Browse Prior Art Database

A NOVEL TEST METHOD TO REDUCE RETEST RATE AND INCREASE FIRST PASS YIELD ON ATE

IP.com Disclosure Number: IPCOM000230740D
Publication Date: 2013-Sep-09
Document File: 4 page(s) / 125K

Publishing Venue

The IP.com Prior Art Database

Abstract

As communication era is moving to IoT, data transfer rates are becoming very high as are SoC communication interfaces. Testing such high data rate interfaces on ATE (automatic test equipment) requires specially designed hardware and instruments, which requires special attention during high volume ATE testing. ATE test engineers are facing this challenge day to day as the design margins to the spec limits become tighter as the technology shrinks the gate lengths of the transistors. Any minor flaw in testing setup or handling process can cause the DUT(device under test) to fail, which causes a yield loss on the perfectly good die causing test overkill situation and directly impacting the gross margin of the product. This paper discusses a method for addressing a bad contact/setup problem during ATE testing to overcome such problems in high volume production line.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 44% of the total text.

A NOVEL TEST METHOD TO REDUCE RETEST RATE

AND

INCREASE FIRST PASS YIELD ON ATE

ABSTRACT

As communication era is moving to IoT, data transfer rates are becoming very high as are SoC communication interfaces. Testing such high data rate interfaces on ATE (automatic test equipment) requires specially designed hardware and instruments, which requires special attention during high volume ATE testing.  ATE test engineers are facing this challenge day to day as the design margins to the spec limits become tighter as the technology shrinks the gate lengths of the transistors.  Any minor flaw in testing setup or handling process can cause the DUT(device under test) to fail, which causes a yield loss on the perfectly good die causing test overkill situation and directly impacting the gross margin of the product. This paper discusses a method for addressing a bad contact/setup problem during ATE testing to overcome such problems in high volume production line.

BACKGROUND

The manufacturing costs of modern complex SOC’s are shooting high and maintaining good lot yield in production line is a big task for product and test engineers.  To achieve their goals, engineers have a list of protocols to follow.  One protocol is to retest failing DUTs that are suspected of having failed due to a set up issue.  This is to regain the yield on some marginal fails due to test setup. This recovery on the retest can cause test time overhead and productivity loss for the production environment, which directly impacts the test cost and reduces the gross margin on the product.  Effort should be made to minimize and eliminate such retests.  The test method proposed here addresses this key issue in high volume manufacturing of modern day SOCs.

PROBLEM STATEMENT

During SOC testing on ATE, certain population of the DUT always fail while sitting in the socket for specific DC parametric tests mainly Vol/Voh type.  Re-execution of the test flow multiple times with the same setup doesn’t help in recovering the yield.  However, when the DUT is picked up and re-seated/re-plunged with handler in the socket and re-running the test flow, it passes the test and becomes a yield recovery.

Thus, ATE testing of silicon is impacted with 2 major issues:

1.       Re-test causes large test time and productivity loss.

2.       Re-testing good die in the production increases test cost and reduces gross margin on the product.

 

ROOT CAUSE

The first pass low yield in the above situation is understood to be based on the three issues.

1)       Low design and process margin to the test spec for the design.

2)      Above combined with increased parasitic components due to POGO pin wear off, socket/load board environment, and

3)      Delta (Machine to machine variability) in mechanical Z-pressure of handler plunger for each DUT insertion into the test socket.

TEST METHOD DEVELOPMENT

Raw data was collected over large sample size spanning multiple production lot material and over 16,000 measurement database was created to...