Tony,

I don't think there's anything wrong with your 3312. The numbers you've provided, however, don't actually mean a whole lot without specifying exact test conditions. Product specs are provided at a specific temperature(+/-), input power, etc. And, we've not even addressed the issue of the last calibration/service date of the unit. But, for the moment let's assume you are testing under the correct conditions and that the unit is in a calibrated state. After allowing for a 30 minute warm up (and it may actually need to be an hour) the frequency drifted less than one half percent over the next 30 minutes according to your numbers. Considering we're discussing a function generator that was designed in the early 70s using technology available at that time I'd say what you're experiencing is pretty damn good! Again, HP products were considered by the vast majority of those who knew what they were talking about :-) to be the best- the "standard". And the 3312 was an excellent, state-of-the-art, best-in-class product when it was introduced. But it's a function generator, not a high stability, low phase noise RF signal generator.

If you're looking for incredible frequency stability then you are using the wrong piece of equipment. Maybe you should consider a newer DDS unit, or probably more likely a true RF sig gen. If you're wanting a product with good spectral purity and lots of functionality then the 3312 is as good as you're going to find for products that are going on 40 years old as it is. And I bet it will even beat several products introduced way more recently.

John