Time Interval Meter

The present invention relates generally to devices for measuring elapsed time, and in particular to a time-interval meter for measuring extremely short time intervals.

Conventional time measurement circuits typically employ direct counting techniques or ratios of counts in frequency /period measurements. One such technique is to gate a digital counter on upon some event, and stop the counter upon the occurrence of a second event. The counter counts clock pulses between the two events, and consequently the measured time interval has an error of .+-. one count. For long time intervals measured by counting high-speed clock signals, the one-count error may be inconsequential. However, for short time intervals, for example, in the sub-microsecond range, the count error becomes significant. In the prior art, this problem is overcome by employing an extremely high-speed clock and associated highspeed counter circuits, with attendant added complexity and high cost.

In accordance with the present invention, a simple and inexpensive time interval meter is provided for measuring extremely short time intervals, such as the time difference between a signal-related trigger and a next successive sampling clock edge in a digital oscilloscope. A timing circuit operable within a predetermined timing window includes a capacitor which is chargeable at two different predetermined rates, with the slower rate establishing a predetermined maximum time interval. In the preferred embodiment, the charging rates are precisely scaled to a ratio of 100:1.

At the interval start, for example, upon receipt of a trigger signal, the capacitor begins to charge at the faster rate. At the interval stop, for example, upon receipt of a sampling clock edge, the charging rate is switched, and the capacitor continues to charge at the slower rate. Also at the interval stop, a counter is activated to count clock pulses during the slow-rate portion of the capacitor-charging cycle. When the capacitor charges to a predetermined voltage level which corresponds to the maximum time interval, the counter is stopped. The count thus obtained at the slower rate is scaled by the fast rate-slow rate ratio, e.g., divided by 100 in the preferred embodiment, to provide an actual measured time interval which is subsequently subtracted from the predetermined maximum time interval to yield the desired time interval measurement between the two events.

It is therefore one object of the present invention to provide a novel time interval meter.

It is another object to provide a time interval meter for measuring short, i.e., sub-microsecond, time interval measurements using lower speed clock and counter circuits.

It is a further object to provide a time interval measurement circuit which employs time scaling during the operation thereof to increase measurement accuracy.

Other objects and attainments of the present invention will become apparent to those skilled in the art upon a reading of the following detailed description when taken in conjunction with the drawings.

FIG. 1 is a schematic diagram of a time interval meter in accordance with the present invention; and

FIG. 2 is a timing diagram showing the time interval measurement.

The preferred embodiment of the present invention is a time interval meter for measuring elapsed time between a signal-related trigger and a next successive sampling clock edge in a digital oscilloscope in order to correct jitter resulting from .+-. one-half sample period error. Referring now to FIG. 1, a pair of edge-triggered D flip-flops 10 and 12 control the operation of the time interval meter in response to a trigger signal applied to an input terminal 16 and a sample clock signal applied to an input terminal 18. The circuit operation will be discussed in detail later in connection with FIG. 2.

A pair of current sources 20 and 22 provide constant charging current for a timing capacitor 24. Current source 20 is connected between a suitable source of positive supply voltage, such as +12 volts, and the emitters of an emitter-coupled pair of transistors 28 and 30. Current source 22 is connected between the +12-volt supply and the emitters of a second emitter coupled pair of transistors 32 and 34. These emitter-coupled transistors provide current switching, as will be described later, and permit only one of the two current sources 20 and 22 to be coupled to the timing capacitor 24 at any given time. The bases of transistors 30 and 32 are connected together to a suitable level of reference voltage, while the collectors thereof are connected together and to one side of the capacitor 24, the other side of which is connected to ground. The collectors of transistors 28 and 34 are both connected to ground, and the bases thereof are connected to the Q and Q outputs respectively of flip-flop 12.

A comparator 40 has its inverting (-) input connected to the away-from-ground side of capacitor 24, and its non-inverting (+) input connected to a precise reference voltage. The output of comparator 40 is connected to one input of an AND gate 42. A clock signal is applied via a terminal 44 to a second input of AND gate 42. The output of AND gate 42 is connected to the toggle input of a binary counter 48. The Q output of flip-flop 12 is connected to the clear input of counter 48. The count data that is produced by counter 48 is send to a processing circuit, such as a microprocessor (.mu.P) 50.

The timing capacitor 24 is resettable by a transistor 54, the collector and emitter of which are connected across the capacitor. The base of transistor 54 is coupled to the Q output of flip-flop 10 via a parallel combination of resistor 56 and speed-up capacitor 58. A resistor 60 is connected between the base of transistor 54 and a suitable source of negative voltage, e.g., -12 volts, to hold the transistor in a normally cut off mode. Transistor 54, while shown as a bipolar transistor, could be a field-effect transistor as well.

The circuit operates as follows: Initially, flip-flop 10 is cleared, so that its Q output is low and its Q output high. Transistor 54 is turned on to saturation, holding timing capacitor 24 completely discharged. Flip-flop 12 is cleared by the low Q output of flip-flop 10, so that its Q output is low and its Q output high. Transistors 30 and 34 are turned on, while transistors 28 and 32 are off, so that current from current source 20 flows to ground through transistors 30 and 54, and the current from current source 22 flows to ground through transistor 34. With the top of timing capacitor 24 virtually grounded, the output of comparator 40 is high, allowing clock signals to pass through AND gate 42 to the counter, which is held in a cleared condition by the high Q output of flip-flop 12 and thus produces no count output. This completes the initial conditions for the time interval circuit.

Upon receipt of a trigger signal at terminal 16, the outputs of flip-flop 10 switch states, releasing flip-flop 12 and transistor 54. Transistor 54 switches off, permitting all of the current from current source 20 to flow into the timing capacitor 24. The charging voltage as a function of time within a predetermined timing window for capacitor 24 is shown in FIG. 2, with the triggering point 80 occurring at time t.sub.o. If allowed to charge at this rate completely to the 2-volt switching level of comparator 40, the 2-volt point would be reached within a predetermined time interval t.sub.T, which is selectable from t.sub.T =200 nanoseconds, 100 nanoseconds, or 40 nanoseconds in the preferred embodiment. These time intervals t.sub.T were chosen for the preferred embodiment to facilitate measurement of the time difference between a trigger signal and a next successive edge of a sample clock at different sweep rates wherein the sampling clock rates are 5, 10, and 25 megahertz respectively.

At some point within the time interval t.sub.T, then, the sample clock edge arrives, causing the Q and Q outputs of flip-flop 12 to switch states, turning transistors 28 and 32 on, and 30 and 34 off, and removing the clear signal from counter 48. At this point, shown by the breakpoint 82 in FIG. 2, current from current source 22 flows into the capacitor while the current from current source 20 flows to ground through now-conducting transistor 28. In the preferred embodiment, current source 20 provides 10 milliamperes (mA) of current, while current source 22 provides 100 microamperes (.mu.A) of current, so that a precise 100:1 scaling ratio exists between the two. Therefore, after receipt of the sample clock edge, the timing capacitor charges toward the 2-volt limit at a one hundredth slower rate, during which time the counter, no longer being held clear, counts the 10-megahertz clock signals arriving via AND gate 42. The slower charge rate is shown as the dashed line 84 in FIG. 2, and it should be noted that the ratio of the slopes is approximately 10:1 to facilitate illustration of the concept. The particular ratio actually chosen depends upon the situation and the measurement accuracy desired.

When the timing capacitor 24 charges to the 2-volt limit, comparator 40 switches and the output thereof goes low, causing AND gate 42 to block the counter 48 from the clock signals. The contents of counter 48 at this point, which have been counted over an expanded t.sub.2 time interval, represent the actual time interval t.sub.2 because of the precise scaling. That is, each count of 100 nanoseconds of the slow charging current is equivalent to one nanosecond at the fast charging current. The microprocessor 50 subtracts the t.sub.2 interval from the predetermined time interval t.sub.T to yield the time interval t.sub.1 between the two events of trigger signal and sampling clock edge.

Circuit imperfections may be corrected by the microprocessor 50 as well. For example, in the saturated condition of transistor 54, the capacitor 24 may actually have a couple of tenths of a volt thereacross, requiring an adjustment of the comparator reference voltage to provide a precise 2-volt time interval window. The microprocessor may correct for this offset by keeping track of minimum and maximum counts received on repetitive cycles and adjust the raw data.

The time interval meter is cleared and reset to the initial conditions upon application of an initialize signal to the clear input of flip-flop 10. The initialize signal may be generated in a number of ways after the count signal is converted to a measurement, and is generated by the microprocessor 50 in this embodiment.

Although the present invention has been described in connection with a particular embodiment thereof, it is to be understood that additional embodiments, modifications, and applications thereof which will be obvious to those skilled in the art are included within the spirit and scope of the invention.