Magnet coils

We will apply a uniform field of approximately 70 Gauss over the TPC volume perpendicular to the beam direction. This field will be created by several turns of water cooled conductor wound in a saddle shape placed in one of several possible positions. The coil can easily be placed anywhere from just outside the TPC pressure vessel to completely outside the hodoscope. More power is required for larger designs of course and we currently favor a position just outside the Sindrum 3 chamber. We have made MC scattering studies of this position.

The location and design of the coil is chosen to

• minimize its interference with the detector components; in case of disassembly for the repairs and maintenance of the detector and coil
• minimize the distortion of the electron trajectories due to scattering and the field itself
• keep the coil power and the resulting temperature rise low to avoid heating the detector and target vessel
• minimize the material and construction costs
• allow easy rotation of the magnet for two perpendicular field directions

The two most studied options are the one wound just outside the pressure vessel, and the ones between the Sindrum 3 and 5 chambers. The smallest one has the disadvantage that the connections would need to pass through notches in the TPC chamber flange. However accessing the coils would be fairly easy because they would roll out with the TPC. There is a lot more room to work between Sindrum 3 and 5 and we currently favor the smallest diameter coils that would fit just outside Sindrum 3. The final decision on the position will be made after we finalize the Sindrum wiring details since the signal wires and magnet wire must pass close to each other.

A prototype of the magnet was used at PSI in May 2000. It produced 70 Gauss cooled with air. It was made of two coils with a cross section consisting of a single layer fixed to a 10 cm diameter beam pipe. With uniform current density $\pm$60$^o$ from horizontal, leaving a 60$^o$ open slot in the field direction which was perpendicular to the beam axis in the vertical direction. The field was measured to be uniform to $<$5% variation over the beam stopping volume. The power was 150 Watt. The scaled-up coil would have, a 30 or 40 cm diameter,a 60/80cm length, 40/60 turns, consume 250/400 Watt using Aluminum tubing and 2 water circuits resulting in less than a 10$^o$C temperature rise. The Aluminum conductor size depends on availability, ease of bending, etc.

The coil would differ from the prototype in that its sectors width of the coil would be reduced from 60$^o$ to 20$^o$ in 2-3 layers. This change would hide the coil behind the frames of the TPC (in solid angles where the data is already degraded by frame scattering). Tracks which go through these frames are already removed from the most pristeen data set in order to minimize the effects of multiple scattering on the track reconstruction. Fortunately this change to 20$^o$ sectors has been shown by calculation to have a minor effect on the field uniformity. Additionally we intend to bin the data for each part of the fiducial volume separately and analyze it based on the field strength at that point.

We plan to procure several samples of conductive tubing and determine which is easiest to wind, insulate, impregnate etc., with the available winding fixtures at LBNL. These designs seem to be quite insensitive to choice of both materials' dimensions (Copper and Aluminum) outside diameter and bore. For example we could probably go to 200 Gauss or higher if we wanted to check the field dependence!