Automatic Signal/Noise calculation

Hello,

If you search the help, you can find more details on the SN calculations. For automatic calculations the last blank in the results set prior to the injection is used. So, if you run blanks throughout the sequence a blank is used for the injections after it until another blank is run. The auto time range calculation is shown below, it uses the RT of the peak in the calculation. 

Marty

An automatically determined time region is calculated according to one of the following algorithms:

• If the reference signal is not long enough

  (EndTime - StartTime < 20*W₅₀)

  • StartTime = starttime (of reference signal), and

  • EndTime = endtime (of the reference signal)

• If the reference signal is long enough, but the peak is situated too close to the starttime

  (t_R-10*W₅₀ < starttime of the reference signal)

  • StartTime = starttime (of reference signal), and

  • EndTime = StartTime + 20*W₅₀

• If the reference signal is long enough, but the peak is situated too close to the endtime

  (t_R+10*W₅₀ > endtime of the reference signal)

  • EndTime = endtime (of the reference signal), and

  • StartTime = EndTime - 20*W₅₀

• If the reference signal is long enough, and the peak is situated far enough away from starttime and endtime of the reference signal

  (t_R-10*W₅₀ > starttime, t_R+10*W₅₀ < endtime)

  • StartTime = t_R - 10*W₅₀, and

  • EndTime = t_R + 10*W₅₀

where

t_R is the retention time, and

W₅₀ is the peak width at half height.

Hi Martin, I am using the setting as below (All peaks, ASTM, Automatic). Unfortunately, the S/N for my peak (Toluene) is not calculated and the warning of " At least one peak where noise ASTM on blank with automatic range could not be computed"  is appeared. My blank (Methanol) has not peaks other than the solvent peak. The blank injection is just before the suitability injection.

What could be the reason and how to solve this? 

Thank you

Kris

Hello,

Without looking at the data, I would guess the issue is the auto time range is calculating less than a 1 min window. The one peak where you have noise calculated has a PW of 0.059 which would give a window just over 1 min. 

Marty

Hello martin.adams , 

The European Pharmacopeia 2.2.46 (Eur. Ph) and the United States Pharmacopeia <261> (USP) have changed their criteria for calculating the signal-to-noise parameter. See attached images.

The width becomes 2.5*W50 on each side.

Could you tell me if in the new versions of Openlab this calculation will be fixed? For older versions will a patch be released to change the calculation?

Thanks a lot!! 

Òscar 

Hello,

CDS 2.8 will update the suitability calculations. I will add the release notes to this thread once finalized.   

Marty

Hi martin.adams , you are exactly right, the peak width is small, therefore it is not giving the final S/N value. When I manually integrated the peak by increasing the peak width, it gave the value. It seems i need to fix the time range for the baseline calculation. Thank you for your reply.