How do you define "poor" stability? Scott is exactly right- units as the 3312 were not necessarily intended for long term stability. His suggestion of letting the unit warm up then checking is very good advice. Most test equipment is speced based upon a warm up period of usually 30 minutes. I had a couple of 3312s over 25 years ago in my design lab and they were fine for lots of the work we were doing. If you have a unit that is truly drifting badly over a short period of time after warm up then I'd say it has a problem. However, I don't know if you'll find a drift spec for that particular unit. Best I recall we (yes, I worked for HP/Agilent) didn't routinely spec drift on the low cost units until DDS came about. But one thing is sure- we were pretty much always better than anyone else with respect to most any specification you picked! Output jitter was especially bad on some competing units.

John