Wire Bonding Tools & Questions
Q: Our company manufactures high power RF amplifiers and we have entered the high power LED market. Recently we have experienced failures because the epoxy we are using cannot dissipate the heat generated by small footprint, high powered devices. Are you aware of an epoxy that can transfer the extreme amount of heat generated by high power devices?
A: Iíve been working with Epoxy Technology, a global manufacturer of epoxy adhesives, which offers a complete line of electrically and thermally conductive material for high power applications. In order to generate the highest thermal conductivity (ThK), silver is the filler of choice. Silver filled epoxy provides high Thk values between 3 Ė 20 W/mK which makes it an ideal interface for removing heat from high power devices. I visited a company which uses Epoxy Technologyís silver filled epoxy to dissipate heat from lens concentrated photovoltaic die arranged in massive solar panels. Thatís about as hot as it can get!
Check out their Web site at www.epotek.com or give them a call at 978 667 3805. They are fantastic people to work with and they will bend over backwards to help you.
Q: We are having a difficult time finding wire pull hooks and shear tools for some of our older equipment. Do you know of a source?
A: Hooks and Shear Tools, Inc. specializes in standard and custom hooks and shear tools for most of the pull testers and shear testers in use today. They will design and manufacture a tool to fit your application and equipment. They also supply pick up tools for die pickup machines and die sorters. I was amazed at the line of products available for wire bonding, die bonding and wafer sawing. Iím sure they will have what you need. For more information send an e-mail to firstname.lastname@example.org or call them at (530) 620 7132 and ask for Dave Hollis.
Q: We have used every possible epoxy dispense pattern we can think of and we still experience voids in coverage. Can you help?
A: Using a piece of glass cut to the exact size of the die is usually the best method for determining the ideal dispense pattern to minimize or eliminate epoxy voids. Granted, glass may not have the exact surface characteristics of the die you are placing, but it will give you a good idea of how the dispense pattern will look after placement. If you are using automated equipment a pattern can be etched into or inked on the glass for pick up using pattern recognition. The glass can be removed, cleaned and used again. Hopefully, It will eliminate or minimize the need for expensive X-raying of the placed part.
Q: The programmable lighting on our 2460-V does not display the programmed value when we load the program. It also will not go to 255 when turned fully clockwise on either the off axis or through-the-lens settings. Is there a problem with the module?
A: Itís possible the module is defective, but more than likely it just needs to be set up correctly. Contact Palomar Technologies for the programmable lighting set up procedure.
Q: It appears our host computer motherboard is intermittent. There is no way to tell if itís the processor or the motherboard itself. Are replacements for either part available?
A: You might be able to find a replacement for the processor, but the motherboards are obsolete and impossible to locate. Palomar Technologies does provide passive slot boards and the half height CPU board which will replace the motherboard in the 2460-V or 2470-V machines.
Q: The 5 volt output of our Xentec linear main power supply in our 2460-II is much less than 5 volts at the printed circuit boards in the bay. How do you adjust it?
A: Follow the white wire marked with number 4 or the two black wires marked with number 8 from the barrier strip of the power supply. There should be three pots on the PCB at the end of the wires. Using a mirror will aid in finding the correct pot. The 20 turn voltage adjust pot is the one you want. Adjust the voltage with the meter on the ends of C21 on the I/O board to exactly 5.00 volts. The voltage at the barrier strip will be higher to compensate for the resistance between the barrier strip and the I/O board.
Q: In one of your Bonding Notes you mentioned something about a burnout free EFO solenoid. What was the part number?
A: Replace with a 24 volt solenoid (standard on the 2460-V). The Palomar Technologies part number is WD-15467. A new bracket is required since the solenoid body is longer. The bracket is Palomar Technologies part number WD-8707B25-001. Hopefully, Palomar still stocks them.
Q: At our company almost every stitch bond uses table scrub. I feel this is overuse of the feature; when should it be used?
A: I always emphasize in my classes that table scrub should be used only as a last resort. Table scrub was intended to enhance stitch bond adhesion when normal ultrasonics are inadequate. Table scrub may be used to compensate for the ultrasonics attenuated by a resonant pin, contaminated thick film or spongy surfaces such as a soft dielectric or flexible lead. The overuse of table scrub results in inconsistent pull tests, an inordinate number of flame-offs, deformed stitch bonds and increased capillary wear and plugging. Remember, ultrasonic energy is applied during the entire table scrub cycle. Table scrub should never be used on ball bonds.
Q: We get a large number of stitch bonds with a torn wire sticking up the center of the stitch crescent. Whatís happening?
A: The adhesion between the deformed wire and the surface gold in the stitch bond is greater than the breaking strength of the wire. When the wire breaks above the stitch rather than at it, a stretched broken wire in the center of the crescent occurs. Try decreasing the force and ultrasonics incrementally until a stitch lift occurs, then increase the force and ultrasonics slightly and you should have the correct parameters. I realize this is counter intuitive but it is the only way to correct this situation. This usually happens on easily bonded surfaces like thin film. The other possibility is low clamping force causing the wire to slip in the wire clamp before it is broken, although I have seen this phenomenon with the correct clamping force with 1 mil and larger wire. Try decreasing the force and ultrasonics before making any clamp tension adjustments. If the condition continues, the wire clamping force is probably too low.
Q: The bond head on our machine slams into the heater stage when we turn the machine off. If we donít remember to move the stage out of the way, the capillary and/or part may be damaged. How do we correct this problem?
A: Ideally, itís always best to move the table to its home position before turning the machine off. This will guarantee the part and the capillary will not be damaged if the bond head falls to its lower limit position when the machine is turned off or a power failure occurs. As an added benefit the machine will spend less time initializing the next time itís initialized. Check the bond head for backlash caused by loose engagement of the rack and pinion gear teeth. Adjust the rack and pinion until all backlash is removed with the holding current on. With the holding current off the bond head should move smoothly through its entire travel without binding. At this point, if the bond head falls to itís lower limit simply increase the engagement of the rack and pinion slightly. Always check for backlash and smooth movement of the bond head after the final adjustment.