The disclosure can be better understood with reference to the following drawings. The elements of the drawings are not necessarily to scale relative to each other, emphasis instead being placed upon clearly illustrating the principles of the disclosure. Furthermore, like reference numerals designate corresponding parts throughout the several views.
The present disclosure generally relates to mass spectrometers and methods for dynamically adjusting an effective sampling rate of a signal from an ion detector so that memory requirements can be reduced. A time-of-flight mass spectrometer in accordance with one exemplary embodiment of the present disclosure, for each mass scan, ionizes a mass specimen, and an ion detector provides an analog signal indicative of detected ion abundance as a function of time. The analog signal is sampled, and digitized samples from different mass scans are summed to define a resultant mass spectrum. The number of mass scans is selected to provide a desired statistical accuracy for the resultant mass spectrum.
During each mass scan, the effective sampling rate of the analog signal is changed. In one embodiment, the analog signal is effectively sampled at a relatively high rate at the beginning of the mass scan as compared to later in the mass scan. Thus, as the mass scan progresses, the number of digitized samples provided for summing per unit of time decreases thereby reducing the number of memory locations required to store the resultant mass spectrum as compared to an embodiment in which the analog signal is effectively sampled at the same high rate throughout the entire mass scan.
As noted above, the time required by an ion to traverse the distance between the electrode 12 and the detector 25 is a measure of the mass of the ion. This time is proportional to the value in address register 21 when the ion strikes the detector 25. Hence, memory 29 stores data that can be used to generate a graph of the number of ions with a given mass as a function of the mass. In other words, the data stored in memory 29 defines a mass spectrum of the mass specimen being analyzed.
Various devices, such as a Faraday cup, multichannel plate (MCP), electron multiplier (continuous structure as well as dynode structure), conversion dynode, Daly detector, and combinations thereof, may be used to implement the ion detector 25. The signal generated by the ion detector 25 depends on the number of ions striking the detector 25 during the clock cycle in question. In general, this number is relatively small, and hence the statistical accuracy of the measurements obtained in any single mass scan is usually insufficient. In addition, there is a significant amount of noise in the system. The noise is generated in the detector 25, analog path, and in the A/D converter 27.
To improve statistical accuracy, the data from a large number of mass scans are summed. At the beginning of the measurement process, the controller 15 stores zeros in all of the memory locations in memory 29 and initiates the first mass scan. When the first mass scan is completed, the controller 15 resets the address register 21 and initiates another mass scan by causing the pulse source 17 to pulse the electrode 12. The data from the second mass scan is added to that from the previous mass scan. This process is repeated until the desired statistical accuracy is obtained.
Thus, each memory address stores a sum of corresponding samples from different mass scans. As used herein, samples are “corresponding” if they are taken at the same time after the start of their respective mass scans. For example, a sample taken at a time, t1, after the start of a first mass scan corresponds to a sample taken at the same time, t1, after the start of another mass scan. The sum of corresponding samples from each of the mass scans is stored in one of the addresses in memory 29 and represents a data point of the resultant mass spectrum. Note that the start of a mass scan refers to the generation of the pulse that ionizes the mass specimen being analyzed by the mass scan.
Further, in embodiments for which the write address register 21 is incremented for each clock cycle, contiguous memory addresses store samples that are consecutive in terms of time. For example, assuming that the address register is four digits and is reset by the controller 15 at the beginning of a mass scan, as described above, the data value stored at address 0000 represents the sum of digital samples taken during the first clock cycle of each mass scan. The data value stored at address 0001 represents the sum of digital samples taken during the second clock cycle of each mass scan, and so on. Moreover, the foregoing is illustrated by
In this regard, address (n) of memory 29 stores the sum (samplen+1) of digital samples taken during the (n+1)th clock cycle after the start of each mass scan, where n is a positive integer less than the total number of addresses in memory 29. Address (n−1) of memory 29 stores the sum (samplen) of digital samples taken during the (n)th clock cycle after the start of each mass scan, and address (n−2) of memory 29 stores the sum (samplen−1) of digital samples taken during the (n−1)th clock cycle after the start of each mass scan. Further, address (n−3) of memory 29 stores the sum (samplen−2) of digital samples taken during the (n−2)th clock cycle after the start of each mass scan, and address (n−4) of memory 29 stores the sum (samplen−3) of digital samples taken during the (n−3)th clock cycle after the start of each mass scan.
Unfortunately, the amount of memory 29 required to store all of the data points of the resultant mass spectrum can be quite large thereby increasing the cost and complexity of the mass spectrometer 10. In this regard, the memory 29 has a number of addresses equal to or greater than the number of data points used to define the resultant mass spectrum. Most time-of-flight mass spectrometers sample the analog signal from the ion detector 25 at a very high rate to achieve a desired resolution and, therefore, create an extremely large number of data points.
As shown by
As described above with reference to
As in
Thus, once the spectrometer 50 of
In a time-of-flight mass spectrometer, heavier mass ions arrive at the ion detector after lighter mass ions. The analog signal from the ion detector 25 as a function of time exhibits peaks that can be identified as originating from ions of specific masses. A pulse in the analog signal is due to ions of a particular mass striking the ion detector 25 over a small duration of time. Ions of the same mass are generally bunched together as they travel toward and strike the ion detector 25 and will be referred to hereafter as an “ion packet.” Thus, ions within the same “packet” have the same mass. Further, pulses of the analog signal from the ion detector 25 will be referred to hereafter as “analog pulses.”
The ions of lighter mass ion packets tend to be bunched closer together than the ions of heavier mass ion packets. Indeed, for lighter mass ion packets, there is statistically less separation in time between multiple ion strikes from the same packet as compared to the statistic separation time for heavier mass ion packets. Moreover, the width of an analog pulse from the ion detector 25 for a lighter mass ion packet is usually smaller than the width of an analog pulse from the ion detector 25 for a heavier mass ion packet. Thus, although a high sampling rate may be desired to adequately sample the analog pulse for a lighter mass ion packet, a lower sampling rate may be adequate to sample the analog pulse for a heavier mass ion packet.
The decimator 66 of
In at least one embodiment, the decimator 66 allows a higher effective sampling rate for analog pulses corresponding with lighter mass ion packets and lowers the effective sampling rate for analog pulses corresponding with heavier mass ion packets. In this regard, the lighter mass ion packets arrive at the ion detector 25 before the heavier mass ion packets and, therefore, are sampled first. Thus, for each mass scan, the decimator 66 initially outputs samples at a high rate. However, over time, heavier mass ion packets begin to arrive at the ion detector 25 and are sampled. Moreover, after a predefined amount of time has elapsed, the decimator 66 lowers its rate of outputting samples such that digitized samples for the heavier mass ion packets are output at a lower rate as compared to the samples for the previously sampled lighter mass ion packets. In other words, the decimator 66 decimates the effective sampling rate of the system 51. The decimating of the effective sampling rate reduces the number of samples to be stored in memory 29 and, therefore, reduces the required memory size of memory 29.
In order to further conserve memory space as the mass scan progresses and, therefore, the desire for higher sampling rates abates, the effective sampling rate can be decimated more than once during the same mass scan as the ions in the packets reaching the detector 25 become heavier. However, once a mass scan is complete, the effective sampling rate is set to the initial rate that is to be used for the lighter mass ion packets to be received for the next mass scan.
In one embodiment, which will be described in more detail hereafter for illustrative purposes, the decimator 66 of
In the example shown by
Note that it is possible for the sampling rate of the A/D converter 27 to be adjusted as is described above for the decimator's output rate. In such a case, implementation of the digital filter 65, which will be described in more detail hereafter, and the decimator 66 would be unnecessary. However, changing the sample rate of a high speed A/D converter can be problematic. Moreover, using a decimator 66 to adjust the effective sampling rate of the system 51 is generally desirable so that the sampling rate of the A/D converter 27 may remain constant.
Reducing the output rate of the decimator 66 reduces the effective sampling rate of the system 51 even though the actual sampling rate of the A/D converter 27 remains unchanged. In this regard, the number of samples used to define the resultant mass spectrum is generally equal to the number of samples output by the system 51 per mass scan, not the number of samples output by the A/D converter 27 per mass scan. Moreover, the Nyquist criterion specifies that a sampling rate must be twice the highest frequency component of the sampled signal in order to generate an accurate representation of the signal. If the sampling rate of the A/D converter 27 is close to the minimum Nyquist rate (i.e., half of the highest frequency component of the analog signal from ion detector 25), then a reduction of the decimator's output rate is likely to reduce the effective sampling rate of the system 51 below the minimum Nyquist rate for the analog signal output by the ion detector 25. Thus, the digital filter 65 preferably filters the digitized samples from the A/D converter 27 such that the Nyquist criterion is not violated due to an adjustment to the system's effective sampling rate by the decimator 66.
For example, for each adjustment of the effective sampling rate by the decimator 66, the digital filter 65 may be configured to filter the samples from the A/D converter 27 such that its bandwidth is reduced by at least the same percentage as the adjustment to the effective sampling rate. Thus, in the embodiment described above referring to
If the output rate of the digital filter 65 is higher than the output rate of the decimator 66, such as may initially be the case when the decimator 66 reduces its output rate and the sampling rate of the A/D converter 27 remains constant, then at least some of the data from the digital filter 65 may be lost. For example, assume that the output rate of the decimator 66 equals the output rate of the digital filter 65 before an adjustment to the output rate of the decimator 66. If the decimator's output rate is reduced by one-half, as described above, then the decimator 66 may be configured to output every other sample received from the digital filter 65. In such an embodiment, one-half of the filtered samples may be discarded or lost.
Alternatively, to provide a more accurate resultant mass spectrum, the decimator 66 may be configured to combine samples to reduce or eliminate the number of samples discarded. For example, if the decimator's output rate is reduced by one-half, as described above, the decimator 66 may be configured to take an average of every two consecutive samples from the A/D converter 27 and to output each of the calculated averages as the digitized samples to be summed by summer 33. An exemplary embodiment implementing such an approach will be described in more detail below with reference to
In this regard,
Once the buffer 77 has accumulated a set of consecutive samples to be averaged, the sample combiner 78 sums the samples within the set and divides the sum by the total number of samples of the set. The result is the average of the samples in the set, and this average is output by the signal combiner 78 to the summer 33 (
In the following description, assume for illustrative purposes that the output rate of the decimator 66 is to be controlled in accordance with the graph depicted by
In block 108, the controller 52 begins a mass scan by causing the pulse source 17 to generate a pulse, which causes ions in the ion source 11 to be accelerated toward the ion detector 25. In block 111, the controller 52 provides control signals to the decimator 66 such that the output rate (RD) of the decimator 66 is equal to the sampling rate (RA/D) of the A/D converter 27 divided by 2x. Initially, x is zero (0), and RD is, therefore, equal to RA/D. Thus, the controller 52 increments the address register 21 each clock cycle similar to the conventional controller 15 of
In
In block 121, the controller 121 compares t to T. Until the end of the mass scan, t will be less than T, and the controller 52, therefore, returns to block 111. Now that x equals one (1), the controller 52 provides control signals to reduce the decimator's output rate to one-half of its original output rate or, in other words, to reduce the decimator's output rate to one-half of RA/D. The controller 52 also provides control signals to the digital filter 65 to reduce the filter's bandwidth to one-half the bandwidth of the analog signal from the ion detector 25.
When the output rate of the decimator 66 is reduced to one-half of RA/D, the sample combiner 78 of the decimator 66 takes an average of every two consecutive samples received from the A/D converter 27 and outputs such average in lieu of the two consecutive samples. Note that, unless otherwise indicated herein, samples are “consecutive” if they are successively taken by the A/D converter 27. In addition, the controller 52 begins to increment the address register 21 every other clock cycle rather than every clock cycle. Further, a value is read out of the memory 29 and summed by summer 33 with the output of decimator 66 only when the address register 21 is incremented. The result of foregoing changes is illustrated by
In this regard, assume that address (n−4) is storing the running sum for the last sample output by the decimator 66 just before the decimator's output rate is reduced from RA/D to one-half of RA/D. In such an example, address (n−4) is storing Samplen−3 as it does in the embodiment depicted by
After one-half of the mass scan is complete, the controller 52 makes a “yes” determination in block 116 of
When the output rate of the decimator 66 is reduced to one-quarter of RA/D, the sample combiner 78 of the decimator 66 takes an average of every four consecutive samples received from the A/D converter 27 and outputs such average in lieu of the four consecutive samples. In addition, the controller 52 begins to increment the address register 21 every four clock cycles rather than every other clock cycle. Thus, an average for four samples from A/D converter 27 is summed every four clock cycles with a corresponding running sum in memory 29.
After three-quarters of the mass scan is complete, the controller 52 makes a “yes” determination in block 116 and, therefore, increments x to a value of three (3) in block 118. Now that x equals three (3), the controller 52 provides control signals to reduce the decimator's output rate to one-eighth of its original output rate or, in other words, to reduce the decimator's output rate to one-eighth of RA/D. The controller 52 also provides control signals to the digital filter 65 to reduce the filter's bandwidth to one-eighth of the bandwidth of the analog signal from the ion detector 25.
When the output rate of the decimator 66 is reduced to one-eighth of RA/D, the sample combiner 78 of the decimator 66 takes an average of every eight consecutive samples received from the A/D converter 27 and outputs such average in lieu of the eight consecutive samples. In addition, the controller 52 begins to increment the address register 21 every eight clock cycles rather than every four clock cycles. Thus, an average for eight samples from A/D converter 27 is summed every eight clock cycles with a corresponding running sum in memory 29.
At the end of the mass scan, the controller 52 makes a “yes” determination in block 116, as well as in block 121. Thus, in block 125, the controller 52 determines whether another mass scan is to be run. If the desired number of mass scans have yet to be performed to achieve the desired statistical accuracy for the resultant mass spectrum, the controller 52 resets x to a value of zero (0) in block 128 and repeats the aforedescribed process for the next mass scan. Once the desired number of mass scans have been performed, the process depicted by
Since the effective sampling rate of the system 51 is reduced during the mass scan, the total number of data points in the memory 29 for the spectrometer 52 is less than the total number of data points in memory 29 for the conventional spectrometer 10 of
It should be noted that it is unnecessary for the effective sampling rate to be reduced by a factor of two for each rate reduction or for the effective sampling rate to be periodically reduced at equal time intervals. Further, it is unnecessary for the bandwidth of the filter 65 to be reduced by the same percentage or at the same times as the effective sampling rate reductions provided that the Nyquist criterion remains satisfied.