As I finished adjusting my 8596E precision reference to precisely 10MHz (via time delta between ref-out and a rubidium reference), my bubble was burst when the tracking generator output (in zero span) at 10MHz was 10.02 something MHz.
I could have sworn that my old E4401A (1.5GHz analyzer) was "spot on" and naively assumed that the tracking generator was synthesized and locked to the reference. I've been informed by users on another forum that the TG may be driven by the LO and I've also noticed dependency on RBW.
Feeling bad, I looked through drift specs on many HP / Agilent analyzers (85xx, ESA, PSA) and for the most part, it is 1.5kHz in 5 minutes. Now I don't feel so bad but the TG is not going to be a precise source for me (being on the cheap side).
It would be a big help if someone could describe the relationship between the frequency accuracy when in zero span, and the other SA parameters which impact this. I realize that the tracking adjusment function probably is adequate for most intended uses.
Thanks,
Den
I could have sworn that my old E4401A (1.5GHz analyzer) was "spot on" and naively assumed that the tracking generator was synthesized and locked to the reference. I've been informed by users on another forum that the TG may be driven by the LO and I've also noticed dependency on RBW.
Feeling bad, I looked through drift specs on many HP / Agilent analyzers (85xx, ESA, PSA) and for the most part, it is 1.5kHz in 5 minutes. Now I don't feel so bad but the TG is not going to be a precise source for me (being on the cheap side).
It would be a big help if someone could describe the relationship between the frequency accuracy when in zero span, and the other SA parameters which impact this. I realize that the tracking adjusment function probably is adequate for most intended uses.
Thanks,
Den
As for the 8596E, their tracking generators do track the instrument's 1st LO, which is locked to the instrument's frequency reference. But, there is also an unlocked oscillator within the tracking generator assembly that is mixed with the instruments locked 600 MHz signal, then mixed with the instrument's 1st LO. So the tracking generator output is thus unlocked as well. This is why there is a tracking peak function on the tracking generator menu, as it will adjust the unlocked oscillator frequency so that the tracking generator output will lineup with the current input frequency, giving the largest possible response. (see attached)
Best regards,
Mike