Home Page Email ONLINE GALLERY: New Work Artist Mission Statement Artist Bio Gallery Blog Wall Hanging Tips NEWS & EXHIBITIONS: News Exhibitions EARLIER WORK: Switchplates Chalkware Figures Pewter Figures Key Racks and Prints Of Reliefs Faux Uprooted Metal HOMEMADE MACHINERY: Spincasted Zinc Vulcanizer Spincaster CNC Router Injection Molder Foundry Compression Molder MISCELLANEOUS: Modeler Page No-Solder Live Steam Skeleton Key Collection Related Links Back

.

Wayne Pearson Relief Artist, Sculptor, Painter Injection Molder Click to enlarge         I designed and built this piston type tabletop injection molder to cast thermoplastics, to explore making plastic model kits.   I designed a type J (iron-constantin) thermocouple-based electronic temperature control (see below), using op amps, a beefy relay, and a galvanometer.   I fabricated a cylindrical heating chamber and piston using nichrome wire, plumbing piping, kaowool, and hand tools.   Molds were made from available aluminum flat stock.  I was able to achieve an adequate sliding fit for the piston and cylinder using emery cloth. Click to enlarge         This is a test wheel made from polyethylene with the injection molder.  This wheel has some flash because I didn't clamp the aluminum mold halves tightly enough (one of my tests). Here are 2 schematics for a thermocouple type temperature controller:       This is a minimal thermocouple schematic. An op amp amplifies the tiny voltage that is created when one of the two thermocouple wire junctions is hotter than the other. In your actual setup, if your meter keeps trying to go negative, reverse the connections of your thermocouple wire. The calibration potentiometer (CAL. POT.) helps tweek out any nonlinearities in your circuitry. The gain resistors (R1 and R2) keep the output voltage from the op amp in a range adequate to drive the meter full scale. The resistor values you choose will depend primarily on what you find to use as thermocouple wire. Any combination of two wires made of different alloys will yield a thermocouple effect, such as copper and steel, but commercial thermocouple wire will yield a predictable linear effect. I was lucky enough to have some Type J (iron-constantin) thermocouple wire in my junkbox. The zero adjustment (replaced by a thermistor in the next schematic) is used to correct for room temperature. Your thermocouple is actually trying to measure the difference between room temperature and the hot junction. Instead of using the ZERO ADJUST to adjust the meter to zero, a more precise actual temperature can be found by ZERO ADJUSTING so that the meter reads room temperature. The meter itself can be a simple 100µA movement. Here is the data for the thermocouple and thermistor I used: TEMPERATURE THERMOCOUPLE THERMISTOR 0° C       (0° F) 20° C     (68° F) 100° C   (212° F) 200° C   (392° F) 300° C   (572° F) 0.00   mVolts 1.59   mVolts 5.27   mVolts 10.78   mVolts 16.33   mVolts 27.28   Kohms 12.09  Kohms 0.9735   Kohms x x       This thermocouple schematic is what I designed and built for use in my homemade thermoplastic injection molding machine.   This thermocouple would also be very useful for controlling the temperature of a vulcanizer.   Of course, you can buy an industrial temperature controller.   They can be expensive though, and you will deny yourself the challenge and satisfaction of making one from scratch.   Also, there are simpler methods - like the method used in the homemade vulcanizer plans.       These schematics are intended as a starting point.   I have been able to keep temperatures within 20° F on my injection molder by using a beefy old relay to switch the 120 volt AC line on and off, sending current through nichrome wire to heat the plastic.   This is not an optimized or tricky design, just one that made use of what I had readily available for electronic components.   I used more op amps than I really needed mainly because they were available in the chips I had, and they kept the functions isolated, to make any troubleshooting easy.   I have made no calculations for component tolerances, since I was only planning to build one temperature controller, but I have included precise measured values of the components I actually used where I thought they could be significant. Copyright © 2004,2007 Wayne Pearson. All rights reserved.