This invention relates to integrated circuits and their testing, debugging and characterization. In particular, this invention relates to the testing and characterization of embedded nets in integrated circuits.
The testing of integrated circuits presents many challenges owing to the extremely small dimensions of the circuit elements and the nets connecting them. The problem is further aggravated by the fact that several of the internal nets and nodes of the circuit are embedded in inner layers of the interconnect. The transient response of internal circuit nets is an important parameter in analyzing the transient behaviour of the circuit. The measurement of the transient response requires the measurement of the voltage at the selected node at various instants of time in response to a step change in applied voltage at a selected point in the circuit.
At present, the voltage at any embedded net is probed using:
A methodology is therefore needed to probe embedded nets. Such a methodology would be immensely useful in designing/testing/debugging/characterization of a time dependent design. Such methodology will also be helpful in providing a correlation between simulation and Silicon, which will aid model development.
To address the above-discussed deficiencies of the prior art, an object of this invention is to provide a device and method for monitoring the transient behaviour of the internal nets of an integrated circuit.
To achieve the said objective, this invention provides an apparatus for enabling the on-chip analysis of the voltage and/or current transition behaviour of one or more embedded nets of an integrated circuit independently of the fabrication process, comprising:
The above apparatus further comprises an on-chip delay characterizer.
The Reference Step Generator is a programmable voltage/current source.
The Comparator is a differential amplifier.
The stepped delay generator is a balanced buffer chain.
The on-chip delay is characterized by a ring counter.
The present invention also provides a method for measuring the voltage and/or current transient response of an embedded net in an integrated circuit independently of the fabrication process comprising the steps of:
Before undertaking the DETAILED DESCRIPTION OF THE INVENTION below, it may be advantageous to set forth definitions of certain words and phrases used throughout this patent document: the terms “include” and “comprise,” as well as derivatives thereof, mean inclusion without limitation; the term “or,” is inclusive, meaning and/or; and the phrases “associated with” and “associated therewith,” as well as derivatives thereof, may mean to include, be included within, interconnect with, contain, be contained within, connect to or with, couple to or with, be communicable with, cooperate with, interleave, juxtapose, be proximate to, be bound to or with, have, have a property of, or the like. It should be noted that the functionality associated with any particular apparatus or controller may be centralized or distributed, whether locally or remotely. Definitions for certain words and phrases are provided throughout this patent document, those of ordinary skill in the art should understand that in many, if not most instances, such definitions apply to prior, as well as future uses of such defined words and phrases.
The invention will now be explained with reference to the accompanying drawings, in which like reference numerals represent like parts, in which:
As shown in
The Comparator produces the following outputs after the comparison at the output Q.
The reference signal generated by the RSG is applied to the SIGREF input of the Comparator. The latching signal produced by the SDG is received by the Comparator as COMPSIGNAL. The node under test whose electrical characteristics are to be derived is connected to the SIGCOMP input of the comparator.
The VSG now generates another voltage VREF2 and the process is repeated. The output comparator VC now goes “TRUE” at the time instant t5. Thus, the delay between the two signals is (t5–t2), and the voltage difference between the two reference voltages is VREF1−VREF2. Using the above two values, the voltage waveform can be determined quite accurately. If the VSG generates another voltage VREF3, the time taken to discharge between the voltages VREF2 and VREF3 can be determined. This logic can be extended until VREF8. Using the above information, the voltage waveform of the node in question can be plotted. If the voltage waveform in question is linear in nature, then the discharge rate can be determined simply by taking only two reference voltages (e.g. VREF1 and VREF2 in the example). The discharge rate would be given as (VREF1−VREF2)/t1TRUE−t2TRUE; where t1TRUE=the time to get the first TRUE output from the VC for VREF=VREF1, and t2TRUE=the time to get the first TRUE output from the VC for VREF=VREF2.
The time steps t1, t2 . . . can be calculated if the delay step is known between any two time steps. The SDG is externally programmable, thus giving the output in digital format. For example, if a two bit counter is used, and each count adds a step, then, if t1 corresponds to 01 and t2 corresponds to 11, it can be concluded that the delay between t1 and t2 is:
mod [[Decimal value (01)−Decimal Value (11)]*Δ]=mod [1−3]*Δ=4Δ.
Similarly, the VSG is also digitally controllable. If both VSG and SDG are implemented on-chip, the step characterization can be done on-chip as explained in the following sections. This makes the entire activity process independent.
A similar methodology can be used for plotting other electrical characteristics on-chip.
A preferred embodiment of the invention consists of these major blocks
Two conditions are defined.
The following table describes the operation of the latch amplifier in the two conditions.
The capacitive element at node 205 is connected as shown in the circuit. A reference voltage is connected to node 206. At the time of the latch on signal, if the capacitor voltage is greater than the constant reference voltage, then a POSC condition arises. If, on the other hand, the capacitor voltage is less than the constant voltage, then a NOSC condition arises.
Any other voltage comparator can also be used for the same functionality as above.
When signal CHAREN is low and signal CONTROL is such that the signal INPUT is passed through the SELECT LOGIC element, the circuit of
The delay step is given as:
Fosc=1/Tosc=1/(2*(N*DELAY+invDELAY+selogicDELAY))
Tosco=2*(N*DELAY+invDELAY+selogicDELAY)
DelayStep=Tosc/2.
There would be an error in the Delay step due to the delay of the invDELAY and selLOGIC DELAY. If the delay chain was large, the above error would be very small. The calculation of delay on chip would make the delay independent of the process change.
As shown in
By controlling the select pins of the mux, various voltages can be generated. Also, by proper designing, the resistance ratio can be made constant and process independent. This makes the voltage generation independent of the process parameters. Knowing the digital values of the select pins of the mux, the voltage generated by the VSG can be calculated.
The whole process of characterization can be automated by using the controller. The controller varies the COMPsignal signal until the NOSC condition is achieved. The controller then varies the reference signal. The whole process is repeated again. The outputs are in the form of the number of delay units used for delay insertion for the first case (N1) and the number of delay units used for delay insertion in the second case (N2). The delay between the two latch on signals which caused the circuit to enter the NOSC region can then be calculated from (N2−N1)*DelayStep. Also, if the reference signals are given externally, then the signal difference is known. If the reference is generated internally, then the reference signal levels will be available as analog or digital values corresponding to the two reference voltage levels. The controller probes the various nodes in the design and stores the digital results for the respective nodes. The results can then be used to probe the characteristics of the embedded nets.
Advantages of the present invention may include:
It is intended that the present invention encompass such changes and modifications as fall within the scope of the appended claims.
Number | Date | Country | Kind |
---|---|---|---|
1629/DEL/2003 | Dec 2003 | IN | national |
Number | Name | Date | Kind |
---|---|---|---|
4161029 | Frye et al. | Jul 1979 | A |
4607214 | Welzhofer | Aug 1986 | A |
5173906 | Dreibelbis et al. | Dec 1992 | A |
5528137 | Riggio, Jr. | Jun 1996 | A |
6182257 | Gillingham | Jan 2001 | B1 |
6369601 | Ishigaki | Apr 2002 | B1 |
6424583 | Sung et al. | Jul 2002 | B1 |
6978411 | Huang et al. | Dec 2005 | B2 |
Number | Date | Country | |
---|---|---|---|
20050174102 A1 | Aug 2005 | US |