If you want to embark on this project be very careful because it uses dangerous mains voltages. You have to know how to deal with.
Also, NEVER connect the Arduino to the PC while leaving the degausser's mains cable plugged in.
The usefulness of the degausserWhen milling, the parallel blocks or the vice jaws can become magnetic. Milling chips stick to them and are difficult to remove. This demagnetizer applies a decreasing alternating magnetic field. At the end of the cycle, the remaining magnetization of all inserted steel parts has disappeared. Excluding neodymium magnets or other permanent magnets.
For the vise jaws, you have to take them out of the vise... or you'll make a bigger coil !
Theory of operationA zero crossing detection circuit provides a "LOW" pulse through an optocoupler to the interrupt input pin 2 of the arduino.
The Arduino uses Timer1 to trigger the triac with increasing lag behind the zero crossing interrupt. The lag grows until it is equal to half a cycle of the ac voltage. So the alternating magnetic field decreases down to zero. The detailed timing sequence is described at the begin of the software file "AC Phase control for demagnetization".
The triac is triggered by means of an opto-triac MOC3021. A low-pass filter (R6-C2) rejects short voltage spikes. R7 limits inrush current when the opto-triac is fired. The two series resistors R11 and R14 are optional. They are in fact not needed as the impedance of the coil itself is about 27Ω at 50Hz.
The whole degaussing sequence is about 2s. The degausser remains inhibited for 2.5 seconds more to allow the generated heat to spread. This can be useful if lot of cycles are launched one after the other.
The coil is made of 4 wafers of 250 turns each of enamelled copper wire 17AWG (~ ø 1.2mm). They are individually insulated with tape and connected in series. This will help not to have to high potential gradient between the wires inside each wafers.
For winding the wafers, a wooden core serves as a winding template. The corners of the core must be rounded so as not to damage the enameling of the wire. In my case, I needed dimensions of 128x40mm for the core. If you need to demagnetize smaller or larger parts or of other shapes, you can adapt dimensions. Also, the gauge of the wires could be reduced to 20 AWG (~ ø 0.8 mm) because high current only flows for a short time and the warming up will be minimal.
But if you change significantly the dimensions, you must measure the resulting inductance and DC resistance of your coil and adapt the number of turns so that the impedance is roughly similar to mine (about 27Ω at 50Hz). More than 20Ω is desirable since we deal with 220Vac mains voltage. The impedance calculation formula is given above...
For the photos I used rectangular flanges. The wire tends to get stuck on the corners while winding. To wind more easily, it is necessary to round the angles and chamfer the flanges inwards
Four (or more) pieces of strong cotton or polyester yarn are first fixed on the core with smal tape pieces. They will serve to hold together the copper wires with a tight knot before removing the coil from the wooden core. The coil is then surrounded with electrician's tape. Finally the 4 wafers are connected in serie and hold tight together with electrician's tape. Pay attention to the direction of the winding when connecting them: the winding end of one shall go to winding begin of the next. One way to check this after connecting, is to power the coils sucessively 2 per 2 with a low voltage DC power supply (first and second, second and third and finally third and fourth). Each time you sense the magnetic field in the center of the coils with a screwdriver or any other iron or steel piece. If the connection is made the right way you can feel the attraction of the magnetic field. If you don't feel it, either the current is much too low or the connection is reversed.
The control board electronics is mounted on a perfboard. To maintain good insulation the unused copper pads are removed with the solder iron. The voltage dependent resistor VR2 is directly soldered on both ends of the coil. The two 2R2/50W resistors are mounted on a 52x80mm aluminum plate 1.5mm thick to diffuse the little generated heat. This plate is screwed on the bottom with 4mm thick nuts as spacers. You can see the two white wires comming from the resistors in the photo below.
For the optocoupler with transistor output you can use mostly any 6 pin optocoupler. Maybe also 4 pin optocouplers could do the job. I used an SL5500 but I've tried with 4N25, CNY17, MCT2 and TIL111. All of them were working well. The voltage dependent resistor VR1 is optionnal; I had not mounted it.
Left, the big yellow part: the home made coil. The black part on the top, next to the right is a salvaged 500mA USB power adapter. The plug was removed and the mains wires directly connected to the PCB. The USB socket was also removed.
To prevent damage to the coil from impact with steel parts, the inside of the coil is protected by a 3mm plastic shell - the gray parts screwed on both sides to the black parts (at the bottom of the photo).
Comments