Homebrew PCBs

> One important thing to know: it's calibrated for an emmissivity of
> 0.97. That means anything light colored is going to read as a lower
> temperature than it really is. Like a silver bottomed iron. Easy fix-
> stick a piece of dark heat resistant tape to the iron. Measure on the
> tape, then measure on the untaped silvery bottom and you have a fudge
> factor. Now you don't need to put the tape on the iron again.

The fudge factor will work as long as the temperature you are trying to
measure is reasonably close to the temperature used to create the fudge
factor. If it is different by a hundred degrees or so, the error can be
fairly large. Their are some surfaces that do not lend themselves to
measurement via single band IR, as the emissivity is too low, and the signal
gets buried in noise. A good example is an uncoated stainless steel oven
shroud. Often times it will act as a mirror, and something as simple as the
distant reflection of a gas furnace will cause huge temperature shifts. Now
if you cover the stainless steel with all sorts of dirt, the effective
emissivity increases, and an accurate measurement is possible. Therefore,
factory cleaning can cause severe headaches and distress for the tech
support guy..... and some level of bewilderment for the design engineer.

I spent 5 years designing these type of units. They are a lot of fun to use,
and will cause much loss of hair during design and will cause many
challenges for tech support.

Other useful tricks.
WD40 works as an emissivity enhancer in some applications, and can easily be
cleaned up.
Dr Shoals foot power is perfect for normalizing the surface emissivity of a
complex shape in order to measure gradients if the spot size is small
enough. Eg, a populated pcb, or even for reflow oven setup.
Red Devil BBQ black is an ideal coating for high temp surfaces, as its
emissivity is constant, unlike many other coatings that have emissivity
changes over time and temperature cycles.
Anodized aluminum irrespective of color or clear has an emissivity of around
0.9
Sandblasting works well to improve the surface emissivity of most metallic
surfaces

Warnings:
The spot size as describe in most literature is not very accurate. Some
times the spot size is determine by a 50% fall off of signal level. Other
times it is 67%, 95% or 99%. The spot size surface distribution is not a box
car shape, it is a function of careful optical design. As such, many times
advertised specs are pretty worthless. Testing in the application is the
only solution.

Emissivity has spectral components, and directional components, and as such
emissivity tables are usually only ball park figures. Not sure if the
Craftsman unit will measure low temperatures, but snow and ice are quite
interesting. Mount a unit on your car and you will increase your personal
safety while driving in bad weather. The snow plows here in MN use IR
thermometers to help control sand/salt mix and deposition rates.

Do not take your unit out in the cold, and then measure something. Most
units require ambient temperature stabilization times of 30 minutes to 90
minutes for accurate measurements. This was a real challenge as our sales
guys would keep the units in the car, then bring them into a customer demo,
and they would drift until they stabilized at the new ambient.

Other interesting facts
The best book on IR thermometry was written by Dave DeWitt when he was at
Purdue. Last time we got together he was finishing a sabbatical at NIST, and
was planning to move there for a full time position, but that was years ago.

There are less than 20 manufacturers of IR thermometers. It is a closely
knit group of some really great guys. We used to meet every year at NIST or
the ASTM meetings and as a team were writing a users guide for IR
thermometry. I went to work for another outfit before it was finished, but
the guide had a lot of potential.

Ray Peacock has an excellent writeup on IR thermometry on his web site.
http://www.temperatures.com/ncsensors.html He was the chair of our ASTM
working group, and also teaches courses for ISA.

Have fun with your unit
Ron