NCT Controller working the final position for SFO described a loss of wake turbulence separation.

Narrative

An A319 was inbound to SFO on the BSR arrival and slow to descend due to his TBM time restriction. I instructed the Airbus to fly heading of 030 and to descend to 4;000 FT. After about 5 miles the A319 was still descending out of 075 and I asked the pilot if he needed a vector east bound for his descent; the pilot said that it wouldn't be a problem so I continued his descent to 030. The A319 was told he was going to be vectored across the localizer for spacing because he was following a heavy and he was cautioned for wake turbulence. I gave the A319 a heading of 250 to intercept the localizer and noticed he was a little high; 6;000; for a 13 mile final and asked the pilot if he was able to shoot the approach at that altitude; the pilot said he would be able and I gave him a clearance for the ILS 28R approach and issued reduce speed to slowest practical speed for wake turbulence. Once I cleared him I began a position relief briefing to another Controller; the whole time watching the A319's profile on the ILS. When the next Controller assumed the position; he again instructed the Airbus to slow to final approach speed and shipped communications to SFO Tower; a mile later the Tower sent the Airbus around. When I was notified the day after of the loss of separation; and upon further review of the tapes; I saw that I used a 4 mile bat instead of a 5 mile bat to represent the minimum for wake turbulence. I also saw that when the Airbus was shipped; there was at least 5 miles which was required; however; with the A319's speed; the overtake reduced separation to 4.45 miles before the Tower sent him around. Recommendation; for this situation I definitely would vector the aircraft for its descent rather than forcing him to fly an unstable approach. I also would recommend using the appropriate size bat when it comes to wake turbulence minima.