### Author Topic: Shaped pulse example  (Read 3416 times)

#### vesco

• Member
• Posts: 2
##### Shaped pulse example
« on: January 05, 2016, 05:13:11 PM »
Hi,

I'm using spinach_1.6.2782 with matlab R2015a

In the shaped_pulse_1 example:

inter.zeeman.scalar=num2cell(linspace(-4,4,31));

produces a single resonance.

Changing to:

inter.zeeman.scalar=num2cell(linspace(1,5,31));

produces multiple resonances.

Am I missing something? Can anyone help?

Thanks,
Vesco.

#### kuprov

• Member
• Posts: 59
##### Re: Shaped pulse example
« Reply #1 on: January 06, 2016, 10:18:15 AM »
The shaped pulse excitation centre is set at zero ppm in those examples.

Check your Y axis scaling in the second plot, it is likely that Matlab simply auto-scales tiny signals that get excited to a negligible extent in [1,5] ppm interval.
« Last Edit: January 10, 2016, 07:07:05 PM by kuprov »

#### vesco

• Member
• Posts: 2
##### Re: Shaped pulse example
« Reply #2 on: January 11, 2016, 01:29:21 AM »
Thanks.

Another newby question:

Regarding the calibration_factor calculation in the example -

How would I convert the power_level for example used in the universal_rotation_grape_1 example

% Power levels
rf_power=2*pi*25000;              % Nominal power level

I interpret this to mean that this is a power level producing a 10 us 90 degree pulse.

How should I scale the amplitudes for use in shaped_pulse?

For example, a 500 step 1 ms pulse made with rf_power = 2*pi*25000

calibration_factor = 100 * pi/2 / (1e-3) ?

« Last Edit: January 11, 2016, 02:21:49 AM by vesco »

#### kuprov

• Member
• Posts: 59
##### Re: Shaped pulse example
« Reply #3 on: January 16, 2016, 04:16:49 PM »
No, that's the power level that corresponds to "1" in the optimized waveform. This scaling is done to improve the internal numerical stability, GRAPE module always uses these fractional powers to keep the Hessian condition number reasonable. Let me take a look at your code, it should be a simple enough edit...