S/N Ratio

[wrighty]

Afternoon all,

i was hoping that somebody could help me with a chemstation query. I am currently developing a method and i am trying to determine the level of quantitation (spelling?). i understand the generic accpetance criteria for this test is a ratio of 10:1 between the signal and baseline noise.

Firstly, on chemstation reports there is an option to print off the performance and noise of the sample but when i do this i just get "-" and no value. Does anybody know how to rectify this?

Secondly, if the first option is no good, are there any recommendations on how to go about determining the LOQ?

Thanks in advance for your help

Steve

[GasMan]

Wrighty,

You must set up parts of your chromatogram where you can measure the noise.

First, make sure that the 'View' mode is for 'Full Menu'. When you are in 'Full Menu' and you click on 'Report', you will see a menu entry titled 'System Suitability'. Click on this and you see an entry 'Edit Noise Ranges'. You should enter times in your chromatogram where there are no peaks coming off. Try to get times at least a minute long.

Gasman

[stevel]

If you go to the Specify Report page of the method, one of the report formats is "Performance plus Noise". After you choose that format, you can edit your method again. There will be a new page in the method that will allow you to choose your noise window(s).

ChemStation calculates noise 3 different ways. The first is 6*SD. For this one, the window that you choose should consist of perfectly flat baseline. Any one-time bumps or baseline drift will be calculated as noise. This method will always give you the highest noise signal, or lowest s/n. It is also the one that ChemStation uses to print its s/n values.

The 2nd method is peak-to-peak. Just like it says, The calculated noise signal is just the difference between the highest point and the lowest point in your chosen window. The same limitations as 6*SD apply so you want to be very picky about your window that you choose.

The 3rd type is the most realistic as far as I'm concerned. The ASTM method breaks the chosen window into 0.1 min intervals and calculates the noise over each. Baseline drift does not present a problem for this calculation so you options for choosing the window are greater. This one will always give the lowest noise level, or the highest s/n. The window must be at least 1.1 min wide for this method to work, though.

[wrighty]

cheers fellas, thats got it sorted

steve