Geo,

Did you auto reset the fast and slow thresholds for the 9.6 usec count? Note the input count well exceeds your total count meaning there is still a lot of low energy noise that needs to be taken care of. Going to the lower 4.8 usec is going to make matters worse in resolution and noise. A 2048 conversion gain is the sweet spot. Use shorter peaking time for high count rate, longer for low count rate and better resolution. But you need do need to reset your filters when you change parameters.

Dud

?

?

Run it for the same time change only one variable at a time

Dud

?

Geo / Charles,

Changing only the peaking time is not going to give you a good idea of what an optimal setting should be. These things interact with each other and need to be set together, The noise base line and valid count line are set with the fast and slow thresholds, when you change peaking time or anything else you need to reevaluate all these settings. The low energy noise (<2keV) we commonly see in these spectra are due to not having properly set all these parameters together.

You can see this with the Total count vs Input count. Input - Total = count noise. That number is minimized using the fast and slow threshold settings with a given Peaking Time. If you have a high count rate use a shorter peaking time .Don’t kid yourself you’ll that you’ll see anything useful <1.e keV without a gas flow so don’t sweat the small stuff when setting this up but make sure the noise down there has been cut off as its going to affect your count time to get a good peak.? ?

The 19.2 usec PT has an input rate of 9.28 valid cps using 3705 counts out of 23,665 while the 9.6 usec PT had only a 4.27 valid cps input rate using only 90,456 counts out of 227,418 noise counts. A shorter PT means more noise and lower resolution.? Noise and pile up are going to lower the number of valid non rejected counts in any given time frame. When noise is minimized by both ?PT and threshold settings this electronic noise is minimized allowing valid counts in a shorter time frame with less noise interference which is slowing you down.

For those using a weak Am source vs a high power X-ray source count time means everything so you need to optimize the PT, set the thresholds, and don’t run anything over a 2048 conversion gain.

Dud

?