Get our free email newsletter

Small Form Factor CDM Testing, Part 2

Air Discharge Options

This column is the second in a three-part series on testing small form factor products for CDM. Part 1 highlighted the issues of CDM testing with the current field-induced CDM (FICDM) testers. The main problems are the pogo pin size vs package or ball bump size and that small form factor products may have very low withstand voltages, where the FICDM testers are known to be unreliable. In Part 2, the first set of possible solutions is presented. These solutions are air discharge techniques just like FICDM. Part 3 will focus on contact-first methods to address these same issues.

Air Discharge Options for Bare Die and Interface Die Testing

CDM Tester with Small Probe to Contact Bare Die Bumps or Pads

A wafer CDM tester is shown in Figure 1. This tester was built by modifying a standard FICDM tester [1]. The basic configuration matches the hardware platform prescribed by ANSI/ESDA/JEDEC JS-002 [2]. To make contact with the small pads of bare die or wafer, the tester has a 7-µm radius needle connected to a ground plane with 1-ohm resistance. One end of the 1-ohm resistor is connected to the inner conductor of the coaxial cable, while the other end is connected to the ground plane and to the shield conductor of the coaxial cable. The coaxial cable is connected to an oscilloscope. There is a charge plate that can hold a 300-mm wafer.

Figure 1: Wafer CDM tester – Block diagram
Figure 1: Wafer CDM tester – Block diagram

A 1.8V power terminal that had experienced frequent failures at the assembly house. This modified tester was used to test the product in three different form factors: first on wafer of 300 mm at full thickness, second bare die which was 8 x 8 mm and a thickness of 400 µm, and finally in a BGA package at 30 x 30 mm in size. The results are shown in Table 1.

- Partner Content -

A Dash of Maxwell’s: A Maxwell’s Equations Primer

Through mathematical precision, Glen Dash offers a comprehensive exploration of Maxwell's Equations, revealing the intricate mechanisms by which electric and magnetic fields interact to explain electromagnetic radiation and fundamentally shape our technological understanding of modern electronics and communications.

Bare die

Package

Wafer

1.8V power terminal

1500V

- From Our Sponsors -

1300V

1250V

Table 1: Product test results

For the 1.8V power terminal, the destruction voltage for the bare die and wafer tests was, respectively, 1500V and 1250V, which easily met the electrostatic control standard of 200V. This shows that electrostatic controls at the assembly house were inadequate and consistent with the level of failures seen. This method does provide a solution to gather on-wafer CDM data, but it does not address the lower voltage stability issues that FICDM testers are known to have. Odds are this still needs to be addressed.

Contact First CDM (CF-CDM)

The contact first with air discharge approach to CDM was developed to address two aspects of field-induced air discharge CDM testing, one of which is especially important for fine pitch packaged devices, and when stressing bare die or at the wafer level.

  • A sharp tip degrades the arc’s properties in air discharge. This method removes that degradation because the discharge is in the chamber, not at the sharp-tip pogo pin.
  • Air discharge quality and repeatability depend on air quality, such as humidity.

An entirely new ground plane and test head were developed and are shown in Figure 2. The ground plane includes a contact pin, which is physically but not electrically connected to the ground plane. The ground plane is lowered in operation until the contact pin touches the DUT. Since the contact pin is floating and has low capacitance, it does not ground the DUT. Next, the rest of the test head continues its downward motion until an air discharge occurs between the pogo pin and the top of the contact pin, very similar to the discharge of a standard FICDM. The discharge pogo pin is connected to the 1-ohm resistor and current measurement electronics. The top of the contact pin and the discharge pogo pin are enclosed in a small chamber flooded with dry nitrogen. This method provides a spark geometry independent of contact geometry. The discharge also occurs in a stable, dry nitrogen environment, eliminating the issue of humidity [3].

Figure 2: Contact first CDM test head, (a) with head separated from DUT and (b) shown in the stress position
Figure 2: Contact first CDM test head, (a) with head separated from DUT and (b) shown in the stress position

Measurement results with the contact first air discharge CDM method were found to meet the requirements of ANSI/ESDA/JEDEC JS002 regarding waveform parameters such as peak height, rise time, peak width, and undershoot, but unfortunately, only marginally better regarding variability from arc to arc. The real strength of the new system has turned out to be its ability to reliably test packages with very fine pitch connections. The method allows using a sharp contact pin to touch the DUT before any discharge.

Summary

The standardized FICDM tester with a pogo pin is unsuitable for bare-die or wafer-level (including 2.5D/3D microbump D2D interfaces) CDM testing. The two air discharge options in the literature today approach the problem differently. One system uses an FICDM tester and modifies it so that it can handle wafers to be charged. It also has a modified pogo pin to contact the pads/bumps on the part. Limited data is available that proves its usefulness as an option for small form factor parts. The second system can contact the part with a secondary small pogo pin. Air discharge takes place in a chamber above the small pogo pin as the primary pogo pin will drop after the smaller pogo pin has made contact. In Part 3, another set of systems will be presented that do not have an air discharge. The associated pros and cons will also be discussed.

Acknowledgements

This three-part publication is part of a summary of a technical report published by the ESD Association. The authors of that technical report would like to be acknowledged here: Robert Ashton (retired), David Epps (AMD), Jared Floyd (ESDEMC), Wei Huang (ESDEMC), David Klien (pSemi), Tom Meuse (Thermo Fisher), Kathleen Muhonen (Qorvo), Friedrich sur Nieden (Infineon), Paul Phillips (Phasix), Michael Reardon (ESDEMC), Masnori Sawada (Hanwa), Jasmine Shen (ESDEMC), Marko Simicic (IMEC), Heinrich Wolf (Fraunhofer)

References

  1. M. Sawada et al., “Study of FI-CDM Discharge Waveforms,” EOS/ESD Symposium 2010.
  2. ANSI/ESDA/JEDEC JS-002, Charged Device Model (CDM) – Device Level, EOS/ESD Association.
  3. Grund et al., “A New CDM Discharge Head for Increased Repeatability and Testing Small Pitch Packages”, EOS/ESD Symposium 2018.
  4. Kathleen Muhonen, “Small Form Factor CDM Testing, Part 1: Problems with FICDM Testing for Small Form Factor and Interface Die,” In Compliance Magazine, January 2026.

Related Articles

Digital Sponsors

Become a Sponsor

Discover new products, review technical whitepapers, read the latest compliance news, and check out trending engineering news.

Get our email updates

What's New

- From Our Sponsors -

Don't Let Regulations

Derail Your Designs

Get free access to:

Close the CTA
  • Expert analysis of emerging standards
  • EMC and product safety technical guidance
  • Real-world compliance solutions

Trusted by 30,000+ engineering professionals

Sign up for the In Compliance Email Newsletter

Discover new products, review technical whitepapers, read the latest compliance news, and trending engineering news.

Close the CTA