DIY MULTIMETER THERMOCOUPLE PYROMETER
The pyrometer is built from a digital multimeter that is capable of reading a type K thermocouple and converting it to degrees Fahrenheit. It does not need to be calibrated as long as the appropriate thermocouple is used. It is very similar to Richard's from Hobbicast. He gave me lots of good advice on the purchase and construction of the thermocouple. Thanks Richard. The multimeter is from Circuit Specialist. Model number CSI9903. It cost $29.95. It is of very good quality and is a good tool to have for its other capabilities as well. It comes with a thermocouple probe but it is light gauge and I doubted that it would perform to the rating of 1832 F (1000 C). I therefore made a higher temperature thermocouple for it without trying the supplied unit. It might perform well though. I wanted to keep it so I could compare future homemade thermocouple readings since thermocouples do not last forever. With shipping the total cost of the multimeter was $36.95.
Harbor Freight tools sells a similar multimeter as well. It is part number 42658-2VGA and sells for $19.99. Enter the exact part number in the "Order From Printed Catalog" and the sell price of $19.99 should come up instead of the normal $32.99. I chose the Circuit Specialist Model over the Harbor Freight because it reads in Fahrenheit and I hoped it might have better quality than the HF model. The HF unit only reads in Celsius. I do not know if the quality is better. Sears also sells what appears to be a similar meter as Harbor Freight (item number 03482322000) for $59.99. Typically Sears sells good tools so it might be good quality. Sears also carries a Fluke meter for $129.99 (item number 03481336000) as well as other similar units.
A thermocouple is simply two wires of different metals that are joined (melted) together on the sensing end. The theory is that when heated a small thermoelectric voltage is produced in the center zone of the decreasing temperature gradient of the wires.† The welded tip does not produce the voltage.† Thus a third material could be used to weld the tip without affecting the voltage.† Manufacturers go to great lengths to produce perfect welds of the wires but a torch should work for us.
†Since the material of the wire is known a unique voltage is read and interpreted into degrees by the multimeter. Thermocouples typically come in several standard types.† These produce a set voltage at a given temperature so that a monitor designed to read the type of thermocouple can display it accordingly. Some monitors can selectively read several different types of thermocouples.† In my case the digital multimeter only reads a type K.† There is a chart at http://www.yankeecontroller.com/color_code.htm of standard thermocouples with their properties and color codes.† A normal multimeter could read a thermocouple voltage but it would have to be converted and calibrated from the voltage to temperature. The connection of the thermocouple wire to the multimeter contact my also behave as a thermocouple itself so calibrating and connections can become troublesome. Since many multimeters are capable of reading thermocouples and do not cost much I would highly recommend purchasing a dedicated unit opposed to reading the voltage and interpreting with a standard multimeter.
These are the actual thermocouples. They do not have a sheath on the sensing end like some. The one on the left is supplied with the multimeter. The wire on the right I purchased and melted the tip to create the couple. I also mounted a similar plug, which will connect directly to the meter. I wrapped the sensing end of the couple with stainless steel wire. This secures the ceramic insulation from fraying and offers higher durability. The wire is extended outside of the tubing in the picture below and will be coated with a refractory wash.† The tube will be grouted with a high temp grout to prevent molten aluminum from going inside.† These will offer some protection from the corrosive aluminum however exposed light gauge wire like this will still not last very long. †I have thought about embedding the thermocouple completely inside of the grouted tube.† This will slow the response time and perhaps not really necessary but would hopefully extend the life of the sensor. I will add the results of such if I do it. †The steel probe will also be coated with a refractory wash on the outside to prevent corrosion and contamination.
I purchased the thermocouple wire from McMaster-Carr Company.† Another excellent source for such is Ebay. There you can find good deals on thermocouple wires and actual probes sheathed in stainless steel.† I used the 20 Ga. ceramic fiber type Kk that is rated to 2200F (part number 3870K66). It cost $2.50 a foot. I ordered 6 ft and made two thermocouples out of it. Since shipping is a large percentage of the cost I decided to stock up and get two. Thin gauge thermocouple wire like this will not last for too many measurements.† I also ordered a Type K, flat-pin, mini male plug (part number 3869K35), which plugs right in to the meter. It cost $4.07. The plug is yellow which is a standard color code for a type K thermocouple. I could have used the one supplied with the meter but then I could not interchange them easily. The thermocouple I built reads the exact same temperature as the supplied one at room temperature. A good sign I thought. Both seem to be correct. With shipping the total McMaster Carr bill was $19.97.
In order to join the two wires on the sensing end all I did was twist them together and touch them with a cutting torch. This melted the wire together into a small bead. It works very well. I left some length on the wire so that I can redo the end if it needs it. The 2200F rating allows for the thermocouple to be directly immersed (with a refractory wash) into the aluminum, which is what I built it for. No probe or sheathing is necessary but the thermocouple will have a limited life. I did construct a wand out of 1/2" 14 ga. square steel tubing that I had on hand. It would be better if it were stainless steel since ss seems to withstand heat better. I believe both metals can still contaminate the aluminum slightly with iron so a refractory wash should be applied which will help prevent corrosion.
With the meter removed. The wire comes out of the square tube and plugs into the meter. The steel base frame provides a secure mount and protection from abuse. It stays in very well. I secured it with some wire like so. I painted the base and upper wand with high temp paint shown below.† I did not paint the submerged portion of the wand.† I purchased the paint that is rated to 1200 F to paint my burner tube with.† Therefore I did not include it in the cost of the unit.† Painting it is not really necessary but makes it look nice.
Typically I plan to insert the wand into the molten metal through the exhaust. I put a piece of sheet metal in the base so it might perform as a heat shield. The wand is bent so it allows the meter to be positioned away from the rising heat. I plan to add a bicycle grip to the handle.
I tested the unit in the furnace air and it went slightly over 2300F for a brief time. I did not mean to get it that hot but it was there before I knew it. That was 100 degrees over the insulation rating. It performed great. I will try not to take it that high again though. The multimeter says its range is to 1832 but it read to over 2300. I guess the rating is for accuracy or the supplied couple. I plan to use it for aluminum only so the 2200F-degree rating should be much higher than needed. Works great so far.
The total out of pocket cost: $56.02
This included all shipping charges. This did include 3 thermocouples as well. The thermocouple supplied with the meter might work fine. If so a good portion of the cost could be shaved off.
I consider it a success. I think it will be a good tool that should last a long time with some preventative measures like a refractory wash.† Even then new probes will have to be made.
If your budget allows I would highly consider the MIFCO unit, which can be found for about $230 for the digital and $200 for the analog (I hear). It is much smaller, lighter and specifically designed for foundry use. I have heard very good things about it. It uses a true probe that would most likely perform very well for a long time. It would most likely be a good investment. I am just cannot justify it at this time. I do plan to purchase it someday though.† If you are not using it for professional work and are in the market for a good multimeter anyway, then the multimeter pyrometer might be an excellent option.† You can have a pyrometer for basically the cost of the thermocouples.† At least that is how I looked at it.† I was about to purchase a new DMM anyway since my old multimeterís screen was on the blink.†