I've had a DSO-2250 9-bit/250MSps/60MHz scope for a decade, bought "new", and it's always had a large ripple voltage (about 1/3 of a division) overlayed on any signal. The ripple sometimes made the DSO barely usable. At some stage I put a filter on the USB cable inline with the power; this didn't work at all.
I finally got around to probing (with its own probe) around the rail voltages inside, which are easy to find because they have large capacitors on them, and saw about 10% more noise on a -5V rail across a 22uF capacitor. One of the inductors in the nearby SEPIC converter had 1/4 of its ferrite sheild broken off (at factory), but didn't appear to cause issues. This board has a lot of evidence of rework.
Connecting a parallel 100uF cap improved things:
Connecting an additional parallel 470uF cap improved things further; the ripple is now competing with the ADC noise (i.e. ± a couple of bits):
I didn't see any notes about this when I looked for a solution years back; just posting now in case it's of use or interest to someone, however unlikely.
I finally got around to probing (with its own probe) around the rail voltages inside, which are easy to find because they have large capacitors on them, and saw about 10% more noise on a -5V rail across a 22uF capacitor. One of the inductors in the nearby SEPIC converter had 1/4 of its ferrite sheild broken off (at factory), but didn't appear to cause issues. This board has a lot of evidence of rework.
Connecting a parallel 100uF cap improved things:
Connecting an additional parallel 470uF cap improved things further; the ripple is now competing with the ADC noise (i.e. ± a couple of bits):
I didn't see any notes about this when I looked for a solution years back; just posting now in case it's of use or interest to someone, however unlikely.