Sense of the Gain of a fluorescence detector

[LCFan]

Hello,

I performed 4 injections of the same solution using different gains of the fluorescence detector. The unit of the ordinate was energy.

Thus, I got a linearity "peak area vs. gain" => higher gain results in higher peak areas.

But in the same way, I observed: higher gain results in higher noise.

The lowest gain I tried was 1, the highest was 20. The S/N ratio was around 100 for all gains. Thus, I get no additional sensitivity by increasing the gain.

Thus, my question: what is the real sense of the gain? To increase peak areas in a way that differences in integration from different injections are relatively lower due to the high areas?

I am puzzled about this.

Thank you very much in advance for your comments.

Florian

[tom jupille]

Thus, my question: what is the real sense of the gain?

For one thing, to match the signal to the optimum voltage range of the A/D converter.

[lmh]

It may depend on your detector. Using a Thermo FDplus on a weak peak, it's quite possible to have a situation where at the lowest gain, electronic noise is nearly as large as the peak. On this instrument, the peak-to-peak random noise (which I'm assuming is down to the electronics and an imperfect A:D converter) doesn't increase terribly with gain, so at a higher gain, the peak benefits.

A lot also depends on how S:N is defined, and what sample you are running. For example, software may calculate S:N using peak-to-peak noise in what it considers to be baseline adjacent to the peak. But if, in fact, there are weak peaks eluting in the "baseline" section, then the measure of S:N is actually the ratio of a strong peak to a weak peak, which doesn't change with gain (up to the end of detector linearity).

[Hfranz]

Changing sensitivity/gain of the detector should have an impact on the S/N. Changing this parameter changes the saturation of the photomultiplier, so increasing it should increase the saturation. Typically, you can optimize the saturation to an optimal value that provides the best S/N. I am surprized to hear that this does not seem to be the case for your detector. For some more background information on sensitivity/gain and other fluorescence-related parameters I can recommend this Fluorescence Method Development Handbook: http://www.dionex.com/en-us/webdocs/113 ... PN2988.pdf

[tom jupille]

I am surprized to hear that this does not seem to be the case for your detector.

If the detector is the dominant source of noise, then you are correct. However, if the dominant source of noise is in the chromatography (rather than in the detection), then it's not surprising at all.

[Hfranz]

I am surprized to hear that this does not seem to be the case for your detector.

If the detector is the dominant source of noise, then you are correct. However, if the dominant source of noise is in the chromatography (rather than in the detection), then it's not surprising at all.

Good point.