In a message dated 12/01/00 02:41:46 GMT Standard Time, danieo@............ 
writes:

>> A ball rolling on a flat is fine for the lower pivot point but my ball is 
attached to the mast and the flat is a small polished hardened steel flat 
attached to the pendulum. This way, the pendulum is less sensitive to the 
position of the ball on the flat. The flat is made from the handle of a small 
flat file which was also hardened.
 Cheers all Danie Overbeek.
 
Jim
    I have been reading your mail with interest. I read the reply submitted 
by Frank and agree with his reply as well as the others. Eleven to twelve 
seconds is about all that
you can expect out of the detector as it is constructed. I could not get any 
more out of mine and it was built the same way.
    The time period is controlled by the upper and lower pivot point. Even 
the smallest amount of friction or resistance at these two points is 
critical. We are talking about extremely small amounts.
    To lower the friction at the lower pivot point, cut the end off flat, 
drill a hole in the end of the pendulum large enough for a 1/4" ball bearing 
to fit in. About half of the bearing protruding from the pendulum. The 
surface that the ball bearing rests against should
be sanded with no: 400 paper then with no: 600 paper. The smoother the 
better. Absolutely no holes drilled to keep the pendulum from migrating. The 
weight of the coil and the lead weight will keep it in place. It will never, 
never slide out while it is in operation. It will slide off sometimes while 
it is being adjusted. Put a drop of super glue in the hole before inserting 
the ball bearing. If you like, put a little more around the edge of the ball 
bearing after the first has hardened.
    The top pivot point. I use .015 music wire for the lower section up to 
the turn buckle. The upper wire from the turnbuckle to the top pivot point is 
..010 thousandth. are you with me? the best source for the wire of the correct 
diameter is, your local music store. The e string of a guitar is used. That 
is the first small string. Cost 50 cents to a dollar. The diameter of the 
strings are written on the package. The smaller diameter on the string at the 
top the longer the period.
    Using this method I can get any time that I desire. I usually run mine at 
about sixteen seconds. Twelve seconds or even less is OK. I think that most 
professional detectors run at fifteen seconds. Any higher and you will be 
adjusting all the time .
    PS I have four detectors up and running at this time.    
                                                  
                                                     John Cole

    Great contributions!! 
    Let's think them through further. We need to have the vertical hinge axis 
fixed, so that the seismometer level can be trimmed easily and also so that 
it can be disassembled and reassembled without too much difficulty. Attaching 
a ball bearing to the top of the fixed vertical frame and fitting a rigid 
flat load plate on the suspension should do this well enough. The load plate 
can be marked with the approximate correct position. With a ball bearing in 
the end of the horizontal swinging arm and a flat plate on the frame, this 
lower point of rotation is NOT fixed laterally. Would it not be better to fix 
this by mounting the flat plate on the end of the horizontal arm and fit the 
ball bearing into the frame?
    I suspect that one reason for difficulties in setting up a Lehman for 
long swing periods may be that the 'hinge axis' of a flexing suspension wire 
at the top of the frame may change as the wire flexes through different 
angles. It also assumes that the wire can't move in the clamp jaws. You 
really need both jaws to be aligned vertically, so that the wire flexes at 
right angles to the jaws. If you use identical lengths of the same wire for 
the top and bottom flexures as Sean suggested, there should be no problem. 
Likewise for rolling contact (John Cole). The rolling resistance of a ball on 
a plate may depend on the surface finish of plate. A small disk of polished 
sapphire (optical stockist / ex IR sensor window?) would certainly be OK, but 
a small glass flat (part of a microscope slide?) glued on with epoxy should 
also work. Hard glass or polished Quartz or Agate would probably be better. 
Agate flats and triangles are used in some Laboratory Balances and sometimes 
in 'grandfather' clocks.
    We also need to consider the long term operation of the system. 
Unlubricated rolling contacts (such as steel bearings) can corrode (and 
probably will in time). Lubricated corrosion protected contacts can attract 
dust and 'particulate matter' and may need to be covered. You can get 
stainless steel ball bearings (from a bearing stockist) and sapphire spheres 
are also made (they are used in some vertical tapered column gas flowmeters).
    Another alternative is to make a crossed shim flex suspension for both, 
but making these may present rather more difficulties for the amateur 
constructor.  
    I hope that these thoughts may be helpful.   
    Regards Chris Chapman

_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)