Isolating the pressure sensor may be a good idea. Even if a bubble won't work, perhaps a long tube that is pressurized at the top with a inert gas that slowly leaks out the bottom. There is probably a way to simplify turning the inert gas reservoir on and off. What pressure does the gas pressure sensor operate at ? Gus > On Feb 19, 2010, at 3:59 PM, Carl Denk wrote: > > Don't think a bubble is applicable. The natural gas well is a 1000 > feet > deep, 6" diameter hole mainly in shale stone. The natural gas seeps > out > of fissures and porosity of the shale. In general the hole is dry, but > there could be water near the bottom. The top 160 feet of hole are > lined > with steel casing with a coupling and plug 3 feet above ground > level. A > 1.25" steel pipe is welded to the side of the casing, and there are > some > valves, and a tee fitting for the pressure sensor, and a visual gauge. > There is no liquid to bubble. :) > > Marcel Duchamp wrote: >> It sounds like you are not quite sure as to whether the failures are >> caused by ESD or damage from incompatible media. >> >> One type of solution that could make both cases better but is more >> complex would be to use a bubbler system. This prevents contact with >> the media and can help to isolate the sensor electrically as well. >> >> Basically, you hang a conduit (tygon tubing, or whatever) down into >> the >> location to be measured. The top goes to a 'T' fitting that splits >> out >> to your pressure sensor and a pressure source. The source can be any >> supply of air pressure that is slightly higher than your full scale >> pressure to be measured. The air pressure line has a solenoid in >> it so >> you can turn it on to purge the line, shut it off, wait a bit then >> take >> a pressure measurement. The air pressure in the tubing will bubble >> out >> the bottom (hence the name bubbler). When it turns off the system >> will >> return to static pressure to be measured. -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist