project : vapour phase soldering
The motivation for this project was to save the travel time and expenses to my printed circuit board assembler that I tend to give smaller series of printed circuit boards to manufacture. Modern circuit cases tend to become harder and harder to solder. A simple and yet for hobbyists feasible approach is a vapour phase soldering machine. The basis of this approach is that vapour condenses at the coldest spot and thus deposits the condensation energy there. Besides that, no overheating occurs at any locations. Luck is with us as there is a series of liquids that boils at a selection of these temperatures where different kind of solder melts. Here I use Galden HT 230 with a boiling point of 230 degrees Celsius. For comparison, leaded solder melts at 188 degrees and leadfree solder melts at 220 degrees. This stuff is rather pricey, in the order of 100$ US per kilogram, which is about half a liter. All the other parts are rather simple and trivial.
The heater is a heating element from hot water jugs. Its nominal power is 1600W at 230 Volts.
It is mounted inside an iron ring to keep it under liquid. The electrical insulation from 230V is
made with cuttoffs from a printed circuit board (FR4).
As case, a chemisty beaker, high form with a volume of 3 liter was choosen for its transparency, which allows to visibly inspect the process.
The copper foil wrapped around was to increase the heat transfer at this height. I'm not sure it is required.
For the printed circuit boards, a tray from stainless steel was cut from hole-sheets.
Assembling the lot is not that difficult nor earth moving. So we just plug it in. Since the full power of the heating element would have been too much to handle, we did choose a variable transformer and after some experimenting found the optimum power to be achieved at 110V for heating up and 80V for keeping it hot.
Boiling is interesting to look at. The vapour condenses at the cold wall and shows visibly how high the vapour is.
The height of the cushion of the condensing vapour is dependent on the input power. The more the power, the higher it goes. Yes, the wires do have power on it. Some fearlessness from electricity is required at this point. I might have to look for some high temperature insulation at a later stage.
Yes, some pyrex glass tubes were added later. The machine is safe now. Not pictured.
A final look at the assembled apparatus.
Even though the boiling point is high, it still is a volatile compound and is to be filled back into the thight bottle after use, otherwise it'd vanish within a few days and keep on destrying the ozone layer.
One of the many chips that was successfully soldered was this leadless chip with 20 mil(0.5mm) pad center distance and 10mil(0.25mm) wide pads. There was absolutely no chance in doing it with an iron manually.
To be continued when I have a BGA or similar to solder...
The next generation
At a later point, I needed a bigger setup to accomodate bigger circuit boards.
The case was from stainless steel, with an airtight sealed lid, as used for food in professional environments. The heating was formed from an array of ceramic resistors rated 5W each in air. I did a few tests and found they do 30W in boiling liquid. The heating could be automated with thermocouples and a controller, but has been used manually with a Variac since.
last updated 5.nov.11 or perhaps later
Copyright (99,2006) Ing.Büro R.Tschaggelar