Here is some email I forgot to post regarding 
some specific questions that I haven't answered yet.

Bob L. has asked about extending the effective period of a "baby"
Benioff seismometer using either the equalization technique of Roberts
or a VBB feedback method. An advantage of the "baby" Benioff 1 second 
seis is that it has a 14.402kg (for Z) mass. A disadvantage is that it 
is a variable reluctance transducer, so the inductance of the coils is 
a major problem. The seis has 8 coils of 125 ohms each, so it can be set 
up with the proper series/parallel combination to match almost anything,
including the 80 ohm, 0.75 second galvanometers of the WWNSS system.
For this a net DC resistance of 65 ohms is realized, which has an 
inductance of 5 henries. This puts the inductive pole (Rs+Rd)/L*2*pi
at 4.5hz, or within the passband of the data. This inductance presents 
a major phase delay problem for a fedback system using the coils.

For electronic recording the coils are connected all in series for
1000 ohms and overdamped with 1200 ohms in parallel. The inductance 
of this configuration is measured at 8 henries (mutual inductance 
of the individual coils makes the various combinations incalculatable).
This moves the inductive pole to about 44 hz.  For a feedback system,
this is still a serious inductance, and the high coil resistance makes
the transfer function unstable at higher frequencies. 

For this reason, I have not attempted to install a displacement
transducer on one of my Benioffs so as to try a VBB feedback. But
I see no reason why an equalization circuit wouldn't work to improve
the long period response to 10 seconds or longer. To really make an
effective improvement for teleseismic recording, 20 seconds or longer
would be needed, which would really push the amplifier noise requirements,
particularly the 1/f noise. 

_________________

On a previous topic regarding brass and stainless steel shim stock:

Chris has suggested a readily solderable stainless steel alloy, EN58.

There is possibly the rub, as it were. McMaster's shim stock is a 303 alloy, 
which McMaster says meets the  specifications of AMS 5515, SAE 30302, 
Fed. Spec. QQS-766, and ASTM A177. It makes no mention of EN58 (which
may actually be included in one of the given specifications.) 

The major difficulty of soldering thin/fragile assemblies and using a 
very corrosive flux is that one needs to use an electronic solder
pencil to control the heat so as not to greatly modify the temper (as
a micro torch would do), and I don't want to contaminate the tips I 
use for electronics with a corrosive flux, so they have to be carefully 
separated and cleaned.  McMaster sells a variety of fluxes that I have 
tried, (and may try again, since I have managed to break a number of the 
box flexures that I made with 0.002" brass.)

Also:
George H. has suggested using a heat-treatable beryllium-copper alloy
that can be heat treated to restore a hard temper after forming in a
soft condition. This would make it a spring, which of course is needed
in a vertical sensor, and I believe that some of the diaphram-style
spring/flexures, such as in the L4-C, are made this way. But for an 
application as strictly a flexure in a fedback system, the ideal flexure
will have NO restoring force, only positional stabiltiy, such as a 
pivot (which unfortunately have large micro-positional noise so they
are not used). Heat treating might allow a thinner material to be used
for the same strength as a thicker soft material with a minimum restoring 
force, so some calculations are needed.

Regards,
Sean-Thomas
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)