Barry,

Regarding regular periodic bumps in the output of your larger VBB
vertical. You mention that they occurr at 5 minute intervals, but
you don't say what the period of the event is. So I will assume that
it is a single sine (-like?) wave at the operating period, like
60 seconds. This is usually the result of a step in displacement,
or a simulation of such in the electronics.

To isolate the noise of a seismometer system the obvious method is
to clamp the boom or mass so it cannot move and produce real output.
This would separate the source possibilities between electronic and
mechanical, where mechanical includes the sensor, pier, cover, garage
floor concrete slab contracting because it is cold, and the thermostat.
Of course, if you have and adjacent seis that doesn't sense the problem,
most of these mechanical/environmental sources are eliminated.
And if you have two VBB sensors with identical electronics, you can 
interchange them.

Clamping the mass is easy to do with a velocity (moving coil) output
system. But any seis with a displacement detector is another matter.
Clamping the boom will usually position the detector at an extreme of
its range and its electronics will saturate (lock up at a supply voltage).
This is ususally not useful since the output is directly coupled.

In the triple feedback VBB that you are using, the displacement detector
IS the source of the broadband output WHEN the feedback is connected.
I don't recall if you are using the VRDT or a capacitive bridge. But
clamping a boom so that the detector is at "zero" is quite difficult,
especially when "zero" is +,- 0.1 micron, and the detector+amplifier
output is 4 volts/micron. Using shims to wedge the sensor to center
can create a wide variety of noise, most of which will take days to
settle to the micron level.

Another test is to remove the sensing vane of the VRDT, which should
balance the reactive bridge IF everything else is balanced. This latter is
difficult to achieve. I just made another "batch" of VRDT sensors, and
after starting with 25 miniature transformers, I ended up with 6 pairs
that matched within 0.5 ohm DC, and only one that matched within 1 ohm
reactively. Of course, the change caused by the sensing vane still allows
a working linear range of several hundred microns at 40 millivolts/micron.
For a capacitive sensor, the sensing plate is the output, and can be
substituted with a pair of small fixed capacitors if they can be matched.

For testing the electronics box, I have an adapter that fits in parallel
with the DB-25 connector to the seis that replaces the VRDT coils with
their reactive equilavents, namely 365 ohms. The resistors are matched
within 0.1 ohm with a meter from a surplus handful of 1% values. Their 
noise is predominantly thermal, and they produce about 10 times the noise 
of the earth background. The advantage of this parallel substitution is 
that the it avoids disturbing the seismometer.

I don't know if any of this will help. If you can separate the problem
into either mechanical or electronic, it will help. I don't know if
the tantalum capacitors can produce spikes if they are operated well
within their voltage ratings; and I assume that your DC power supplies
are well regulated.

Regards,
Sean-Thomas

_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)