Hi Jeff, sorry for the delay. It looks like the mount is
the problem, not something you've screwed up with the guiding parameters.
To start, the backlash test you did demonstrates pretty clearly that you have a
lot of reversal delay in Dec - it takes over 3 seconds to reverse direction on
the Dec axis and get any actual movement. You can see that in the graph
results - the "flat top" is what we're looking at:
Each of the data points represent a guide pulse of 500ms, so
the reported delay value of 3.3 secs is pretty easy to see. Unless you can
fix this mechanically, it's going to be difficult to get good guiding results
unless you switch to uni-directional Dec guiding. This backlash problem
also affects your calibrations unless you manually clear the backlash before the
calibration begins. You have to do this every time you calibrate.
With PHD2 looping camera exposures, use your hand-controller to move the mount
north ('up' button)until you see the stars in the display clearly start
moving. Then start the calibration.
Next, look at the raw tracking performance of your mount in RA
for the 14 minute session starting at 22:06 - this shows what the mount was
doing with guiding corrections removed (RA in red):
You can see the long-period oscillation from the uncorrected
periodic error but this is actually guided out reasonably well. But look
at all of the ragged, abrupt guide star movement that shows up in various
segments of that curve. Many of the excursions are both abrupt and large
and these aren't going to be handled well by guiding. This is why your RA
performance is always much worse than Dec - and as it turns out, your worst-case
results near Dec = 0 would be 35% worse than this.
You should probably seek EQ-specific help on what you can do
to improve on the mount behavior. Other forums have more contributors with
that sort of specific knowledge.