DSA815 Autotune Accuracy

[jstarr]

Both my DSA815 and DG4062 are enabled with external reference (Thunderbolt GPSDO 10Mhz freq standard).  On getting familiar with the DSA815 i noticed that regardless of the input frequency from the DG4062 that the Autotune function always sets me to a RBW of 300kHz and Span 10Mhz and the CF or Peak-CF leaves me always off by exactly +16,666 kHz.  This is true for any frequency from 10-150mhZ.  I know the Freq readout accuracy requires sufficiently narrow Span and small RBW but the Autotune error is so consistent I wonder if there is an internal calibration to improve it?  I have run the auto-calibrate without improvement and the equipment has been on for hours.  This +16666 error is consistent from day to day... so is there some internal adjustment possible?

Happy New Year to all.

john

[iDevice]

Same here, don't worry.
It probably has something to do with the step width used to scan the full span when the autotune function searches for the highest peak.
This search has to be performed quickly so these steps are quite coarse and probably the result of a simple binary counter, converted to floating point.
Autotune is meant to "find something to display quickly" not to make any accurate measurement, it just put you approximately there and then it is up to you to decide how you will setup the analyzer's parameters to best suit your intended measurement.
Same thing for the span and RBW, these default settings are used whatever autotune finds probably because they are the most susceptible to display something meaningful.

So, it's pretty much normal and has nothing to do with calibration error.
Besides, if you just do a peak detection marker, you will see that it is spot-on the freq +/- 1 digit.