Noise and drift

As well as noise arising due to particular electrical or optical devices, noise may be caused by external sources, such as mechanical vibrations, or pick-up of external electrical signals. Detector noise tends to increase with the age of the instrumentation. Certain components have limited lifetimes and are considered as expendable. An example is the deuterium lamp of a UV absorbance detector, which typically has a rated life of 2000 h or so. After this time the lamp will probably not fail completely. 

---

There are three different types of detector noise, short term noise, long term noise and drift. These sources of noise combine together to give the composite noise of the detector. The different types of noise are depicted in figure 3. 

Figure 1.18 Methods for calculating short- and long-ten noise and drift for chromatographic detectors.

Many instruments utilize a double beam principle in that radiation absorbed or emitted by the sample is automatically compared with that associated with a blank or standard. This facilitates the recording of data and corrects for matrix effects and instrumental noise and drift. Instrumentation for the generation of radiation is varied and often peculiar to one particular technique. It will be discussed separately in the relevant sections. Components (b) and (c), however, are broadly similar for most techniques and will be discussed more fully below. 

Obviously the optimum potential for detection of analyte X in this mobile phase is a compromise a higher potential will increase the peak height, but will also increase the baseline offset and thus the baseline noise and drift. Lower potentials will decrease the peak height, but also decrease the noise and drift. 

To avoid noise and drift contribution by amplifier electronics, the optimum range in practice is 5 to 25%. 

The working electrode must be fitted in its holder, tight enough to prevent mobile phase from leaking into the space between the working electrode and its fitting. In practice this process appears to be a major source of noise and drift and is very difficult to overcome. 

All detectors need to be verified for meeting their noise and drift specifications. These tests are often performed with a dry cell that is sealed and has no actual solvent flow going through it. Some detectors have on-board diagnostics for these tests, with digital performance readouts. 

The first experiment of this set (row 1) is used to establish the noise and drift characteristics of the detector. 1 mL/min water (with 1% methanol) is pumped through the system. The system is allowed to equilibrate then, a 1 pL water blank is... 

The parameters that require qualification for a UV absorbance detector are wavelength accuracy, linearity of response, detector noise, and drift. These determine the accuracy of the results over a range of sample concentrations and the detection limits of the analysis. 

A test for noise and drift would normally follow the recommendations of the instrument vendor. 

Detector sensitivity is one of the most important properties of the detector. The problem is to distinguish between the actual component and artifact caused by the pressure fluctuation, bubble, compositional fluctuation, etc. If the peaks are fairly large, one has no problem in distinguishing them however, the smaller the peaks, the more important that the baseline be smooth, free of noise and drift. Baseline noise is the short time variation of the baseline from a straight line. Noise is normally measured "peak-to-peak" i.e., the distance from the top of one such small peak to the bottom of the next. Noise is the factor which limits detector sensitivity. In trace analysis, the operator must be able to distinguish between noise spikes and component peaks. For qualitative purposes, signal/noise ratio is limited by 3. For quantitative purposes, signal/noise ratio should be at least 10. This ensures correct quantification of the trace amounts with less than 2% variance. The baseline should deviate as little as possible from a horizontal line. It is usually measured for a specified time, e.g., 1/2 hour or one hour and called drift. Drift usually associated to the detector heat-up in the first hour after power-on. 

In practice, the real signal is never ideal. Systematic and random errors often occur due to, e.g. unresolved or badly resolved peaks, non-linearity (resulting in concentration-dependent peak shapes), noise and drift. 

A chromatogram without noise and drift is composed of a number of approximately bell-shaped peaks, resolved and unresolved. It is obvious that the most realistic model of a single peak shape or even the complete chromatogram could be obtained by the solution of mass transport models, based on conservation laws. However, the often used plug flow with constant flow velocity and axial diffusion, resulting in real Gaussian peak shape, is hardly realistic. Even a slightly more complicated transport equation... 

When characterizing copolymers, it is necessary to have two detectors in series, e.g., a refractometer with either a UV detector or an IR detector. An IR detector is preferred for the detection of polyalkenes at elevated temperatures because baseline noise and drift are much less than for the refractometer detector. 

Noise and Drift. Electronic, pump, and photometric noise poor lamp intensity, a dirty flow cell, and thermal instability contribute to the overall noise and drift in the detector. Excessive noise can reduce the sensitivity of the detector and hence affect the quantitation of low-level analytes [13,14]. The precision of the... 

Nowadays, most chromatographic software is capable of calculating the detector noise and drift. Typically, the detector should be allowed to warm up and stabilize prior to the test. Temperature fluctuations should be avoided during the test. The noise and drift tests can be performed under static and dynamic conditions. For a static testing condition, the flow cell is filled with methanol, and no... 

Make sure that the flow cell is clean and free of gas bubbles when performing the detector performance tests. Dirty flow cell and gas bubbles are the main reasons for poor results for detector noise and drift. 

A stable temperature must be maintained when performing the noise and drift tests. 

In reality, the performance of the LC system will deteriorate over time, especially for noise and drift. If the performance verification tests do not pass the predetermined acceptance criteria, an impact assessment should be made to evaluate the effect of the failure on the quality of the data generated by the system. The impact assessment should cover all the analyses done on the system since the last performance verification, as there is no effective way of determining when the failure occurred. The system suitability data generated together with the analyses will be very useful in demonstrating that system performance is adequate for the application at the time of analysis, so that any data generated are reliable [19]. 

It would be very convenient if the TCD had noise levels down in the nanovolt region. Amplifiers can be built without too much difficulty which contribute no more than 25 nV of their own noise. Unfortunately, the TC detector is subjected to many extraneous influences. It is the fluctuations in these which dominate the noise and drift seen on the recorder. 

Spurious peaks could occur due to volatile compounds emitting from the flow controller diaphragm. This problem can be solved by use of metal diaphragms. Some manufacturers have used short molecular sieve traps between the flow controller and the inlet system, but, if a poor inlet system is used, the sample can flash-back upon injection and condense on the trap, later causing spurious peaks or noise and drift as the sample slowly comes back off the trap. 

Figure 7.14 Recognized types of noise short-term noise, long-term noise, and drift.

Refractive Index Detectors These detectors respond to changes in refractive index (positive or negative) arising from the presence of a compound in the eluent. All the factors which can affect refractive index must be carefully controlled (e.g. temperature, eluent composition, pressure) otherwise noise and drift will limit the sensitivity. Thus the chromatograph is best placed in a thermostatically-controlled cabinet and good pumps are desirable to minimise pressure fluctuations. Changes in eluent composition will also cause spurious changes in refractive index. 

Detector noise is the term given to any perturbation on the detector output that is not related to an eluted solute. It is a fundamental property of the detecting system and determines the ultimate sensitivity or minimum detectable concentration that can be achieved. Detector noise has been arbitrarily divided into three types, short term noise, long term noise and drift all three of which are depicted in figure 4. 

---