There seems to be lots of anecdotal evidence in the responses, and a few technical problems with the measurement setup of his original post. Is he following any one of many documented procedures here?
His 400MHz oscilloscope can't properly measure 390MHz with a 400MHz probe, as an example.
He should be calculating his system bandwidth, and would require a probe bandwidth 2-3 times greater than his signal under measurement to be effective.
I would expect there's other free documentation on this, and HP was particularly good on these older bench briefs, etc. Tektronix a bit less so.
This could be concluded if the procedures were referenced.
https://download.tek.com/document/60W_18324_0_0.pdf
