Feedforward



next up previous contents
Next: Feedforward and Feedback Up: Overview Previous: Feedback

Feedforward

Feedforward is designed to prevent intensity variations of the beams from making the scavenger beam energy vary. Without any correction, longitudinal wake fields (beam loading) would reduce the scavenger beam energy by 2.0 GeV as all three bunch intensities (positron, electron, scavenger) went from 0 to .

The trick to correcting this is that we know the beam intensities well before the beams are accelerated down the LINAC. Each beam is stored in a damping ring for at least 8.3 milliseconds (1/120 second) before being accelerated. As very little beam is lost on extraction from the ring, the intensity measured with BPM strips 5.3 milliseconds before extraction can be used to estimate the amount of beam loading. Feedforward does this using a combination of analog and digital electronics (but no micro-processors). This electronics resides in the domain of the DR12 micro. There are various gains, offsets and readbacks available on the VAX via the DR12 micro.

This estimate of the beam loading is then sent in digital form over a twisted pair communications link to the second feedforward electronics box which lives in the domain of the EP01 micro. The communications link is similar (but slower) to the one used for the klystron veto system. This feedforward box in EP01 (which contains the DACs that feedback wrote to) calculates the phase shifter settings needed to compensate for the beam loading. These voltages go to the phase shifters and voilá we have compensated for the beam loading. All of these calculations take about 100 micro-seconds (considerably less than the available 5.3 milli-seconds) and are repeated at the full beam rate.



SLAC Controls Software
Fri Nov 4 11:34:56 PST 1994