The disclosure is related to time interpolator circuits.
A time interpolator circuit increases the resolution and accuracy of time measurements beyond the limits of digital circuits. Time interpolators are commonly used in frequency counters, which are instruments that measure the number of cycles of a repetitive signal per second. In a typical reciprocal frequency counter, a digital circuit counts the number of electronic clock pulses that occur per cycle of a signal to be measured. The frequency is then proportional to the reciprocal of this number. In actual practice most frequency counters count the number of clock pulses that occur during a large number of signal cycles. Thus the counter may start counting clock pulses at one cycle of the signal and stop millions of signal cycles later.
As an example, suppose that a frequency counter has a clock that runs at 10 MHz and the signal to be measured has a frequency of roughly 1 GHz. The counter begins counting clock pulses at one signal cycle and stops 100 million signal cycles later. Suppose that 999,437 clock pulses are counted between the first and 100 millionth signal cycles. This result means that the frequency of the nominally 1 GHz signal is actually about 1.000563 GHz (100 million signal cycles divided by 0.0999437 seconds). It may not be exactly 1.000563 GHz, however, because the time between the first signal cycle and the first clock pulse, and the time between the last signal cycle and the last clock pulse, haven't been measured. A time interpolator is a circuit that accounts for these fractional times to improve measurement accuracy.
An early time interpolator circuit example is described in “Electronic interpolating counter for the time interval and frequency measurement” by Bagley and Brooksby (U.S. Pat. No. 3,133,189), and numerous variations and improvements have followed. Many interpolators rely on the charging characteristics of a capacitor connected to a current source. The voltage across such a capacitor is:
where V is the voltage, C is the capacitance and I is the current flowing into the capacitor. If I is constant, as is the case with a good quality current source, then:
Thus, the voltage across the capacitor is directly proportional to the time during which current is allowed to flow into it. Furthermore, this voltage can be measured quite accurately and precisely with an analog-to-digital converter.
Despite the long history of interpolator circuits, room for improvements exists. Thus what is needed is a simple, accurate interpolator circuit appropriate for modern frequency counters and similar devices.
The time interpolator circuit described below provides good performance in frequency counters and other instruments. It is simple, inexpensive and easily interfaced to digital circuits.
Time is converted to voltage in the circuit of
Buffer amplifier 140 and its feedback network resistors 145, 150 are not required if ADC 155 is able to directly measure the voltage across capacitor 115 without discharging it. In other words, the voltage measurement should not affect the capacitor voltage by more than the precision needed in a particular interpolation application. Whether or not a buffer amplifier is necessary depends on the input characteristics of the ADC.
The operation of the circuit of
The capacitor voltage ramp shows the voltage across capacitor 115. The capacitor charges linearly with time when the current gate signal is V1 and stops charging when the current gate signal is zero. The voltage drop across diode 110 prevents capacitor 115 from charging when switch 125 is turned on. In other words the voltage drop across the diode is greater than the resistance of switch 125 in its conducting state multiplied by the current supplied by current source 105.
The reset signal shows when capacitor 115 is discharging. When the reset signal is VR circuit node 165 is connected to ground through switch 130 and capacitor 115 discharges. When the reset signal is zero, switch 130 is off (i.e. non-conducting) and the capacitor may charge depending on the state of switch 125.
The circuit of
At time t1 in
where I is the current supplied by current source 105 and C is the capacitance of capacitor 115.
This voltage remains constant and is available to be measured by buffer amplifier 140 and ADC 155 from time t2 until time t3. At time t3 the reset signal changes from zero to VR and switch 130 turns on, connecting circuit node 165 to ground and discharging capacitor 115. The voltage across the capacitor rapidly decays to zero. Sometime before the next interpolation interval begins the reset signal changes back to zero at time t4 and the circuit is returned to the state it was in just before time t1. A new interpolation interval starts at time t5.
The time interpolator circuit of
In
Event counter 335 counts input events, e.g. pulses or cycles from input 305. Time counter counts clock events, e.g. pulses from an internal clock or reference clock 315. The counters send their results as digital data to microprocessor 345. Interpolator 100 is started and stopped by the count synchronization logic via switches 125 and 130 as described above. The interpolator also sends its digital results via output 157 to microprocessor 345 which carries out the necessary calculations to estimate the frequency of a signal at input 305. Interpolator 100 may interpolate between clock pulses at the beginning and end of a counting interval. The results of a frequency measurement may be displayed by display 350.
As another example, interpolator circuit 100 of
In one implementation the interpolator described above provides approximately ten picosecond repeatability and is ready for a new measurement in less than one microsecond. Low-voltage logic interfacing is convenient when switches 125 and 130 are implemented as FETs. Regardless of specific implementation the circuit provides a simple, precise and inexpensive time-to-voltage conversion capability.
The above description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Thus, the disclosure is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Number | Name | Date | Kind |
---|---|---|---|
3133189 | Bagley et al. | May 1964 | A |
3204180 | Bray et al. | Aug 1965 | A |
3983481 | Nutt et al. | Sep 1976 | A |
4433919 | Hoppe | Feb 1984 | A |
4439046 | Hoppe | Mar 1984 | A |
4620788 | Giger | Nov 1986 | A |
4719608 | Genat et al. | Jan 1988 | A |
4764694 | Winroth | Aug 1988 | A |
4772843 | Asaka et al. | Sep 1988 | A |
4870629 | Swerlein et al. | Sep 1989 | A |
4875201 | Dalzell | Oct 1989 | A |
5132558 | Rustici | Jul 1992 | A |
5191336 | Stephenson | Mar 1993 | A |
5199008 | Lockhart et al. | Mar 1993 | A |
5206889 | Unkrich | Apr 1993 | A |
5333162 | Condreva | Jul 1994 | A |
5359404 | Dunne | Oct 1994 | A |
5521696 | Dunne | May 1996 | A |
5631553 | Bose et al. | May 1997 | A |
5684760 | Hunter | Nov 1997 | A |
5703678 | Dunne | Dec 1997 | A |
5703838 | Gorbics et al. | Dec 1997 | A |
6137749 | Sumner | Oct 2000 | A |
6246737 | Kuglin | Jun 2001 | B1 |
6324125 | Frankowsky et al. | Nov 2001 | B1 |
6822485 | Kattan | Nov 2004 | B2 |
7202716 | Chao et al. | Apr 2007 | B1 |
20090009455 | Kimura | Jan 2009 | A1 |
Number | Date | Country | |
---|---|---|---|
20120281806 A1 | Nov 2012 | US |