I used to use a microfilter by millipore. You cut out the band
froze it, smashed it up b4 loading it and then spun it out and ethanol
precipitated the filtrate (UFC3 OHV 25). This was Ok for cycle sequencing
on an ABI autosequencer.
Since then I've used an ultracentrifuge with eppendorf appadters
to purify fragments in the range of 100bp and above for cloning and this
seems to be pretty cheap too.
TLA100 60,000 4C for 15 minutes, take supernatant and etoh ppt. This will
probably be OK for automated sequencing. I don't know where the protocol
came from originally just ppl in the lab were using it and it works for
Alistair Forrest (PhD student)