okay with me. I certainly have more than my fair share of LPMs to measure the powers of lasers at many different levels.

 

[WizardG]

I've got a bunch of tungsten carbide plates, 30 x 30 x 0.5 mm. What about using something like arctic silver epoxy to glue a thin carbide plate to a TEC sensor? These plates are a dark grey so, at least in the visible range, I wouldn't think there would be too much measurement shift due to wavelength, and they can take an awful lot of heat.

 

[lasersbee]

One other problem with using a 0.5mm Plate glued to a
TEC that I can see is that it greatly increases the time
to get a stable reading due to the additional mass in
front of the TEc's ceramic plate.

Another problem I can see is the melting temperature
of the solder used to assemble the TEC... Going over
that will destroy the TEC and it's characteristics.

Jerry