I did read that post before I posted my question.
My question was one of clarity for us.
Has anyone done such a test with a second SA to look at the TG signal alone to see if it in in the TG signal itself or is it a combination of the TG/SA on the 815.
Can you post screen shots showing the TG output of the 815 on the other SA you tested with if you've done this?
I'm curious to see if its purely the TG or the combination of both on the 815
FWIW, a few weeks ago I did a test where I connected the triggers of a a scope and the 815 and let the TG sweep. Regrettably it was inconclusive.
There is a definite discontinuity that occurs at 10MHz, but it was significantly less than that shown on the SA. I tried it at a number of levels too, in case it could be shown that the TG was operating non-linearly somewhere in the frequency chain: the error the 815 SA shows is much less with lower TG power, but the relative TG discontinuity remained the same at the 10MHz transition irrespective of power.
At that point I assumed non-linearity in the SA's mixer, but that doesn't add up either as you'd expect that behaviour to exist over a fairly broad frequency range.
Although I realise it's been diagnosed unfixable in software by Rigol, I am not convinced. if it were a hardware fault, I'd expect the glitch to exist over a wider range of frequencies. Having said that, it is possible that there is some interaction between the internal 10MHz reference and the behaviour seen, it does seem rather coincidental that the ref is at 10MHz as is this trace glitch. If that were the case though, I'd expect the error to be worse at lower TG powers rather than better.
So in short the jury was out for me, but as it doesn't affect me (for now!) I didn't bother to pursue it any further.