Heated Stage for Thermosonic Wedge Bonding

Introduction: Heated Stage for Thermosonic Wedge Bonding

About: I'm an electrical engineer. By day I design chips, by night I like making stuff that is unnecessarily complex.

First off, I'd like to thank Jeff Keyzer for his donation of a PID controller to this project and his exepriences building his hotplate for surface mount soldering. Take a look at his project in his MightyOhm blog at http://mightyohm.com/blog/2009/01/diy-pid-controlled-soldering-hotplate/

Thermosonic compression bonding is used to make electrical connections between a bare die chip and other components including other chips, printed circuit susbtrates, packaging frames, etc. A combination of heat, ultrasonic vibration, and compression form a metallic bond between a wire and a pad on the die.

I'm interested in making use of MMIC (Monolithic Microwave Integrated Circuits) chips in the > 30 GHz range, parts which for the most part do not come in solderable packages. So I'm left with wirebonding, something usually relegated to industrial and academic institutions.

Piecing together a wirebonding setup can be costly, as most wirebonders new are >$10,000. I found a great deal on a manual WestBond bonder for $50, and a stereo microscope to go with it. Among the things missing though was a heated stage to hold and heat the workpiece.

Commercial heated stages are costly, and for something like $50 I put together my own, one which I believe is as good if not better than commercially available units.

Here are some features that I wanted:

- Heat to 150C and keep it there with reasonable regulation
- Have a smooth plastic base to allow easy maneuvering under the bonder arm
- Adjustable height for use with machined microwave housings, or bare substrates (need about 1/2 inch of adjustment range)

Step 1: Parts List

Here's what I used to make my heated stage:


Block of aluminum ~3x2x1 inches
Plate of aluminum (for baseplate) 1/4 inch thick
Standoff holders/adjusters 1/2 inch aluminum rod stock
Standoffs - 1/4 inch diameter FEP plastic (a PTFE-like material)
Cartridge heater (1/2 inch diameter, 400 watts (which is far too much, going to switch to a 150 watt heater)
Computer power cord (used for supplying power to the heater)
20 gauge Teflon insulated wire (for grounding the heated portion)
Nylon 'expando' sheath to bundle the wires together
Miscellaneous stainless steel screws and set screws (4-40 size)
UHMW plastic (static dissipating type) 1/8 inch thick.


Omega CN9000 PID controller (Thanks Jeff!)
Rocker switch
IEC power socket
Aluminum housing
Rubber grommet for wires
hookup wire and shrink tubing for insulation

Most of the parts here can be purchased from McMaster Carr and Digikey. Getting a good deal on a PID controller is key, as they are ~$200 new.

Step 2: Machining the Block

The block forms the heated surface and provides thermal mass for the thermal feedback system. I chose to go with a 3x2x1 inch block because its big enough for most of my work, and small enough to fit easily on the wire bonder's work surface.

I machined the block square on a milling machine and bored a 1/2 inch hole in one edge, and 4 holes for standoffs on the corners. I used a 0.257 inch reamer for finishing the standoff holes which improves the inner surface finish.

Another feature I decided to include was a trough around the perimeter of the block in order to use tiny C-clamps to hold down a workpiece to the stage. This was accomplished again on the milling machine using a 1/8 inch end mill.

Step 3: The Baseplate

The baseplate is just a piece of 1/4 inch aluminum plate, with a sliding surface made from UHMW Polyester static dissipating plastic. I made this on the milling machine. The center section was hogged out to allow room for screw heads for the standoffs, cable strain relief, and so on.

Step 4: Adjustable Standoffs

Being able to adjust the height of the work stage is critical. The wire bonder I'm using has no height adjustment and you get very little vertical movement. Additionally its better to work in middle of the range of the arm travel, rather than at its extremes.

So to achieve vertical adjustment I decided to use high temperature plastic rods held in place by tubing that has a slot on the side for using a set screw.

The 'tubing' was actually made by drilling out the center of 1/2 inch solid rod, and machine one side flat. This gives a very nice surface for a screw to tighten against. I have just over 1/2 inch of travel with this arrangement.

Step 5: Other Stuff for the Heated Stage

There are a few other details that I added to the heated stage. First, I added a grounding wire to the stage since the plastic insulates the heated stage from the grounded and static dissipating slider plate. I also added a cable strain relief block, and thermocouple attachment.

The attachment method for the thermocouple is now different from what is shown. The picture shows an eyelet holding the thermocouple in place, but it proved to be far too insulating. I have since replaced it with a small aluminum block with a small recess machined out. This provides a much better thermal contact to the heated block, dramatically improving the response time.

Lastly, I made a couple C-clamps for holding down modules. These can fit anywhere on the perimeter of the heated block.

Step 6: Control Box

The control box is very simple. It has a PID controller, switch, and power receptacle. There's also a bolt going through the back for grounding.

Step 7: Improvements

I initially chose a 400 watt heater for this project, and it turns out that it is far too much. The rate at which the heater heats the block is so fast that the PID controller can't shut it off fast enough. There are settings in the PID controller to allow for different rates, but this is so fast that it would be hard to regulate well. I'm going to replace the heater with a 100 or 150 watt version and that ought to make things much more stable.

Again thanks to Jeff for his insight and PID controller.
Take a look at his excellent surface mount soldering hot plate:

Be the First to Share


    • Make It Bridge

      Make It Bridge
    • Game Design: Student Design Challenge

      Game Design: Student Design Challenge
    • Big and Small Contest

      Big and Small Contest



    3 years ago

    Have you measured the amplitude of the sine going to the bonder transducer? I wanted to compare with mine. It's hard to find data on these. I was surprised that it wasn't as high as expected: about 10vpp at 64khz


    14 years ago on Introduction

    I'm interested to know what you're building that utilises MMICs Also, kudos for such a nice build.


    Reply 14 years ago on Introduction

    Millimeterwave ham radios. I've been building microwave ham radios for years, and now that I'm working at 47 GHz and above, wire bonding bare MMIC dies becomes necessary.