This post authored by John Coonrod originally appeared on the ROG Blog hosted by Microwave Journal.

Wire bonds keep everything in place on a printed-circuit board (PCB). They are used to attach passive and active components as well as integrated circuit (ICs) to a circuit substrate, and even to connect one circuit substrate to another. Wire bonds can be formed with a variety of different wire bonding machines, including manual and automatic models. In all cases, the goal is to achieve a low-resistance connection with good mechanical integrity and high reliability. But this seemingly simple goal depends not only on the type of substrate material and its parameters but numerous wire-bonding parameters, including the temperature, time, and applied force when making a wire bond.

Forming Wire Bonds

Wire bonds can be formed by various methods, including ultrasonic bonders, in which the energy for the weld comes from ultrasonic force, and thermocompression bonders, in which thermal energy is applied to form a wire bond. In addition, a thermosonic bonder uses a combination of ultrasonic and thermal energy to form a wire bond. Types of wire bonds include ball bonds and wedge bonds. Interconnection bond wires are typically formed of gold (Au), copper (Cu), or aluminum (Al) wire.
Wire bonding machines used with high-frequency PCBs include ball bonders and wedge bonders. Ball bonders are faster, but wedge bonders tend to deliver higher-reliability bonds. Ribbon bonders are essentially wedge bonders in which flat ribbon wire is used instead of round wire to provide a large cross section at the heel of a bond and, presumably, a higher-reliability bond.

Establishing workable wire-bond parameters

Establishing workable wire-bond parameters depends not only on the type and thickness of the substrate material, but on the type of metal plating and plating thickness on the substrate and even the dimensions of the bonding pad. For example, softer circuit-board materials such as PTFE can present challenges for forming wire bonds because a soft material can absorb more energy than a harder material and deform during the wire-bonding process.

When evaluating the quality of a wire bond, electrical testing is conducted to establish that a low-resistance path has been formed, while pull tests are typically performed to determine the strength of a wire bond. For example, making wire bonds on substrates with smaller bond pads can be more difficult than with a substrate having larger bond pads, since the vertical pressure from a wire bonder is much greater on a small pad than on a larger pad. If a substrate’s bond pad is too small, the force of attaching a wire bond can deform the bond pad, or even force it beneath the surface of softer substrate materials. Deformation of the bonding pad and/or substrate material can also occur as a function of the bonding temperature, when temperatures higher than the glass transition temperature (Tg) of the substrate material are used. Bond-wire suppliers often provide recommendations for the maximum bond pad size (in mils) for a given type and diameter of their bond wire.

Rogers Study on Wire Bonds

Because the choice of substrate material plays a major role in the ultimate quality that can be achieved with a PCB’s wire bonds, we recently performed a study on wire bonds formed on high-frequency substrate materials. The results of this study helps material users better understand how different wire-bonding parameters are needed for different types of materials. Sample boards were manufactured with different Rogers’ materials as part of the study to better understand how such parameters as device finish, wire type, and wire diameter can impact the quality of a wire bond.

Two different wire bonders were employed in the study, an automated wire bonder and a manual wire bonder, both using thermosonic bonds with 1-mil gold bond wire. Substrates were plated with 50 microinches of nickel and 200 microinches of Type III grade A gold. FR-4 was used as the reference material in the study, which also included RO4000® hydrocarbon ceramic laminates and RT/duroid®5880 glass-microfiber-reinforced PTFE composite PCB material from Rogers. Low-cost RO4000 laminates are engineered to be processed like FR-4, while RT/duroid 5880 represents the challenge of forming wire bonds on a softer PCB material.

The study (available upon request from the author) establishes safe “starting points” for making low-resistance, reliable wire bonds on each of the materials. It details a number of different parameters for a wire bonder, including the temperature of the stage on which the substrate is mounted, the bonding power and force, and the time required to form the wire bond for each material.

For example, the reference material, FR-4, has the shortest processing time but the highest stage temperature (+130°C), highest applied force, and greatest amount of bonding power of the materials studied. The RT/duroid 5880 material, because it is a “soft” PTFE-based composite material, worked with the lowest stage temperature (+80°C), less bonding power, and considerably less bonding force. The study even cautions that a lower stage temperature may be required for PTFE-based materials, along with a stabilization period for the material to reach a level of thermal equilibrium once mounted on the wire-bonder stage.

The study points out that its results are to be taken as starting points for setting wire-bonding parameters. In handling PTFE-based materials, for example, more reliable bonds may come as a result of decreasing the bonding force while increasing the time required to form the wire bond. The type of plating used with soft PCB materials can also impact the reliability of the wire bonds and the PCB in general. The study offers tested starting points known to deliver good results in terms of wire bond electrical performance and reliability for these materials, and PCB users are invited to modify those bond-wire parameters in their quest for the perfect PCB wire bond.

Tagged with:  

62 Responses to Aiming For The Perfect Wire Bond

  1. Jeff Gibala says:

    I’m interested in seeing the study that was conducted. How can I get a copy?

  2. Rick Wylie says:

    I’d like to see this study as well. Thanks in advance.

  3. Daniel van Wyk says:

    Hi

    We would also like to receive a copy of the article, if possible.

    Thanks beforehand.
    Rgds

  4. sharilee says:

    Hi Daniel. It’s on its way to you via email. Let us know if you have any questions.

  5. Tash says:

    Hi, I’m interested in reading the full study, could I have a copy?

    Thanks

  6. sharilee says:

    Hi Tash. I just emailed the study to you. Let us know if you have any questions.

  7. Jan Nordbo says:

    Hi, I’m interested in reading the full study, could I have a copy?

    Thanks

  8. sharilee says:

    Hi Jan. Certainly. I’ll email that to you. Let us know if you have any questions.

  9. David Escudero says:

    Hi! Please, if I’m still on time, I would like to receive a copy of the study.

    Thanks

  10. Wei Zheng says:

    Hi
    It is a valuable study.

    Can I have a copy of your study.

    Thank you so much!

  11. Wolf Haslin says:

    Please provide a copy of this study. The RO8580 and RO5870 is of special interest
    Thanks
    Wolf

  12. sharilee says:

    Copies of the study have been emailed to everyone. Let us know if you have any questions.

  13. Theo Pantazopoulos says:

    May I have a copy as well please?

  14. Billy McDonald says:

    Hi
    I am interested in reading the full study, could I have a copy?

    Thanks

  15. Hello, I’d also be interested in reading the full study. Could you please send me a copy?
    Thanks in advance.

  16. Phil Harm says:

    Greetings! Please send me a copy of the full study. Thanks in advance!!

  17. Sofan Sofyan says:

    Hi sharilee,
    i’m interested to read the full study, kindly sent me the full copy,
    thanks a lot…

  18. Jay B says:

    Can I please have a copy of the study sent to my email. I am really interested on the outcome.

    Thanks,

    jay

  19. Burt J says:

    I also would like to see the results of this study. Can you please send me a copy?

    Many thanks.

  20. Steve Riesbeck says:

    Hi Sharilee, I’d love to see a copy of that report too. Duroids have always been a challenge to wire bond too. Thanks ! Steve

  21. Hamza says:

    Hey .. Can I get a copy of this study ??

  22. KS says:

    I would like to get a copy of this study too

  23. Simon says:

    Hello Sharile, Could I also get a copy of this study.
    Many thanks.

  24. Frank says:

    Can I have copy of the study please?
    Thank you!

  25. Ziha says:

    Hi Sharilee, i would love to include this study in my References for my final year project. Could i have a copy?
    Thank you 🙂

  26. Eric says:

    Hello Sharilee,

    may I have a copy of the study as well?

    Cheers

  27. lothar baier says:

    can you email a copy to lothar.baier@utd.edu ?

  28. Dhanraj Biradar says:

    Hi, I’m interested in reading the full study, could I have a copy.

  29. Greg W says:

    Hi — like others, I would like to have the study. Thank you!

  30. Elizabeth B says:

    I would like to read the full study. Thank You.

  31. Sean A says:

    Please could I get a copy of the study

  32. Derron says:

    I am interested with the study, may I ask for a copy.
    Please consider.

  33. Matt says:

    I would like to see the study on aiming for the perfect wire bond. Thank you!

  34. Paul Mellon says:

    May I have a copy of the study report?

Leave a Reply

Your email address will not be published. Required fields are marked *