I haven't posted here in a while, but thought I might chime in here. I built an extremely sensitive displacement sensor to use with one of my Lehmans. Now I could see all kinds of drift artifact that I could not see before. [ By the way, the beech tree that I speculated (on this reflector) was causing tilt artifact when fully foliated...confirmed. Half of it collapsed under sticky spring snow storm in '00, so I removed the whole tree. No change last fall or this spring.] Anyway, since I had already gone to all the (excessive) effort I was willing in order to make the system mechanically stable, I decided to use a jack screw and stepper motor to move a chunk of lead back and forth on the slab to automatically cancel out drift. This worked OK, but seemed a bit crude, so I finally decided to use the force balance approach, which simply establishes a feedback loop to hold the boom in place, and takes the acceleration component from the feedback loop. It also allowed me to go with a pure digital phase locked loop detection method of extraordinary precision, since you now are only concerned with "any displacement at all" verses "accurate representation of displacement"in your primary detection loop. The result is a real time stream of digital acceleration words which you COULD clock right into the parallel port on your PC, but I, who prefer analog recording on cash register paper, instead run them through a D/A converter and integrate it in my recorder...JUST LIKE THE MAGNET AND COIL DID ORIGINALLY. So it proves that you CAN, in fact, through ample application of ingenuity and fanaticism, keep improving a system until it works almost as well as it did before... Tom On Fri, 25 May 2001 10:09:26 EDT ChrisAtUpw@....... writes: In a message dated 25/05/01, mja6042@............. writes: It seems like the thing to do then is build Lehman with a displacement sensor similar to the one found on a Shackleford-Gundersen seismometer in place of the magnet and coil on the boom of the Lehman. Hi there Mark, Drift is likely to be the biggest problem. Lehmans are very sensitive to tilt effects. The longer the period, the greater the tilt sensitivity. [edited] The normal coil assembly used on a Lehman is sensitive to the rate of change of position and isn't worried by small drifts. Have a look at:- A Force-Balance Seismometer by Karl Cunningham on the PSN Website.I haven't posted here in a while, but thought I might chime in here.= DIV>I built an extremely sensitive displacement sensor to use with one of<= /DIV>my Lehmans. Now I could see all kinds of drift artifact that I could= DIV>not see before. [ By the way, the beech tree that I speculated (on=20 thisreflector) was causing tilt artifact when fully foliated...confirmed.= =20 Halfof it collapsed under sticky spring snow storm in '00, so I removed=20 thewhole tree. No change last fall or this spring.] Anyway, since I had= DIV>already gone to all the (excessive) effort I was willing in order to=20 makethe system mechanically stable, I decided to use a jack screw and=20 steppermotor to move a chunk of lead back and forth on the slab to=20 automaticallycancel out drift. This worked OK, but seemed a bit crude, so I=20 finallydecided to use the force balance approach, which simply establishes a<= /DIV>feedback loop to hold the boom in place, and takes the acceleration= DIV>component from the feedback loop. It also allowed me to go with a=20 puredigital phase locked loop detection method of extraordinary=20 precision,since you now are only concerned with "any displacement at all"=20 verses"accurate representation of displacement"in your primary detection=20 loop.The result is a real time stream of digital acceleration words which=20 youCOULD clock right into the parallel port on your PC, but I, who=20 preferanalog recording on cash register paper, instead run them through a= DIV>D/A converter and integrate it in my recorder...JUST LIKE THE MAGNET= DIV>AND COIL DID ORIGINALLY.So it proves that you CAN, in fact, through ample application=20 of ingenuityand fanaticism, keep improving a system until it works almost as well = as=20 itdid before...TomOn Fri, 25 May 2001 10:09:26 EDT ChrisAtUpw@....... writes:In a message dated 25/05= /01,=20 mja6042@............. writes:
It seems like the thing to do then = is=20 build Lehman with a displacement
sensor similar to the one found on a= =20 Shackleford-Gundersen seismometer in
place of the magnet and coil on = the=20 boom of the Lehman.Hi there Mark,=20
Drift is likely to be the = biggest=20 problem. Lehmans are very sensitive
to tilt effects. The longer the=20 period, the greater the tilt sensitivity.[edited]= DIV>= DIV>The normal coil assembly= used on=20 a Lehman is sensitive to the rate ofchange of position and = isn't=20 worried by small drifts.
Have a = look=20 at:- A Force-Balance = Seismometer=20 by Karl Cunningham
on the PSN = Website.=20
Regards, Chris Chapman=20[ Top ] [ Back ] [ Home Page ] Larry Cochrane <cochrane@..............>