That sounds like the POES plots.
http://www.sec.noaa.gov/pmap/
Note that they are actually predicted auroral visibility levels based on proton flux density from one pass of a polar orbiting satellite, they are not actually satellite photos. See this for info:
http://www.sec.noaa.gov/pmap/Intro.html
http://www.sec.noaa.gov/pmap/html/200504051034S16.html
These are real images (far ultra violet), but are not updated as frequently:
http://sprg.ssl.berkeley.edu/image/latest_wic.html
STD am will download all these images for you automatically. It's a bit of a heavy data load for a dial-up connection though.