1. Field of the Invention
The present invention relates to a comparator with a latching function that determines a level of a differential signal.
2. Description of the Related Art
Differential signals are used for transmitting a signal between semiconductor devices at a high speed. A test apparatus for testing a semiconductor device with a differential interface is provided with a comparator (timing comparator) that receives a differential signal outputted from the semiconductor device to be tested (hereinafter, referred to as a DUT: Device Under Test) and that performs a comparison/determination of a value of the signal at a certain timing and latches a comparison/detei iination result. The test apparatus determines whether the DUT is good or not, specifies a defective portion, or evaluates its performance, based on an output signal from the timing comparator.
The data transmission rate between semiconductor devices keeps on increasing in recent years, and the response speed needed for a test apparatus is increasing in accordance with it. Accordingly, a timing comparator is needed to have a higher-speed response with higher gain than before.
A comparator with a latching function is disclosed in FIG. 1 of Patent Document 1. The comparator comprises a first-stage differential amplifier and a latch circuit latching an output from the amplifier. With the structure, there is a problem that the range of an input voltage may be narrower as compared with that of a power supply voltage of a differential amplifier. When using a Rail-to-Rail amplifier perfoi ming in the full range between the power supply voltage and the ground voltage to deal with the problem, an operation speed is decreased in exchange for that. In addition, when a potential difference between the two input signals is small, there is a possibility that a circuit is unstable to cause an oscillation. When hysteresis is supplied thereto to deal with that, a dead-band occurs in exchange for that, causing the sensitivity to be decreased.
(Patent Document 1) U.S. Pat. No. 6,211,705B1, Description
(Patent Document 2) U.S. Pat. No. 4,558,241, Description
(Patent Document 3) U.S. Pat. No. 4,608,670, Description
(Patent Document 4) U.S. Pat. No. 4,461,965, Description
(Patent Document 5) U.S. Pat. No. 4,485,317, Description
The present invention has been made under these circumstances, and one of the exemplary purposes of an embodiment of the invention is to provide a comparator with a latching function that has a high-speed response.
An embodiment of the present invention relates to a comparator that receives a pair of differential signals consisting of a positive differential signal and a negative differential signal, and that determines an amplitude level of the pair of differential signals at a predetermined timing and latches a determination result. The comparator with a latching function comprises: a positive input terminal to which the positive differential signal is inputted; a negative input terminal to which the negative differential signal is inputted; a first fixed voltage terminal to which a first fixed voltage is supplied; a second fixed voltage terminal to which a second fixed voltage is supplied; a positive line in which the positive differential signal is propagated; a negative line in which the negative differential signal is propagated; and a comparison amplification unit that compares a level of the signal in the positive line with that of the signal in the negative line and latches a comparison result, wherein the comparison amplification unit includes: a first inverter, an input terminal of which is connected to the positive line, an output terminal of which is connected to the negative line, and a power supply terminal on one side of which is connected to the first fixed voltage terminal; a second inverter, an input terminal of which is connected to the negative line, an output terminal of which is connected to the positive line, and a power supply terminal on one side of which is connected to the first fixed voltage terminal; and a first activation switch that can selectively switch between the state where the comparison amplification unit is inactivated by outputting the first fixed voltage to the other power supply terminals of the first and second inverters that are connected in common and the state where the comparison amplification unit is activated by outputting the second fixed voltage. The comparator outputs a signal corresponding to at least one of the signal in the positive line and the signal in the negative line at a timing after the comparison amplification unit is activated.
According to the embodiment, a high-speed response can be realized by using a dynamic comparator as a comparison amplification unit.
Another embodiment of the present invention relates to a test apparatus provided with the aforementioned comparator with a latching function. The comparator with a latching function receives a differential signal outputted from a device under test, and determines a value of the signal at a specified timing.
It is to be noted that any arbitrary combination or rearrangement of the above-described structural components and so forth is effective as and encompassed by the present embodiments. Moreover, this summary of the invention does not necessarily describe all necessary features so that the invention may also be a sub-combination of these described features.
Embodiments will now be described, by way of example only, with reference to the accompanying drawings which are meant to be exemplary, not limiting, and wherein like elements are numbered alike in several Figures, in which:
The invention will now be described based on preferred embodiments which do not intend to limit the scope of the present invention but exemplify the invention. All of the features and the combinations thereof described in the embodiment are not necessarily essential to the invention.
Herein, “the state where a member A is connected to a member B” includes not only the state where the member A is physically and directly connected to the member B but also the state where the member A is indirectly connected to the member B via another member that does not affect electrically the connection state between them. Likewise, “the state where a member C is provided between a member A and a member B” includes not only the state where the member A and the member C, or the member B and the member C, are connected directly, but also the state where they are connected indirectly via another member that does not affect electrically the connection state between them
An embodiment described below relates to Automatic Test Equipment (ATE) for testing a device under test (DUT) with a differential interface.
The ATE 100 comprises a pattern generator (PG) 1, a timing generator (TG) 2, a waveform shaper (FC: Format Controller) 3, a driver DR, a comparator (also referred to as a timing comparator) TC, and a determination unit 5. The driver DR and the comparator TC are collectively referred to as pin electronics 4. It is noted that the structure of the ATE 100 is merely an example, and architectures different therefrom may be adopted.
The pattern generator 1 generates a timing-set signal (hereinafter, referred to as a “TS signal”) and supplies it to the timing generator 2. The timing generator 2 generates a periodic clock CKp and a delay clock CKd based on the timing data specified by the TS signal, and supplies the periodic clock CKp to the pattern generator 1 and the delay clock CKd to the waveform shaper 3, respectively. The pattern generator 1 generates addresses ADRS indicating each of the blocks, which are a plurality of memory areas the DUT 102 has, and a plurality of test pattern data items Dt that are to be written in each of the plurality of blocks, and supplies them to the waveform shaper 3.
The waveform shaper 3 generates a test pattern signal St corresponding to the test pattern data Dt generated by the pattern generator 1, based on the delay clock CKd supplied by the timing generator 2. The waveform shaper 3 supplies the addresses ADRS, which are supplied by the pattern generator 1, and the test pattern signal St thus generated to the DUT 102 via a write driver DR.
The pattern generator 1 generates in advance expectation data Dexp, data to be outputted by the DUT 102 in accordance with the addresses ADRS and the test pattern signal St, and supplies the data to the determination unit 5.
The comparator TC receives from the DUT 102 a differential signal Do corresponding to the addresses ADRS and determines a level of the signal at any predetermined timing such that the comparator TC latches a determination result and outputs it to the determination unit 5 in the subsequent stage. The determination unit 5 compares a level of the data Do read by the DUT 102 with that of the expectation data Dexp supplied by the pattern generator 1, such that it is determined whether the DUT 102 is good.
The comparator TC receives differential signals Do consisting of a negative differential signal IN and a positive differential signal IP, and determines an amplitude level of the differential signals Do at a timing set by a control signal (φ3), and latches a signal indicating a determination result.
The comparator TC comprises an input hold unit 10, a comparison amplification unit 20, a reset unit 30, an output unit 40, and capacitors of a first capacitor C1 to a fourth capacitor C4.
A positive differential signal IP is inputted to the positive input terminal TP and a negative differential signal IN is inputted to the negative input terminal TN. The power supply terminal 50 (first fixed voltage terminal) is supplied with a power supply voltage Vdd as a first fixed voltage, and a ground terminal 52 (second fixed voltage terminal) is supplied with a ground voltage VGND as a second fixed voltage.
The positive differential signal IP (signal HP held by the input hold unit 10) is propagated through the positive line LP. The negative differential signal IN (signal HN held by the input hold unit 10) is propagated through the negative line LN.
The comparison amplification unit 20 includes a first inverter 22, a second inverter 24, and an activation switch 26. The comparison amplification unit 20 is structured with a form referred as a dynamic comparator or sense latch.
The first inverter 22, the second inverter 24, and the activation switch 26 are CMOS inverters. An input terminal of the first inverter 22 is connected to the positive line LP, and an output terminal thereof to the negative line LN. A power supply terminal (source of a PMOS) on one side of the first inverter 22 is connected to the power supply terminal 50 (first fixed voltage terminal).
An input terminal of the second inverter 24 is connected to the negative line LN, and an output terminal thereof to the positive line LP. A power supply terminal (source of the PMOS) on one side of the second inverter 24 is connected to the power supply terminal (first fixed voltage terminal) 50. The other power supply terminal (source of a NMOS) is connected in common to the other power supply terminal (source of the NMOS) of the first inverter 22. That is, the first inverter 22 and the second inverter 24 are cross-coupled.
The activation switch 26 applies the power supply voltage (first fixed voltage) Vdd or the ground voltage (second fixed voltage) VGND to the other power supply terminals (sources of the NMOS) of the first and second inverters 22 and 24 that are connected in common. A third control signal φ3 is inputted to the input terminal of the activation switch 26.
While the third control signal φ3 is being negated (low level), an output voltage from the activation switch 26 is equal to the first fixed voltage (power supply voltage Vdd), allowing the first and second inverters 22 and 24 to be inactivated. When the third control signal φ3 is asserted (high level), an output voltage from the activation switch 26 is equal to the second fixed voltage (ground voltage VGND), allowing the first and second inverters 22 and 24 to be activated.
The first capacitor C1 and the second capacitor C2 are coupled to the positive line LP and the negative line LN, respectively. The first capacitor C1 and the second capacitor C2 may be formed explicitly by using an MIM (Metal Insulator Metal) capacitance; however, when the capacitance values thereof are proper, they may also be formed by using the parasitic capacitance (gate capacitance of transistor and wiring capacitance). In addition, the first capacitor C1 is provided between the positive line LP and the grounding, and the second capacitor C2 between the negative line LN and the grounding; however, the positive line LP and the negative line LN may also be coupled by a single capacitor.
The input hold unit 10 is provided in the stage preceding the comparison amplification unit 20. The input hold unit 10 has a function of electrically separating the input terminal of the comparison amplification unit 20, the positive input terminal TP, and the negative input terminal TN. The input hold terminal 10 holds the differential signals IP/IN at a timing specified by the first control signal φ1. And then, the input hold unit 10 outputs the differential signals HN/HP thus held to the comparison amplification unit 20 in the subsequent stage.
The input hold unit 10 includes a first positive switch SW1p, a second positive switch SW2p, a first negative switch SW1n, a second negative switch SW2n, a third capacitor C3, and a fourth capacitor C4.
The first positive switch SW1p and the second positive switch SW2p are provided in series between the positive input terminal TP and the positive line LP. The third capacitor C3 couples a path between the first positive switch SW1p and the second positive switch SW2p and the ground. The first positive switch SW1p is turned on when the first control signal φ1 is asserted (high level), and the second positive switch SW2p is turned on when the second control signal φ2 is asserted (high level). In the state where the first positive switch SW1p is on and the second positive switch SW2p is off, a potential of the third capacitor C3 (positive hold signal HP) follows the positive differential signal IP. When the control signal φ1 is negated (low level) at a certain timing, the first positive switch SW1p is turned off and the positive differential signal IP at the timing is held by the third capacitor C3. When the second control signal φ2 is asserted and the second positive switch SW2p is turned on, the positive hold signal I-IP thus held is supplied to the positive line LP.
The first negative switch SW1n, the second negative switch SW2n and the fourth capacitor C4 are also connected in the same way to hold the negative differential signal IN.
In the same way as the first capacitor C1 and the second capacitor C2, the third capacitor C3 and the fourth capacitor C4 may also be farmed by using an MIM capacitance or the parasitic capacitance. Instead of providing the third capacitor C3 and the fourth capacitor C4, the line in which the positive hold signal HP is generated and the line in which the negative hold signal HN is generated may be coupled by a single capacitor.
When the circuit in the stage preceding the positive input terminal TP and the negative input terminal TN has a property by which a signal to be inputted to the comparison amplification unit 20 is switched off, in other words, when a signal in the positive line LP or the negative line LN does not affect the circuit in the preceding stage, the input hold unit 10 itself or some switches within the input hold unit 10 can be omitted.
The reset unit 30 is provided to initialize the potential of the positive line LP (positive output signal OP) and that of the negative line LN (negative output signal ON) to the first fixed voltage (power supply voltage Vdd). Specifically, the reset unit 30 includes a first reset switch SWr1 and a second reset switch SWr2. The first reset switch SWr1 is provided between the positive line LP and the power supply terminal 50 (first fixed voltage terminal). The second reset switch SWr2 is provided between the negative line LN and the power supply terminal 50. The first reset switch SWr1 and the second reset switch SWr2 are turned on when the fourth control signal φ4 is asserted (high level), and turned off when the fourth control signal is negated (low level). When the first reset switch SWr1 and the second reset switch SWr2 are turned on, the potentials of the positive line LP and the negative line LN are immediately reset to the first fixed voltage (power supply voltage Vdd).
Even when the first reset switch SWr1 and the second reset switch SWr2 are not turned on, the potentials of the positive line LP and the negative line LN are, although slightly, pulled toward the power supply voltage Vdd by a transistor included in the comparison amplification unit 20. Accordingly, if the comparator TC operates sufficiently at a low speed, the reset unit 30 can be omitted.
The output unit 40 includes a comparator element 42 and a latch circuit 44. The comparator 42 compares the positive output signal OP with the negative output signal ON and converts the result to a single-end signal (SO). The latch circuit 44 latches, when the fifth control signal φ5 is asserted (high level), the value of the output signal SO. The latched signal LP is outputted to a logic block in the subsequent stage. The output unit 40 only has to convert the SO signal to that with a signal form suitable for the circuit in the subsequent stage and to retime the signal; hence the circuit form thereof is not limited. In addition, because the output unit 40 is a component appendant to the comparator TC, the unit may be omitted.
The structure of the comparator TC in
Before the time t1, because the first control signal φ1 is at the high level and the second control signal φ2 is at the low level, the input hold unit 10 performs in a tracking mode and the hold signals HP/HN follow the input signals IP/IN. Because the third control signal φ3 is at the low level during this period, the comparison amplification unit 20 is inactivated. Further, because the fourth control signal φ4 is at the high level, the potentials of the positive line LP and the negative line LN are fixed to the power supply voltage Vdd by the reset unit 30.
When the first control signal φ1 makes a transition to the low level at the time t1, the values of the hold signals HP/HN are held to those at the time t1. At the time, the fourth control signal φ4 is at the low level, allowing a reset by the reset unit 30 to be released.
When the control signal φ2 make a transition to the high level at the subsequent time t2, the second positive switch SW2p is turned on, and charges are transferred between the first and third capacitors C1 and C3, allowing the potentials OP and HP to be averaged. The same is true for the negative line.
Subsequently, when the third control signal φ3 makes a transition to the high level at the time t3, an output signal from the activation switch 26 has a voltage equal to the ground voltage VGND, allowing the comparison amplification unit 20 to be activated. When the comparison amplification unit 20 is activated, a potential difference between the positive output signal OP and the negative output signal ON expands to reach its saturation.
Subsequently, when the second control signal φ2 makes a transition to the low level at the time t4, the third and fourth capacitors C3 and C4 are separated from the comparison amplification unit 20.
Subsequently, when the first control signal φ1 makes a transition to the high level at the time t5, the input hold unit 10 is brought back to the tracking mode.
Subsequently, when the fifth control signal φ5 makes a transition to the high level at the time t6, a signal corresponding to the positive output signal OP and the negative output signal ON, that is, an output signal from the comparator 42 is latched. The fifth control signal φ5 is later to be negated at an appropriate timing.
When the third control signal φ3 makes a transition to the low level at the time t7, the comparison amplification unit 20 is inactivated. During the period between the times t7 and t8 (hatched period), the potential of the positive line LP (negative line LN) is slightly pulled toward the power supply voltage Vdd by the comparison amplification unit 20 in the inactivated state.
Subsequently, when the fourth control signal φ4 makes a transition to the high level at the time t8, the positive output signal OP and the negative output signal ON are fixed to the power supply voltage Vdd by the reset unit 30 (reset performance). This state is the same as that before the time t1. The comparator TC subsequently repeats the same performance.
According to the comparator TC, the input signals IP/IN can be sampled at a high speed at a desired timing. Specifically, the comparator can follow a signal with a 100-ps cycle time (10 GHz). In addition, because the comparison amplification unit 20 is provided with a dynamic comparator, the comparator has a very high sensitivity (high gain) and is tolerant of self-oscillation. Furthermore, the comparator can receive the differential signals IP/IN swinging from the ground voltage VGND to the power supply voltage Vdd (Rail-to-rail input).
Subsequently, some variations of the comparator derived from the structure in
Specifically, one power supply terminal (source of NMOS) of a first inverter 22a is connected to a ground terminal 52 (first fixed voltage terminal). One power supply terminal (source of NMOS) of a second inverter 24a is connected to the ground terminal (first fixed voltage) 52, and the other power supply terminal (source of PMOS) is connected in common to the other power supply terminal (source of PMOS) of the first inverter 22a.
An activation switch 26 applies a power supply voltage (second fixed voltage) Vdd or the ground voltage (first fixed voltage) VGND to the other power supply terminals (sources of PMOS) of the first and second inverters 22a and 24a, which are connected in common). A third control signal φ3# (# indicates a logic inversion) is inputted to the input terminal of the activation switch 26a.
While the third control signal φ3# is being negated (high level), an output voltage from the activation switch 26a is equal to the first fixed voltage (ground voltage VGND), allowing the first and second inverters 22a and 24a to be inactivated. When the third control signal φ3# is asserted (low level), an output voltage from the activation switch 26a is equal to the second fixed voltage (power supply voltage Vdd), allowing the first and second inverters 22a and 24a to be activated.
The reset unit 30a is provided to initialize potentials of the positive line LP (positive output signal OP) and the negative line LN (negative output signal ON) to the first fixed voltage (ground voltage VGND). Specifically, the reset unit 30 includes a first reset switch SWr1a provided between the positive line LP and the ground terminal 52 (first fixed voltage terminal) and a second reset switch SWr2a provided between the negative line LN and the ground terminal 52 (first fixed voltage terminal).
In other words, the first fixed voltage is equal to the power supply voltage Vdd and the second fixed voltage is equal to the ground voltage VGVD in the structure of
The reset unit 30b is provided to initialize the potential of the positive line LP (positive output signal OP) and the potential of the negative line LN (negative output signal ON) to a predetermined voltage Vc. The predetermined voltage Vc may be arbitrarily set, and may be set to, for example, the midpoint voltage of the power supply voltage vdd and the ground voltage VGND or the common voltage of the differential signal IP/IN. Hereinafter, Vc is referred to as a center voltage. The reset unit 30b includes a voltage supply 31 generating the center voltage Vc and reset switches SWr1b and SWr2b.
The state before the time t2 is the same as that in
In the time chart illustrated in
Subsequently, when the first control signal φ1 makes a transition to the high level at the time t4, the input hold unit 10 is brought back to the tracking state.
Subsequently, when the third control signal φ3 makes a transition to the high level at the time t5, the comparison amplification unit 20 is activated. The subsequent performance is the same as that of the time chart in
As stated above, even if transition timings of some control signals are changed, the comparator TC can sample the differential signals IP/In. It is noted that a person skilled in the art can understand that the time chart illustrated in
The first inverter 22d comprises transistors PMOS, NMOS1 and NMOS2. The first inverter 22d performs an inverse amplification on the signal inputted to the gate of the transistor NMOS1 and outputs it from the drain thereof (drain of the PMOS). The gates of the transistors PMOS and NMOS2 are connected in common and biased by the output signal OP from the second inverter 24d. The second inverter 24d is structured in the same way as the first inverter 22d and connected symmetrically to the first inverter 22d.
Subsequently, the performance of the comparator TCd in
In the comparator TCd in
A performance speed of the stand-alone comparison amplification unit 20d in
When there is a mismatch between the threshold voltages of the respective transistors NMOS1 in the inverters 22d and 24d that are used for the input unit of the comparison amplification unit 20d, the mismatch is, as it is, added to an input offset voltage of the comparator. However, when using the comparator TCd for a semiconductor test apparatus, the offset can be cancelled by initializing the apparatus, resulting in a less major issue.
The following variations are possible for the comparator TCd in
1. a structure in which, in the same way as
2. a structure in which, in the same way as
While the preferred embodiments of the present invention have been described using specific terms, such description is for illustrative purposes only, and it is to be understood that changes and variations may be made without departing from the spirit or scope of the appended claims.