Inexpensive M/S setup for experimentation?

["Steve Duncan" swduncan]

It seems like the barriers to making one's own stereo boundary mic out of c=
apsules are fairly low - sites like Rob D's & Curt O's cover the basic conc=
epts, and the components are inexpensive and fairly easy to find.  But for =
M/S it seems much less simple, which got me thinking - what would be the ea=
siest/cheapest way to experiment with M/S, starting with a handheld recorde=
r and no mics? What are the options?

The cardioid part seems the easiest - there are several choices. There is a=
t least one free plugin for decoding M/S on the computer, but a Rolls MX310=
 or the like could do it in the field, using a cable with two connectors wi=
red out of phase at one end, right? That leaves the figure-8 mic, which see=
ms to be the hard part as there aren't many inexpensive options.

Has anyone been doing this without a heavy investment?