Ok, so I'm getting into guiding, got my Polemaster regime down pat (or so I thought), guiding seems to be going well (i.e. it is actually guiding after a few teething errors such as guiding on a hot pixel

) but my total RMS error is around 1.5 to 2 arc seconds and I believe less than 1 a-sec is the goal.
So, I'm looking at my graph, referring to the pdf for PHD2, thinking it's a little saw toothed which is over compensating.
Thought I would run the Guiding Assistant first, see what it suggests.
Coming back with big PA errors (11.3 arc minutes !!! - see attached screenshots) and recommending limiting exposures to 2 seconds to combat drift ! wtf ? The polemaster is accurate to within 30 arc seconds but even if I'm doing it badly I can't believe it's as bad as 11 arc minutes !
If the mount isn't exactly level could it cause this degree of error ?
I'm going to try the Sharpcap PA tool tonight but I really thought I had PA really accurate with the Polemaster.
I looked at the PA drift alignment tool in PHD last night but didn't seem easy to do having never drift aligned before. I think the sharpcap program is a little easier to follow? Abit more visual like the polemaster software?
What are your thoughts ? I was doing up to 4 minute subs last night that seemed fine - round stars, but maybe I need to be more accurate for clear sharp nebulosity?
What about the sample graphs - overcompensating ?
Thanks in advance
Equipment:
HEQ5 Pro with Rowan belt mod
AT65-QED
Canon T2i
380mm guide scope, helical focuser with mono skyris cam