Monday, June 16, Notes Present: E. Barsotti, A. Byon, R. DeMaat, G. Drake, J. Elias, S. Hahn, S. Orr, K. Schuh, A. Tollestrup Power Supplies Collecting data sheets and specifications is on-going with the idea of acquiring candidate supplies from 4 vendors for evaluation. Some observations by Stan: switchers trade size for noise, the filters are large objects vendors mislead, quiet modules are not parallelable to the current we want but the noisy versions are - difference between 5 mV peak and 70 mV peak supplies will operate on both 60 and 400 Hz mains 3-phase versions presently fail a FCC criterion for ground currents Magnetic fields and RABBIT A & B Supplies Tests in the village with the 400 Hz geberator still there. Gary observed magnetic fields which had essentially no load dependence -> transformer leakage. There was a decrease as the primary voltage was lowered with no evidence for core satyuration, just stray field. A simple mu-metal shield wrapped around the transformer cut the field by a factor of 6. The rf choke then stood out as the next level guilty party. The next step is tests at wedge 21 using the strip chamber pedestal width as the diagnostic. Recycling the supplies will require all new filter capacitors at ~$100 per unit. Gary agreed to a proposal that 12 supplies be modified now and burn-in testing be started. Digital Links Working Group LVDS has about the same ratio of common mode disturbance to signal as does ECL, and the swings are smaller so the em transients are smaller too. The shield termination reduces the transient by about a factor of 2. Using a transformer at the driver gives another factor of about 2 but degrades the risetime noticably. Charlie claims overload and asks for relief. Tests with the strip chamber on the wedge for a link with LVDS terminated pairs & em termination of the shield and the noise blaster driving 8 pairs showed no effect. However, Gary said the RABBIT was being operated in correlated double sampling mode. Plan is to redo this with the RABBIT single ended.