Here's how we apply gradients on our system:

1) within the pulseprogram we define, for example,
    ....
    #define GRADFAC 2
    #define GRADLENGTH 800u
    #include <gradients.def>
    ....
2) the gradients.def file is placed in the pulse sequence directory
3) gradients of fixed length GRADLENGTH = FACTOR*400u ,
   and fixed # elements (100)
4) hardware: blanking device BGB30 and 10Amp gradient amplifier BGU-10
   :c4 (rcp4) for gradient-unblanking (Bruker B-GB30)
   :c16 (rcp16) negative gradient on pin rcp16(gpol) of BGU-10
   these signals can be taken from the BackPanel plugs (? BP1 or BP2)
5) the gradient file (N_8shape8.r) consisting of 8 elements is loaded once
   in the gradient-waveform memory only, independent from DS and NS.
   negative gradients are generated via the hardware rcp16 signal.
6) the experiment is repeated twice for each value of t1:
   once as
    .....
    GRAD(11,2)
    .....

   and once as
    ....
    NEGGRAD(31,2)
    ....

7)  the created echo-antiecho signal can be processed with
    MC2=echo-antiecho