Page 1 of 1

resfields artifacts, pepper with eigfields instead?

Posted: Mon Feb 15, 2016 4:30 pm
by myersw802
In the code below (plot attached) you can see different numbers of transitions present depending on the range specified. By comparing resfields and eigfields, it's clear that resfields is missing transitions in this case. Is it possible to run pepper with eigfields? Any examples of brute force code, if not?
thanks.
-Will
~~~

clear all; clc;

sys=struct('S',3/2,'g',2,'D',20000,'Nucs','135Cs','A',200,'lw',0.1);
exp=struct('mwFreq',9.5,'Range',[1 700],'nPoints',2000);
opt=struct('nKnots',128,'Method','matrix');
[x,y] = pepper(sys,exp,opt);

% here, the Range is increased, leading to large differences in low field
exp.Range=[1 1500];
exp.nPoints=round(2000*15/7);
[xL,yL] = pepper(sys,exp,opt);

% see that no low field transition is observed in resfields
exp.CrystalOrientation = [0 90].*pi/180;
[xe,ye] = eigfields(sys,exp,opt);
yme = max(ye);
[xr,yr] = resfields(sys,exp,opt);
ymr = max(yr);

%% plotting
figure(1); clf; hold on;
plot(x,y+15,'k','LineWidth',1.5); plot(xL,yL,'r','LineWidth',1.5);
plot((xe*[1 1])',(ye/yme*[0 max(y)])'-1.05*max(y),'b','LineWidth',1.5);
plot((xr*[1 1])',(yr/ymr*[0 max(y)])'-2.10*max(y),'m','LineWidth',1.5);
set(gca,'LineWidth',1.5); xlim([0 1500]);
xlabel('B_0 (mT)'); ylabel('EPR simulation [a.u.]');
text(1150,0.75*max(y),'1. range=690mT','Color','k');
text(1150,0.25*max(y),'2. range=1490mT','Color','r');
text(1150,-0.5*max(y),'3. eigfields','Color','b');
text(1150,-1.5*max(y),'4. resfields','Color','m');
pbaspect([3 1 1]); box on;

Re: resfields artifacts, pepper with eigfields instead?

Posted: Tue Feb 16, 2016 8:02 pm
by Stefan Stoll
Have you tried Opt.Threshold=0? This bypasses the automatic transition selection in pepper and resfields and forces ES to calculate all transitions.

Re: resfields artifacts, pepper with eigfields instead?

Posted: Wed Feb 17, 2016 3:53 pm
by myersw802
Okay, RT(F)M, thanks. That was it. We know from field vs. freq. calculations there a huge range of transition probabilities, so it is exactly as I now see stated in the 'pepper' documentation.