Optics

The following applies to the horizontal plane. The vertical plane is unimportant, so long as there is no large beam loss.

All subsequent focci are images of the beam at SC260. Therefore the beam size at this point must be minimised. The FWHM should not exceed 1mm.

Subsequent images are subject to substantial 2nd-order distortion, ie. the beam has an asymetric "tail" (LHS in SC482, RHS in FS670). This distortion will be corrected by the magnetic multipoles.

The settings of the frontend quadrupoles vary slightly according to the ion source. For plasma sources, the voltages should be increased very slightly to compensate for the closer ion source. Other quadrupole settings should not need tuning. Good QP-arrays are HRS_SEP_1-6-7.csv (Mk.7 source) and HRS_SEP_1-5-10.csv (Mk.8 source).

Steering should be minimised wherever possible. Large corrections cause additional beam distortion. If the beam is off axis, turn off all (horizontal) steering: EXTX (centred at 2048), QS30, XY120, XY200. Then see what correction is needed to bring the beam on axis. Ideally the beam should be centred on WG110 and WG210, XY120 and XY200 should be off, and QS30 and EXTX should correct in the same sense. If a large correction is needed, use EXTX, QS30, and XY120 (all working in the same sense) to centre the beam on SC260 (the slit SL240 can be used to callibrate the zero point on the scanner). Then adjust XY200 and MAG90 until the beam is centred on SC482 and WG580 (again, the zero point on SC482 may be checked/callibrated by using the narrow slit in WG470). Note, at the moment it seems impossible to achieve perfection. The beam is 8mm (or 1 wire) to far right on WG580. Under investigation!

It is difficult to callibrate MAG60 so that the final focus is perfectly centred on FS670. Sometimes it is easier to centre the slit SL675 on the beam rather than vice versa. In cases where the beam is radioactive and invisible on the scanners, the slit positioning must be done in conjunction with a radiation detector (as per the rubidium-74 run in June 2000).