The accurate transfer of fluid materials has been a human endeavor since the beginning of time. The caveman dipped a stick in his container of coloring and dabbed it on the cave walls. Surprisingly, that dispensing technique is still used today in the form of “pin transfer” — dipping a fine pin into an elaborate fluid pot and then “touching off” the fluid onto substrate transfer fluid materials. Looking closely at the opportunities for inaccuracy, however, shows the limitations of this dispensing technique.
A more immediate predecessor to current dispensing techniques is the needle inoculation, similar to a medical injection. The hypodermic needle allowed one to position the fluid exactly where it was required and then push the plunger to urge the fluid out of the barrel. The best part was that the syringe barrel and hypodermic needle were inexpensive and could be thrown away when finished — cleanup and no cleaning solvents required. You can imagine the first fly in this ointment. The hypodermic needle was designed to pierce skin and muscle and slide right in — not exactly a quality for an industrial dispensing tool. The idea of being inoculated with industrial strength glue was not appealing to the operator. And so the “industrial blunt” dispensing needle was created. The other hazard was the potential of dropping those glass syringe barrels on the floor of the factory. (Yes, they all began as glass.) However, plastics were introduced quickly thereafter.
The glass tube was replaced with plastic. The hypodermic needle was replaced with a blunt. But something needed to be done about that human hand. Some operators were really good at it. Every time they pushed the plunger, the exact same amount of glue came out of needle, in exactly the right place. But when that person was out for a day, production went down by 50 percent and rejects were up. Repetitive motion stress syndrome and carpal tunnel syndrome also reared their heads.
Compressed air was then introduced, and it made a significant difference. Now the plunger could be driven down by air pressure and the hand could relax. The new challenge was to create a device to control the air pressure accurately enough to produce consistent deposits of material at the end of the needle. The theory of these controllers was that a timed blast of controlled pressure air was delivered to the syringe barrel, forcing the stopper down. Time/pressure dispensing worked well for materials that were not too thick and not too compressible, and for deposit sizes that were not too small.
When the time/pressure controller attempts to force a heavy material from a syringe barrel, very high air pressure must be used. In some cases, the syringe barrel just simply is not robust enough to withstand that amount of pressure, causing it to fail, usually catastrophically. Strike One.
Some fluid materials are quite compressible. Sufficient air pressure is applied to the material barrel to compress the material and then displaces some amount out the end of the barrel. Then, you would like it to stop. However, there remains a certain quantity of stored energy that wants to release. If the energy is not released, the fluid continues to drool out the barrel. Controllers then added drawback (or suck back). At the end of the dispensing shot cycle, the controller returns to delivering a vacuum to the barrel. This addition helped, but also negatively impacted accuracy. Strike Two.
The consistency of time/pressure dispensing was also affected by two inescapable factors. As the amount of material in the barrel decreases, there was less mass that the air pressure acts on. As the amount of material in the barrel decreases, there is a greater volume of air required to fill the syringe barrel and cause pressure on the stopper. Strike Three.
Next came valves. The first dispensing valves were on/off devices. Consistent air pressure is maintained on the material reservoir and the on/off valve allows material to flow. This is still time pressure, except now the pressure is constant on the reservoir and timing is applied to how long the valve is in the open state. Although improved, accuracy is still impacted by a changing amount of material in the reservoir. The other problem with valves is that really heavy materials require high fluid pressures on the reservoir to make them move. Thus, a very robust valve, one with a sturdy seal, is required to hold back the high fluid pressure.
In the mid 1980s, the need-to-needle dispense solder paste increased. Multi-level substrates and populated boards were some of the factors that prevented the use of screen printing. Additionally, there was a demand for smaller solder dots. Attempts to use valves were unsuccessful as the on/off action of the valve caused the solder balls to get crushed and accuracy was not good enough. After a considerable number of failures and testing, it was discovered that the needle tube needed to be roughly seven times the internal diameter of the largest particle in the paste.
The challenge of smaller dots caused time/pressure practitioners to increase the pressure on the material until it was being hammered by high pressure to dispense each dot. This produced the unwelcome side effect of driving all the metal to the bottom of the syringe, separating it from the flux. This led to the conclusion that the correct solution for dispensing paste was a pump not a valve. The theory was that material should be moved to the pumping chamber under low pressure and then the pump would increase the pressure on the material to drive it out of the right size needle tube.
All dispensing to this point had been some form of “extrusion” dispensing. This can also be characterized as “analog” dispensing. In analog dispensing, a flow is initiated, allowed to continue until the right amount of material has been extruded, and then the flow is terminated. The alternative is “digital dispensing.” In digital dispensing, a very small increment of fluid is dispensed, and then that dispensing action is repeated until the right amount of material has been dispensed.
The evolution of dispensing has enabled some of our newest and most appealing technologies. The path from “cave painting” to “digital dispensing” has been stimulated by the ever-changing demands and requirements of new electronic devices and new fluid material developments. I can't wait to see what's next.
Gary helmers, vice president, may be contacted at Creative Automation Company, 11641 Pendleton St., Sun Valley, CA 91352; (818) 767-6220; Article originally published in Solid State Technology Magazine, July, 2004.
Article originally published in Solid State Technology Magazine, July, 2004.