As usually, we (especially I) will not be responsible for any damage to you,
your ego, your stuff, tools, house or anything else you own (or not own).
The things decsribed on this page may be potentially harmful to you and/or
your environment (including other people and/or animals).
Be careful and don't overheat the wax. Don't burn yourself or spill the
wax around as it is hard to remove.
Thus, I'm not responsible for anything. YOU HAVE BEEN WARNED.
My sister likes to melt old candles and dip new ones from time to time. If you don't know how it works: You just melt the old candle's wax in a water bath to prevent it from overheating and dip a wick into the molten wax. This way, the candle's diameter grows layer by layer, dip by dip.
As you can imagine, you need lots of dips to make a candle of normal diameter and it gets boring quite fast. (Some may call it ruminant, but I call it boring in this case.)
I dug out an old project which already had a MCU, a stepper motor driver, the matching stepper motor and shaft with ball bearings attached to the stepper. All I had to do was to remove some of the additional mechanical stuff which was mounted too and add a rope drum.
I made a small rope drum which consists of a piece of carboard tube and two bottle caps which fit exactly into the tube (by accident). I slid it onto the shaft and it fit perfectly.
The last step was to modify the MCU's program. The dip heights are now manually configured after power up and the MCU dips the wick in every approximately 30s. Additionally it stops a few seconds with the bottom of the candle just over the liquid wax surface to melt away the way-tails which otherwise start to form. The MCU runs acceleration and deceleration ramps (well, not really ramps, because I just increase and decrease the timer values which produces 1/x behaviour, but it works quite well) to prevent the stepper from loosing steps while accelerating and decelerating.
I didn't want to keep the water boiling all the time because it would waste lots of energy and you have add supplemental water all the time. I remembered that I once made a thermostat for PCB etching which was already set to about 50°C or 60°C (I don't remember exactly) but I never used. I modified it's setpoint to 90°C which suits the new needs.
The stepper could be somewhat stronger to dip thicker candles. Additionally you can't (of course!) leave the assembly completely allone but I can do something else while the machine dips candles.
To get a nice and evenly shaped candle, the wick must be taken out after the first few dips and rolled on an even surface to remove ripples. This procedure must be repeated every 4 to 5mm thickness growth.
I wasn't satisfied with the maximum load the stepper could handle but I didn't want to exchange it. The main problem with steppers is that they have the highest torque when they don't run. The faster you step them, the higher the impedance of the internal coils gets (and the back EMF increases too) and the coil current drops. The common solution is to use a power supply with much higher voltage than the steppers nominal rating and switching the coils on and off very fast so that the average (the inductance smoothes it) current is the nominal current. Well. I didn't have a way to measure the coil current on that board. Therefore, I made an open-loop control. I implemented a 8 level soft-PWM in the MCU, fed the board with 24V instead of the 12V nominal stepper voltage, measured the current and selected the PWM-level which produced the same current as with 12V operation.
While it's only holding the candle, it runs with "12V equivalent power" and if it starts to move it, the MCU switches to full power.
Additionally, I made a real tappet because the rope drum started to slip. (Formerly the drum was only held by friction.)
Well. Neither the stepper controller nor the thermostat were made for this project so I won't give you the details here. Maybe I'll publish them later, but there isn't anything special about them.
Copyright (C) 2009 by Wiesner Thomas
Last change: January 2nd 2009
