Measuring temperature from heater?

March 25, 2005 9:26 AM

Where can I find a thermometer which will accurately measure infrared heat in 10 seconds or so, and reads in the 300-1000 degrees Fahrenheit?

I'm building an infrared paint remover. I'd like to measure the temperature that the paint actually reaches using my setup.

Extra credit a) How do I limit the heat output of a quartz infrared heater element electronically (modulate voltage?)

Extracredit b) How do I build a 'cut off' circuit? Basically, I want to string a therocouple out underneath the heater elements such that if the temperature gets too hot I cut power to the device.

posted by daver to sports, hobbies, & recreation (10 comments total)

bolometer might be a word to google for?
(they can measure the heat on a polar bear's feet at a distance of half a kilometer. or so i'm told.)
posted by andrew cooke at 9:57 AM on March 25, 2005

I don't know anything about thermometers.

I know abit about the extra credit parts though. The cutoff part is fairly simple in theory. You need two parts: a potential relay and a magnetic contactor. The contactor will probably run you $15-30 bucks. It's basically a magnetic switch: you input your AC 120V on one side, and output it it on the other. There will be two inputs that when voltage is applied across them, the switch closes magnetically (slams shut actually). You may be able to find surplus versions of these. They are typically used on power machinery, the reason being that the voltage that closes the contacter can be very low while the voltage that is being switched can be very high. This means, for example, that you can turn 480V 3 phase current on and off using a regular light switch and 120V power, keeping yourself well away from the high voltages. Now, the potential relay is basically a device that will turn on when the voltage reaches a certain amount. I don't remember how most thermocouples work -- does the voltage drop or rise when the temperature rises? This will change whether you want the thermocouple to be in series or parallel with the potential relay. EIther way, what you're trying to do is set up a situation where when the thyristor causes the voltage to climb or drop to a certain point, the potential relay turns on, and the contactor turns on or off.

This is a very broad overview, just me thinking out loud about how I might actually do it. Actually doing it would probably take me several hours of thought and some research. It's been a while since I've fiddled around with this kind of thing.

I tried to do some research as to how to control the temperature of the quarts. I suspect you may be right with your voltage theory, as that would definitely be the easiest way to build temperature controls into a heater like the one you have pictured. You could try your theory.... you'll need either a potentiometer (a variable resistor) or a switch with several settings (usually a rotary switch). The switch with several settings would probably be fine and would lend itself better to experimentation. Keep in mind that with 120VAC resistors can draw a lot of current and thus dissipate a lot of wattage (heat). I built a phase converter recently and due to the way I had some resistors mounted, they burned the paint on the metal box they were in. Would have melted stuff if left to their own devices.

Anyway, if you're trying to modulate the voltage with resistors, you'd basically want to put resistors in series with the heating elements. The more resistance you have in series with the elements, the more voltage gets dropped across the resistors, and the less gets dropped across the elements. Your multi-way switch would basically control the path of the current. Say you have a 4-way switch. position 1 could be off, 2 could be 1Kohms, 3 could be 10Kohms, 4 could be 50Kohms, etc. You'd probably need some time experimenting and a good multimeter to find out what is going to work for you.

Good luck, have fun and please, be safe.
posted by RustyBrooks at 10:00 AM on March 25, 2005

Extra Credit B is a pretty easy one. There are tons of different temperature switches which do exactly what you need. The have the thermocouples and circuitry built in.

An infrared thermometer depends on knowing the emissivity of the surface it is measuring. Emissivity ranges in value from 0.0 to 1.0, where a black body is 1.0. If you're measuring a painted surface, and one that is change state (from solid to liquid) and maybe color, the infrared thermometer is going to give you inaccurate readings.

I am assuming you have googled "infrared thermometer" and found the options out there. In general, I don't like them, except for very specific, controlled cases. They are good for measuring engine temperatures, for instance.

I second what RustyBrooks says, be safe. Playing with voltages and resistance, intentionally creating heat is a deangerous combo. Please make sure you are comfortable with what you are doing.
posted by jonah at 10:15 AM on March 25, 2005

Infrared heat? Isn't infrared light/heat measured in Kelvins?
posted by fandango_matt at 11:30 AM on March 25, 2005

If it is of interest, I have a Craftsman 82327.

As far as I can tell, It seems to work reasonably well.
posted by Boobus Tuber at 12:32 PM on March 25, 2005

Cool, thanks for the suggestions so far! Trust me, I 'm planning on being safe and careful.

Let me re-state my thermometer problem. What I ultimately care about is the temperature of the paint. Most of the thermometers I can by at a store around here seem to top out around 600 degrees. Any thoughts on where I could find a contact thermometer which would measure upwards of 300 into 1000 degrees F?
posted by daver at 12:40 PM on March 25, 2005

Omega is the source for all your temperature measurement needs. It looks like you'll be pushing upwards of $700 to get an infrared scope capable of doing what you need it to do. You should be able to do it with a thermocouple for much cheaper if contact is acceptable.
posted by mr_roboto at 12:57 PM on March 25, 2005

Here's the problem: infrared heat is merely radiant energy. It doesn't really have a "temperature" until it strikes an object and excites it. You could get a thermometer, but would be able to measure only the ambient heat (i.e. the temperature of the air surrounding the thermometer), or of whatever the thermometer itself heats up to when struck by infrared light.

You want the temperature of the paint, which will depend upon its color, water content, etc.

It seems to me that the craftsman dealie, above, is probably the closest you're going to come to what you want to know...
posted by curtm at 2:15 PM on March 25, 2005

The Raytech Laser Thermometer reeds temperatures up to 275 C (525 F) with reflected laser light. All the professional cooks (and amateurs with around $180 to throw around) are crazy about it.

www.raytek-northamerica.com/cat.html?cat_id=2.3.5&PubSessID=034ca17fdbd07b2ee90d3cda56c1a046&PHPSESSID=publicRaytekNorthAmerica

Models that read up to 600 C (1100 F) and 760 C (1400 F) are available, but at MUCH higher prices.

www.flw.com/rayt_ol1.htm
posted by KRS at 2:28 PM on March 25, 2005

A thermocouple in contact with the paint being stripped might be a way to go if a noncontact thermometer turns out to be too expensive. They can be used to measure fairly high temperatures (depending on the exact metals they're made of). They're also pretty small. Normally the thermocouple itself and the thing that connects to it and displays the temperature are separate. I don't know where you'd go to buy a set though.

Re controlling the temperature: instead of a variable resistance, I'd go with an on-off controller that cycles faster than the element itself will change temperature. (Say, a few times a second.) If you use a resistor, you'll end up dissipating a lot of heat in the resistor itself (possibly comparable to the amount of heat emitted by the quartz heater). As a first try you could use an incandescent-light dimmer switch, as long as it's rated to handle the wattage.

As for extra credit (b), is this a safety device, or a temperature-regulation device? If safety, you could look for a 'thermal fuse' (ideally a self-resetting one, at least for the experimentation phase).

curtm: These noncontact thermometers are basically calibrated bolometers. They heat up from the infrared radiated by the target, and by measuring that temperature rise and making a double handful of assumptions about the target's emissivity and so forth, they display an educated guess of the target's temperature.
posted by hattifattener at 12:25 AM on March 26, 2005