I am a newcomer to your group and like several of you have just purchased the DSA815-TG spectrum analyzer, for me a replacement for a 7+ year-old Instek GSP-810; that I was fortunate enough to be able to sell on eBay for US$1300, greatly subsidising the cost of the DSA815. I find it to be a very capable instrument, far superior to the Instek which cost $2500 when new.
Also like another poster I too have a rubidium standard and thought I would post my findings.
Here's a shot of the test setup:
And the 10 MHz reference at 2 kHz span:
You can see that the peak marker reads 9.999973 MHz, 27 Hz low, however in calculating the instruments marker uncertainty according the for formula presented in the specifications:
± frequency indication × frequency reference uncertainty + 1% × span + 10% × resolution bandwidth + span/(# points -1)
frequency indication = 9999973 Hz
frequency reference uncertainty = 1e-6 (it's brand new)
span = 2000 Hz
resolution bandwidth = 100 Hz
# points = 601 (from the instrument specs)
I get: 9999973 * 1e-6 + 0.01 * 2000 + 0.1 * 100 + 2000 / (601 - 1) = ± 43 Hz
, making the -27 Hz error well within spec.
I tested several other span/rbw combinations and found the marker readouts to all be within spec...
Swallowtail333, I would be very interested in how well Ultra Spectrum works with the '815; please let us know.
I have also contacted Rigol to ask if it would be possible to add an "inverse" display function in any upcoming firmware update as I feel a "black on white" display would be much easier on the eyes than the black screen--which I find to be an odd emulation of one the technological limitations of CRT technology.
This would be much easier on my eyes:
It reminds me of the control software provided with a DDS signal generator I obtained some time back, the publisher went so far as to create virtual legacy style frequency range selection buttons--on a fully digital instrument?: