Electricity: measuring and testing – Fault detecting in electric circuits and of electric components – Of individual circuit component or element
Reexamination Certificate
2000-06-22
2002-06-04
Karlsen, Ernest (Department: 2858)
Electricity: measuring and testing
Fault detecting in electric circuits and of electric components
Of individual circuit component or element
C324S765010, C324S1540PB
Reexamination Certificate
active
06400164
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to electromagnetic interference measurements (EMI) of digital systems and, more particularly, to EMI characterization of integrated circuits (ICs) at multiple clock speeds.
BACKGROUND OF THE INVENTION
Electromagnetic radiation emitted from a digital system is mainly at the fundamental frequency of its switching operation and that frequency's harmonics; for systems with multiple clocks, multiple fundamental frequencies and harmonics will be present. At other frequencies, system radiation levels are typically undetectable.
International as well as national regulatory standards exist which specify allowable levels of EMI emissions from unintentional radiators. The main purpose of these standards is to protect the radio frequency spectrum for equipment licensed to operate at appropriate frequencies. Because the regulated frequencies reach into the gigahertz range and the allowable signal levels are typically in the microvolt range, the test set-up when measuring EMI often influences the test results. Therefore, in addition to allowable signal levels, these regulations also specify measuring methods in an attempt to standardize the compliance tests and improve their repeatability. These factors increase both the length and cost of typical EMI tests.
Typically, EMI regulations directly apply only to full systems, such as a personal computer, and not directly to system components, such as video cards or microprocessors. An accepted industry practice, however, to test the electromagnetic compatibility (EMC) of system components is to use the “substitution method”. In this method, a compliant system is used, and an original component is replaced with a new component. If the system is still compliant, then the new component is determined to be electromagnetically compatible.
Because the microprocessor is one of the major contributors to system level EMI, EMC assessment of microprocessors is a concern of CPU manufacturers. This assessment encompasses two facets: verification of compliance to EMC standards and component characterization. When performing the substitution method to verify EMC compliance, multiple tests of multiple components on multiple systems are needed to provide any degree of confidence in the test results. Component characterization, the other facet of EMC assessment, identifies the effects that design choices have on EMI emissions. For purposes of characterization, a high number of package design options (e.g., package layer stack-up, chip-cap configuration, etc.) are evaluated and compared to arrive at valid manufacturing decisions.
One common feature of both facets is the large number of tests which need to be performed in order to properly assess a microprocessor's EMC. The far-field tests typically used for radiated emission measurements are both time consuming and expensive and the component industry needs an alternative method to perform EMC assessment. The Society of Automotive Engineers in developing their standard SAE J1752-1, entitled “EMC Measurement Procedures for Integrated Circuits”, has investigated using near-field EMI measurement methods for ICs; but these methods typically involve specially manufactured test boards attached to a modified Transverse Electromagnetic Mode (TEM) cell and introduce measurement errors when operated above one gigahertz. Other near-field measurement methods are also known but have the drawbacks that repeatable measurements are difficult to obtain and measurements above 1 GHz are outside the range of current near-field probes.
Regardless of the method in which EMI data is acquired, the current method of testing EMI performance of different packages involves operating the package at a single operating speed, or core speed, and measuring the resulting EMI emissions at a number of individual frequencies (i.e. spot frequencies). The measured data is linearly interpolated to fill in the gaps between the spot frequencies. When the relative EMI performance of two or more packages is desired, the data collected for each package is compared. This method of testing and comparing package EMI data has a number of shortcomings; in particular, the EMI performance of a package operating at one core speed is not necessarily a good indicator of its performance at other core speeds. A measure of EMI performance of an integrated circuit which considers performance at a number of different core speeds is necessary.
SUMMARY OF THE INVENTION
The present invention addresses the need for an improved EMI testing method which not only considers EMI performance of an integrated circuit operated at a single speed but also considers the EMI performance at other operating speeds.
This, as well as other needs, are met by the present invention which provides a novel method of testing integrated circuits at a number of clock speeds and then forming an EMI characterization of the IC based on its EMI performance at the different clock speeds. The resulting EMI characterization allows more meaningful comparisons between an IC's different EMI performances than the previous method of collecting EMI emission data while operating an IC at only a single operating speed.
The needs are also met by embodiments of the invention which measure EMI emissions of an IC operated at a first core speed, measure EMI emission of the IC operated at a second core speed, and then use the EMI emissions data from these two speeds to form a characterization of the IC's EMI performance.
The foregoing features, as well as other aspects and advantages, of the present invention will become more apparent from the following detailed description, claims and drawings.
REFERENCES:
patent: RE35644 (1997-10-01), Rogers
patent: 6201403 (2001-03-01), Rollin et al.
patent: 6242925 (2001-06-01), Schutten et al.
patent: 11-262330 (1999-09-01), None
Advanced Micro Devices , Inc.
Karlsen Ernest
Tang Minh N.
LandOfFree
Method for comparing package EMI performance at multiple... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for comparing package EMI performance at multiple..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for comparing package EMI performance at multiple... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2967898