(Full 7 Pages Here)

Introduction to IEEE P1687/IJTAG

CJ Clark, Bill Tuthill,
Intellitech Corp. 69 Venture Dr, Dover, NH,

Abstract – This paper introduces the basics of the IEEE P1687

Standard for Access and Control of Instrumentation Embedded
within a Semiconductor Device. The standard introduces a new
language called Instrument Connectivity Language for
describing how to connect internal IEEE 1149.1 Test Data
Registers inside a device. IEEE P1687 leverages the IEEE
1149.1 TAP to access these Test Data Registers via Procedural
Description Language extensions to the open source Tool
Command Language. This paper represents the views and
experience only of the authors and does not represent a position
of the P1687 or 1149.1 working group. The standard is still in
development as of May 2010 and the information here may be
Figure 1. Life-Cycle JTAG re-use
subject to change.
Great progress has been made over five years of
development, however we will not speculate on when the
Keywords: IJTAG, P1687, TAP, JTAG, 1149.1, Instrument, DFT, standard will go to ballot or when it will pass balloting. A lot
DFx, BIST of hard work from many smart engineers is required to
develop a standard. One fellow already said it was „done‟ but
I. INTRODUCTION that was nearly four years ago now. 4 We think CEOs should
not be in the position of predicting when engineers will finish
The IEEE P1687 working group is developing a standard the work in an IEEE standard.
for connecting and communicating with on-chip DFx IEEE P1687 is poised to be a very useful standard. We
structures via the 1149.1 TAP. Using the TAP for on-chip have read white-papers and attended conference presentations
access to DFT was presented throughout the 1990s by Sun however that have hyped the proposed standard like a blender
Microsystems, Motorola, Xilinx, LogicVision and others. on a late night infomercial. One explanation for the hype is
(Despite this fact, the press and board test companies still that those positioned as „boundary-scan‟ players, have found
continue refer to 1149.1 as “boundary-scan” even though that that boundary-scan tool sales have run out of steam. This may
is only one half of the standard). Early presentations show also explain the non-1149.x numbering chosen for the
how the TAP could be used for access to on-chip health standard; a „new‟ capability would allow new feature pricing
monitors, BIST, BISRs and other test IP in 2003.1 The P1687 and potentially not be included in the usual 1149.x annual
effort started in 2004 and the IEEE PAR (Project maintenance contract.
Authorization Request) was approved in 2005. On chip DFx We are actively working in the WG and developing solutions
structures and infrastructure IP such as BIST and scan are in P1687, but our view is pragmatic in the rate of adoption.
called „instrumentation‟ by the working group 2. Standards and tool vendors don‟t drive the industry, the
The purpose of the standard is to provide a description of industry uses useful standards that uniquely solve a problem,
how to connect the on-chip instruments and describe a and are as easy as or easier to implement than another method.
language for communicating with the instruments via 1149.1 Here are some standards related to test and their status:
Test Data Registers. The goal is greater test re-use throughout
the SoC life cycle using common languages of ICL/PDL/TCL IEEE 1149.1 - Standard - Widely adopted
and operation via the 1149.1 TAP (see Figure 1). The IEEE 1149.4 - Standard – little/no adoption
working group has had many language proposals for IEEE 1149.5 - Withdrawn - System Level Standard
describing the JTAG communication to the instrument IEEE 1149.6 - Standard – Good acceptance for AC-coupled
including C API libraries and in-house developed languages. IEEE 1149.7 - New Standard - little adoption – Royalties5
Several members were in favor of TCL and Intellitech‟s IEEE 1532 - Admin Withdrawn - bit/pof and SVF work fine
presentation in late 2008 probably was the tipping point in IEEE 1500 -Standard – little adoption, NXP, Virage, Synopsys
getting agreement on TCL and the use model. 3
Completion of a standard does not guarantee wide-spread
adoption, especially when competing technologies exist.
Business models are probably the biggest driver. IEEE 1500
1
http://www.intellitech.com/pdf/internal-jtag.htm
2 4
Eklow, Bill, Bennets, Ben, “New Techniques for Accessing Nelson, Rick, “SJTAG, IJTAG claims premature”, ITC06
Embedded Instrumentation: IEEE P1687 (IJTAG)”, ETS 2006 Interview. http://www.tmworld.com/article/320833-
Guest_commentary_IJTAG_SJTAG_claims_premature.php
3 5
http://www.intellitech.com/powerpoint/IJTAG-presentation- http://standards.ieee.org/db/patents/pat1149.html
by-Intellitech.pdf
IJTAG/P1687 Tutorial NATW 2010

http://www.intellitech.com IJTAG P1687 1687