Agilent 33220A 20 MHz Function Generator

[jason123]

looks like most ICs have some marker / dots on them. I took some photos below. I wish I were 20 years old.. no, this 50 plus can hear it. Seems to change with frequency output - I hear it most prominent at 650 Hz and up.. as you change the output frequency, it's like you're playing a musical instrument.

[CapLeaker]

Look at the tab of Q1301 and Q1302 and some other components around that area... almost like Agilent was saving on solder?

[jason123]

Well, this is a bit anti climactic / bitter sweet - so I had taken the transformer off the board to replace it - a replacement is not available, this is custom transformer specifically made for Agilent. In order to get the plots R_J asked for, I soldered it back on, and miraculously the audible noise seems to have disappeared! I still got the scope captures at the output pins of the transformer, reference to pin 8 of the transformer "ACOM"
Seems to have been solder joint issue? Does anyone have other suggestions? Thank you for the help folks - great team as always. hover over the plots below as the name indicates where it was taken

[stj]

a fix is a fix.
in the old days we used to resolder all transformers and heatsink-mounted parts before even thinking about the fault.
cracked solder was that common

[CapLeaker]

Yes! That could have been an issue! Just not enough solder and with all these years of heating and cooling or a little corrosion they can cause weird issues. Basically anything that gets hot and the transformer. Just about the same we used to treat old stubborn TV’s. Call it a win! To me it just really looked liked Agilent was skimping on solder. My next step would have been to resolder all these components with 60/40 solder and see what that gives.

[jason123]

Yes! That could have been an issue! Just not enough solder and with all these years of heating and cooling or a little corrosion they can cause weird issues. Basically anything that gets hot and the transformer. Just about the same we used to treat old stubborn TV’s. Call it a win! To me it just really looked liked Agilent was skimping on solder. My next step would have been to resolder all these components with 60/40 solder and see what that gives.

Not too late, I can still do that.. Thank you all for the help

[jason123]

a fix is a fix.
in the old days we used to resolder all transformers and heatsink-mounted parts before even thinking about the fault.
cracked solder was that common

interesting - Lesson learned! Thank you