HP 3314A sine-wave short discontinuity

[the Chris]

I've recently observed that my HP 3314A exhibits a slight discontinuity between the end and start of a new sine-wave period. This is also visible in the FFT as sidebands at 1/2 and 3/2 of the main frequency. The screenshots show a comparison with a proper output from an UTG1022X(-PA).

Did anyone have had something similar and located the origin or has a guess where to look at first on the PCBs?

[wn1fju]

I have had two HP 3314A's.  One exhibited what you are seeing and one did not.  Guess which one I kept?  None of the calibration adjustments got rid of the one with the distortion.

Take a look at the triangle waveforms and see if there's a "blip" at the corners.  The sine waves are generated from the triangle waveforms via a custom hybrid IC.  I would think if the integrator generating the triangle waveforms isn't perfect, the "errors" will propagate through as distortion on the sine waves.  Then again, perhaps there were slight differences in the precision of the hybrid sine-shaper IC that HP produced.

There has been a discussion on eevBlog about ringing on the triangle waveform for the 3314A.  People were speculating that it was a mix of board layout and parasitics that affected the capacitance in the integrator. 

[the Chris]

Thanks, makes sense. Seems to be a good starting point. I'll check the triangles, but I think I saw something there as well, just not so obvious.

BTW: love your website, I visit it regularily.

[MathWizard]

So looking at that nice old machine
https://www.eevblog.com/forum/testgear/hp-3314a-function-generator-teardown-explanation/

So say something like the "hybrid sine-shaper IC" really is faulty. With all the advances in IC's, and all the fast FPGA's and A/D chips and whatever else is off the shelf.....how hard would it be to homebrew some replacement for chips in these older digital machines ?

Even if it's a lot of work and money, I bet there's people that would do it for fun.