No abstract

No abstract

AbsttactAn advanced, high-performance BiCMOS technology has been developed for fast 16Mb SRAMs. A split word-line bitcell architecture, using four levels of polysilicon and two self-aligned contacts, achieves a cell area of 8.61prrP with conventional 1-line lithography and 7.32pm2 with 1-line phaseshift lithography. The process features PELOX isolation to provide a 1 .Opm active pitch, MOSFET transistors designed for a 0.80p.m gate poly pitch, a double polysilicon bipolar transistor with aggressively scaled parasitics, and a thinfilm polysilicon transistor to enhance bitcell stability. A quadruple-well structure improves soft error rate (SER) and allows simultaneous optimization of MOSFET and bipolar performance. IntroductionFast. hiah density static RAMS rewire both a small bitcell area an( a high'performance piocess. At thel6Mb generation, a bitcell area of less than 9.Op.rn2 can be achieved using 1-line technology with multiple layers of polysilicon and self-aligned contacts. Use of phase-shift lithography provides for even smaller bitcell areas. A symmetrical split word-line bitcell and thin film polysilicon transistors serve to enhance cell stability. Speed requirements can be met by adding a high performance, double polysilicon bipolar transistor. With careful design, bipolar parasitic capacitances can be even further reduced without increasing process complexity. With the aggressive scaling of bitcell areas at the 4Mb generation and beyond, SER has become a primary reliability concern for SRAMs. In this technology, a quadruple-well has been developed to provide diode isolation for the memory array while allowing for simultaneous optimization of MOSFETs and bipolars.

logical conclusion of this process. Yet it is called a "new" turbine.Looking ahead, I expect to see for this type, before it is finally established as a standard, an entirely new type of gate developed. The swivel gate, while obviously correct in connection with the inflow turbine, is illogical when applied to the axiflow type. L. A. Barnes: In regard to the 2.5 in. which Mr. Wilkins took off the runner. I assume he took that off in small amounts at a time. If he did not I would like to ask how he determined that amount? And, if he did take that off a little at a time I wonder if he could tell the success, that he had in the successive cuts, that is, whether it was progressive or whether he arrived at it all at once. R. Wilkins: In answer to Mr. Stauffacher, Curves No. 5 andNo. 6 were taken on identical scroll cases with different runners and under the same head on the same pipeline. On No. 4 there was sufficient noise to make talking difficult, that is, the hum was so loud that you had to raise your voice to be understood in the power house; this gradually decreased up the pipeline yet you could hear the hum at the penstock. With the seventeen vane runner in an identical scroll case just beside it ο α the same pipe line, there was no noise that you could hear at all. There was no vibration that you could feel with your hand, yet the hydraulic pressures inside the pipeline were just as violent as those in the other unit. Both these, however, are considerably smaller than some of the other curves. Fig. 13, for instance, gives no hum; it is not regular enough to give a hum and it didn't give an actual noise, yet that unit vibrated so that we were afraid to run it for

No abstract