Why a Vault
A Force Balance Horizontal seismometer was designed and built along the design concepts of Dave Nelson's and Brett Nordgren's Inyo Force Balance Vertical seismometer. The FBH was installed in the basement wine cellar next to the 2 Lehman and FBVseismometers. the FBH was covered with an insulated box to cut down on air draft movements affects on the seismometer. A typical 24 hour response is shown in the figure.
This is completely useless. The only time when there is a stable trace is when we are sleeping and not walking around the house. Plan B is required. How about moving the FBH outside and away from the house? A search on the internet found a very useful video showing geophysicist Anne Sheehan from the University of Colorado installing a new seismic vault out in the field near Brush CO. The installation used 42 inch diameter plastic storm drain pipe. In my situation I could use a pipe with maximum diameter of only 36 inches in diameter. Locally I could purchase storm drain pipe for about $38/foot. 
However on New Year's Day I was walking past a near by construction site that was using what looked like the kind of pipe I needed (maybe they may have leftovers). The next day, a holiday, I stopped by with a tape measure to check out the dimensions. Sure enough, the pipe was 36 inches in diameter. I stopped by when construction started again and spoke to the manager. He said to come back the next day and they may have some extra. Sure enough, they had a 7 foot section left over and I could have it. I rushed home and to get the station wagon and the manager even had a helper to assist getting heavy section on the car roof.
The only suitable location for the vault was in the center of the backyard redwood deck. The deck was built with joists on 24 inch centers. So I removed a section of the deck approximately 4 x 4 feet in size. the opening was then re-framed so that the removed section could be put back in place and removed when vault access was required. See figure.
Notice the 2x4 at bottom of figure with a bracket attached on the end. That is the tool used to lift deck section slightly in order to get a hand hold and lift up and slide it to the side. A contractor was hired to dig the hole and remove most of the dirt. A jackhammer was required to dig the very hard clay. 
Almost hard as rock, but something to worry about, it did crumble and I could detect that it was just a little moist.
A form was built for the concrete slab at the bottom of the vault. The upper end of the picture points north. A short section of the storm drain pipe was used to access to the bottom of the vault without stepping in wet concrete. The slab is 6 inches thick.
The concrete used a mix of 50/50 II-V cement and sand and no rebar. The slab did not contact the sides of storm drain pipe. Concrete was mixed and dropped in place. Notice the concrete vibrator of to the right in figure.
Ladder was used to access the plywood platform and was then it was lifted out of place in order to have working room.
A temporary 'A' frame was built to allow placing the storm drain pipe in place. A come-along was used to lower the pipe in place.
The pipe is 51 inches long and when in place, top edge is just slightly above the ground surface.
Dirt was back-filled around the outside of the pipe and packed down frequently. After installing the storm drain pipe and covering up the vault using the lid shown in upper left, moisture was noticed on the upper area of the vault. Something had to be done to seal off the exposed dirt and the vault bottom. Two-part closed cell polyurethane was mixed and poured in place and up the same level as the concrete slab. No excess moisture has been detected after adding the foam. The cream colored foam can be seen at the bottom of the vault. A shelf was installed in the vault for holding tools, etc. while servicing the unit. A service ladder was built out of 1/2 inch rebar. Not shown in the picture is the 2 inch thick styrofoam installed on the side of vault at the bottom. Initially the foam was 2 feet high from the vault bottom. Recently 12 inches of the foam was removed, starting from the bottom. The foam was taking up too much space. In addition a circular section of foam is used to cover the bottom half of the vault. The idea is to insulate the bottom of the vault area from temperature changes from the outside. The outside lid cover is also insulated with the closed cell polyurethane foam.
A 3/4 inch conduit was installed from the house basement to the upper portion of the vault. Seven RG179 coax cables, a twisted pair, and a CAT5 cable was pulled through the conduit.
The electrical interface is shown in the figure. The panel shows 7 RCA type connectors for seismometer data, 2 connectors for DC power and a phone type connector for device that monitors temperature and humidity. Typical vault temperature is about 56 degrees Fahrenheit (mid April) and humidity is 71% and dropping as long as the cover is in place. A box of silica-gel is used keep the humidity under control. temperature is very stable and only varies about 0.2 of a degree during 24 hours. The Sonoma FBH seismometer was moved to the vault on April 9, 2012. The silica-gel is replaced every 3 to 4 months with freshly dried out material.
The 2 inch thick styrofoam insulation can be seen in the lower section of the vault. When closing up the vault, 2 inch thick styrofoam sections are placed on top of the side insulation.
Plastic sheeting is placed on top of vault lid in order to keep the surrounding area dry. A wet/dry vacuum is used to suck up rain puddles that may have collected on top of the 10 x 10 foot plastic sheet.
Seismic Date in the Vault
The initial installation did not use a box enclosing the unit in order to cut down on the affects of convection air currents. As one can see air currents have a large effect and this is with the Y-Scale set to 40, whereas, normal is between 2 and 5. The enclosed air volume is about 14 cubic feet and this must be just too much for the seismometer to handle. Insulated enclosures were built using 2 inch thick styrofoam and used in wine cellar to enclose both the FBV and FBH seismometers. However there is not enough room at the bottom of the vault for such cases. A new case was built using 1/4 inch thick polycarbonate. The clear plastic allowed one to observe the PCB LEDs and monitor operation. The dimensions were such there was about 1/2 inch clearance around the seismometer. The lower figure shows how important such a case is to correct operation.
Now we have a response that looks reasonable. Y-Scale is down to 5. The 2 quakes shown are the Valparaiso Chile and Eastern New Guinea shakers recorded on April 17th. Notice the quiet traces recorded before the local winds come up. The Force Balance Horizontal seismometer is no longer sensitive to people walking around the house (though sensitive to walking on deck and surrounding areas) but the vault is now closer to 3 close by trees and making the FBH very sensitive to extra noise when the winds pickup. The conclusion, build a Force Balance Vertical Seismometer unless you very far noise sources such as nearby trees or people walking around.