Instrumentation for local events (frequencies > 1 Hz) is easier but IMO less interesting than the teleseismic band below 1 Hz. With even a rudimentary long-period seismometer you can see magnitude 6+ events in the same hemisphere and magnitude 7+ events worldwide, learn how the different wave components propagate, how we know that the Earth has a liquid core, etc. As a bonus that frequency range is less susceptible to local noise sources. The stuff you have to watch out for is not so much cars and footsteps as things like wind, or sunlight hitting the concrete slab you've placed the seismometer on.
For a geophone-based solution (mainly high frequency but maybe some teleseismic capability too), I recommend checking out the Raspberry Shake product.
The classic amateur long-period design, of amazing simplicity, is Lehman's "garden gate" seismometer from the Amateur Scientist column in Scientific American, July 1979. Speaking from experience, this is an excellent gateway drug

For a really serious, vertical broadband design with force feedback, google "FBV seismometer"; sadly there has never been a proper web site designed for this but several of the builders have nice pages describing theirs, e.g.
http://www.groundmotion.org/main.html. The "Yuma" is the most recent design for home builders and competes favorably with professional instruments costing tens of thousands of $$. It has a flat response from 20 mHz to 30 Hz. From a location in Seattle, mine sees most M5+ anywhere on the Pacific rim of fire, usually several events every day. At this level the limiting factor is not the instrument but the site. It helps if you are in the middle of a big continent to reduce microseism noise from the oceans, as does placing the instrument in a temperature-stabilized vault cut into bedrock. OTOH mine sits on a concrete basement floor atop hundreds of feet of glacial till, and I'm happy with the results.