Depos that are created between the response plane and the wire planes are currently dropped. The induction they produce is truncated starting at some finite field response. This starting point depends on the X location of the depo. Further, given the nonuniform fields near the wires it is not trivial to even find the nearest drift path. These irregularities can not be correctly handled with the optimized "zipper" algorithm.
The simplest approximate correction is to "back up" the depo in space and time given the nominal drift velocity. Ie, if a depo is at (x,t) it is moved to (X_rp, t - (X_rp-x)/v) where X_rp is the X location of the response plane. Residual error will be worse as (x,t) approaches collection. There, velocity can change substantially and simply projecting back along nominal drift lines will not place the "backed up" depo exactly in the write spot.
The full fix would be to perform a detailed convolution with truncated field responses as special case for these depos. For this tick, just back 'em up.
Depos that are created between the response plane and the wire planes are currently dropped. The induction they produce is truncated starting at some finite field response. This starting point depends on the X location of the depo. Further, given the nonuniform fields near the wires it is not trivial to even find the nearest drift path. These irregularities can not be correctly handled with the optimized "zipper" algorithm.
The simplest approximate correction is to "back up" the depo in space and time given the nominal drift velocity. Ie, if a depo is at
(x,t)it is moved to(X_rp, t - (X_rp-x)/v)whereX_rpis the X location of the response plane. Residual error will be worse as(x,t)approaches collection. There, velocity can change substantially and simply projecting back along nominal drift lines will not place the "backed up" depo exactly in the write spot.The full fix would be to perform a detailed convolution with truncated field responses as special case for these depos. For this tick, just back 'em up.