[Sneap] Fast closing gate-valves

[Vincent Needham]

>
> Do any of you have such valves on your system?  
>

Back when we put in the KSU superconducting linac, we decided that it
would be prudent to protect it and its cold surfaces from both the
Tandem and the beam-line sides, especially given that we mostly have
students working those things. The cost of a commercial system was
prohibitively expensive, so we built our own.

The valves themselves are pneumatic angle valves with beefed-up springs
and a short rod extending from the piston through a small hole in the
housing. A catch  hooks a notch in the rod; applying air to the piston
lifts the rod out of the housing, the catch engages, making the valve
cocked in the open position. 

The sensor is a simple spark gap placed as far up- or down-stream of the
protected system as possible. A circuit monitors the gap; if a vacuum
accident allows the pressure at the sensor to reach around 1e-5 torr, it
will spark, and the circuit fires a high-voltage pulse that fires the
squib on the valve catch. The catch releases, and the springs ram the
valve shut. The circuit also triggers the normal automatic vacuum valves
to go to a safe state.

Our sensors are typically about 40 feet from the valves, with a variety
of 2-to-4 inch beamlines and apertures between them. In testing and
accidents, we typically see transient pressure rises to around 1e-5 torr
in the protected system. 

Fortunately we have had few occasions where this system was required to
work!

This question has come up here before, so you might want to search the
archives for more information.

-- 
Vincent Needham - http://jrm.phys.ksu.edu
JR Macdonald Lab, Physics Department
Kansas State University
Manhattan, Ks  66506-2604