Hi, E. Blasberg wrote on Sun, 15 Feb 2004 12:52:11 +0200: ... > Well, they're expecting very large peaks on the order of microseconds > and they were afraid that too low a sampling rate would miss the true > peak. Acquiring analog signals always inherits the risk of 'missing' the peak (in the sense of not getting its highest value). There's almost ever some analogue filtering that changes the amplitude and almost ever also the phase af a signal. The point is to decide what is acceptabel and what not. IF one, however, is only interested in the resulting amplitude or the best time approximation of any peak, one could estimate those with some calculation. Search your waveform for any raw peaks. Get the nearest neigbours around the most extreme value hou have measured (snippet). Perform a quadratic fit on each snippet for A = at^2 +bt. Do a mathematical differentiatiation and set this equation to zero (find the most extreme value of the fitted equation): 0 = 2at+b. Get you value of t from that and calculate your value of A from the fitted equation. Voila!
This may increase the timing resolution by at least 10 fold or even better, depending on the clock stability of the DAQ device. Just my € 0.02! Greetings from Germany! -- Uwe Frenz ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Dr. Uwe Frenz Entwicklung getemed Medizin- und Informationtechnik AG Oderstr. 59 D-14513 Teltow Tel. +49 3328 39 42 0 Fax +49 3328 39 42 99 [EMAIL PROTECTED] WWW.Getemed.de
