User:Jstapko/EngLab/ScrollSaw/PowerSupply/Ballast

Ballast Resistor Sub Project
The ballast resistor sub project is a component of the power supply project for the scroll saw based hot wire cutter. It is intended to reduce the output voltage of the power unit to a level which heats the hot wire element enough to cut foam quickly, but not enough to burn foam away from the wire, which has the undesirable effect of increasing the kerf of the cut.

Ballast Resistor First Attempt
During initial measurements, it was noticed that simply connecting a 20 amp ammeter into the circuit to measure the current drawn by the heating element reduced the color (and therefore the temperature) from bright orange to dull orange/red.(for more information on this effect, see Color and Temperature, by A. Gavrin and G. Novak)  Following this lead, it was found that the resistance of copper wire was enough to have a pronounced effect on the current delivered to the element. It was then decided to try to build a resistor out of custom made lengths of copper wire. The first attempt was only to check the concept, and consisted of a mass of clip leads and battery cables connected just well enough to prove the concept, not nearly a permanent solution. The results were encouraging enough to invest time in more permanent version.

Second Attempt
The second attempt was to use Ethernet cable scraps of 4 inches to a foot in length, (very rough estimate) with the intent being to wire all strands in parallel so that they could carry the heating current of roughly 5 to 6 amperes. At least two hours were spent stripping, un twisting, and re twisting, a dozen or so scraps of Ethernet cable to effectively make single conductor heavy gauge stranded wire out of them. This task was completed, but was ultimately abandoned on the grounds of several drawbacks:
 * The wire turned out to actually be much heavier than needed to carry the 5 amperes, so more linear feet of it, and thus more weight/bulk, would be needed to make the required resistance than would be needed with smaller wire
 * A brief search for current carrying capacity of wire lead to a discussion suggesting that wires of a multiple conductor cable wired in parallel to act as a single one runs contrary to recommendations of the National Electrical Code (possible link: National Electrical Code online)
 * To attach the Ethernet scraps together, they would need to be soldered and shrink tubed, but it should be possible to find a longer single length of insulated wire, which would prevent the need to consume nonrenewables such as the materials needed to make shrink tubing, run the soldering irons, or manufacture solder.
 * Finding a wire that more closely matched the needs of the project might also reduce the labor time necessary to implement it, which would boost productivity.

Third Attempt
The third attempt at the ballast resistor is based on an electrical appliance line cord, with a grounding wall plug (NEMA type 5-15P), with approximately 16 gauge wire. (more information on NEMA plug designations) In the first test, two conductors of the 3 conductor cord were placed in series with the power unit, and the arrangement was tested. The first challenge in this effort was to connect the power supply wire to the wall plug prong on the line cord. They could have been soldered, but it was desired to keep any changes as reversible as possible, and it is nearly impossible to get all solder off of a soldered component without sanding it off. In the case of the line plug, sanding could damage the metal plating that is put on the prongs to inhibit corrosion. There is also a hole in the rectangular prongs of the line cord, to accommodate a locking device (link to locking device) The presence of the hole raised the possibility of connecting the power supply with screws, as on binding posts. There were 3 options for attaching wires to the screws:

The holes are quite small, so the screws were necessarily small. When the first option, wrapping by hand, was used, there was a tendency or the wire to spread apart and come loose under the screw head. Either of the other two options should improve reliability. It was decided to use solder instead of a ring lug. because if the system didn’t work, a ring lug would then have been needlessly destroyed, while solder could be undone. It was also partly based on the belief that the amount of solder used would be cheaper than ring lugs to replace, although this assumption could erroneous. When I attempted to solder the wire ends into loops, I found the soldering iron tip to be completely untinnable, and had to take time to clean the tip. With the tip cleaned, I proceeded to solder the wire into loops and bolt them to the ends of the flat prongs of the appliance plug, through the holes provided for a locking device. This mounting arrangement worked quite well, but two sections of the wire was insufficient resistance; the heating element was still a bit too hot. To add a bit more resistance, I connected the ground wire of the appliance cord in series with the two power wires, but ultimately had to solder a jumper between one flat prong of the plug and the ground pin, because the ground pin had no hole through which to put a bolt. Thus, ultimately both screw connections and direct soldering were used to connect all three wires of the appliance cord into the circuit. This still proved to be insufficient resistance, necessitating the development of alternate plans. Some ideas for alternate plans include:
 * Wrapping around the threads by hand and tightening
 * Crimping a ring lug to the wire and screwing it to the prong
 * Soldering the wire into a ring shape;
 * discontinuing efforts in the current direction until it is established that the current power supply components can deliver the needed current to the new heating element

using wire nuts on the cut end of the cord
 * a fixed inductor, using wire wrapped around either a bunch of nails or printer rollers, held in a wood frame
 * a variable inductor, in the style of the variometer, with one winding pivoting with respect to another, held fixed
 * a variable inductor in the style of the constant current regulating transformer
 * a variable inductor in the style of the fixed inductor, except having a movable tap as in the Lionel train rheostat or the [James G. Biddle rheostat]
 * a variable resistor, based on the [General Electric Resistance Bridge], in parallel with the fixed inductor. This is to allow the bulk of the current to pass through the inductor so as not to overheat the bridge resistance elements, but to allow for very fine control of heat by using the bridge controls.

Next Steps

 * check to see if 6.3 volts heats about 4 inches of the smaller nichrome wire to cutting temperature
 * look for coil forms or materials with which to build a frame for the inductor or variometer
 * draw a schematic for the current power supply as found
 * draw a schematic for the General Electric resistance bridge
 * try to find a manual for the GE resistance bridge
 * look for long lengths of 16 gauge or so wire
 * try to take high quality images of my own Lionel rheostat with which to build a wiki page on the lionel rheostat for future reference
 * try to find an iron cored variometer picture
 * build a page on the fixed inductor sub project, featuring the images from coil winding and experiments with the cord
 * build a wiki page on the J.G. biddle rheostat, with images of my personal unit

Things this page needs

 * link to wall plug lock, ideally one showing cross section and explaination
 * link to nice page discussing American Wire Gauge and giving sizes, resistance per thou ft, weight, etc.
 * pictures of work done on the project so far
 * links back to the power supply project page, the scroll saw project main page, and the heating elements page
 * amendment to the story to complete it, possibly with a link to the soldering iron tip cleaning effort
 * discussions of future directions, such as
 * a link to the soldering iron tip cleaning page, which first needs to be created
 * a link back to the engineering lab projects home

Back to Scroll Saw Main Page