Integration Settings

Integration component controls two factors:

• Duration of each individual integration event within overall cycle;
• How data produced by the Detector during this event are handled.

A single, complete Detector accumulation cycle is the smallest element Integration can operate on - Integration does not affect how data are manipulated within Detector acquisition routine.

Integration component has two tabs:

• Parameters tab contains setting relevant for routine operation;
• Setting tab contains more general setup and troubleshooting settings.

Parameters tab

There is no explicit control over duration of an Integration step. Instead, initial integration duration is defined as sum of all included sub-steps (in their normal sequence):

• Probe shutter pre-roll  - time allowed for probe source to open;
• Discard - number of Detector cycles that should be purged or discarded;
• Average – number of Detector cycles that should be averaged together;
• Probe shutter post-roll – time allowed for probe source close.

Thus, for a single Detector accumulation of 2 sec, Integration cycle may be:

[100 ms to open shutter] + [2 sec × discard 2 cycles] + [2 sec × average 10 cycles] + [100 ms to close shutter] = 24.2 sec

Discard

Discard specifies the number of Detector acquisition cycles that should be discarded. During these cycles data are being acquired and read from the detector, but are not saved. The purpose of Discard is to allow an option of purging the detector from pre-exited signal and for the entire system to come to reproducible cycle timing.  

Average

After the number of Discard cycles is elapsed, program will continue to read, without interruption, the Average number of cycles. Data acquired during these cycles are averaged together. For multi-channel detectors averaging is done channel-by-channel – upon completion of Integration step data are of the same type and dimension as input data from the Detector. Resulting data are later passed on to Processing. 

Cycle factor

Initial timing defines duration of the first cycle only. Duration of all subsequent cycles is calculated by multiplying duration of the preceding cycle by cycle factor. If cycle factor equals to 1.0, duration of each subsequent cycle will remain the same. Application of cycle factor can be delayed for a number of integration cycles if after value is non-zero. Note that after value specifies number of starts of Integration steps (effectively overall acquisition cycles) and not individual accumulations by the Detector. Cycle factor is especially useful for time-resolved measurements as it allows to slow down data flow as reaction gets progressively slower.

How cycle duration is calculated

Duration of the first cycle is calculated from anticipated duration of Detector accumulation and Discard and Average settings – this becomes duration of an Ideal Cycle. Shutter steps are considered to be an overhead and not a part of the cycle for timing purpose.  At the end of each cycle the Ideal Cycle duration is multiplied by cycle factor and stored for further calculations on the next cycle. The number of complete Detector accumulations is calculated from the Ideal Cycle duration. If additional complete accumulations can be done within the Ideal Cycle, number of Detector accumulations is increased accordingly. If increase in the Ideal Cycle is less than a complete accumulation, Discard and Average parameters do not change on this cycle. On the next cycle, however, the Ideal Cycle will increase again and may accommodate additional complete accumulations.

Keep average

Keep average option controls how the count of additional cycles is distributed between Discard and Average steps. When Keep Average is enabled, the number of accumulations averaged together will not change, while Discard steps will vary with cycle duration.

When Keep Average is disabled, all changes in the number of cycles will be applied to Average. If Discard has been set to zero before the measurement, not data will be discarded throughout, while the number of spectra being averaged may increase from 1 to 1000. In some cases increase in the number of spectra being averaged is beneficial because it will improve S/N ratio. In other cases it may be best to keep average constant.

Shutter

Shutter box controls operation of a probe shutter which is directly related to data acquisition. Shutter box is not present when I/O module is not available. The role of probe shutter is very different from that of exposure shutter(s) described here.

Shutter wraps all other steps within Integration and includes two sub-steps:

• Pre-roll – delay from the beginning of integration step to the moment when acquisition is requested from the Detector;
• All data acquisition and integration;
• Post-roll – delay from the end of last acquisition to moment when data are passed along to the Processing.

For probe shutter Post-roll delay can be typically set at zero, i.e. no wait for shutter closing. Pre-roll delay of 100-150 ms may be sufficient in most cases.

For fast, time-resolved measurements it may be necessary to force shutter to open position and disable its operation. Disabling probe shutter will skip this step altogether, reduce overhead on the cycle, and improve time resolution.

Setting tab

Settings tab contains mostly parameters relevant for troubleshooting of software/hardware issues. The only practically relevant parameter is Beep on data option.

• Beep on data – provide audible feedback of data arrival from the Detector. The last data in Integration has different sound pitch that other data. Besides its direct purpose of notification about arrival of new data, beep has an important psychological role of providing company and comfort in long solo experiments.
• Data, Post-roll, and Next step call – control mechanisms by which control is passed on to corresponding steps.
• Direct – calls function directly; this is faster way but in some cases in the past is has caused overflow error when data acquisition is calling itself recurrently.
• Message – uses Windows message that is sent though the central messaging system; Message takes longer to deliver, is less predictable, and is more affected by other processes on the system than direct call, but it allows to clear stack and to avoid overflow.

If overflow occurs, one Message anywhere within overall cycle should be enough.  

• On overflow – optional action to take if new data have been produced before previous data have been processed and purged.