About 6 months ago I've build the Mega2560 version with drv8825. Nema17 1.8deg and belt mode MXL 60:15. I have to say I am impressed and very happy with the results except for some inconsistencies working with PHD2. The AstroEQ settings are
Axis steps/rev -3 686 400
worm stteps /rev - 25600
sidereal speed Ival - 310
resolution - 0.352
The thing is that some times I manage to get guiding within 1arc"(even without proper polar alignment) , but in most of the cases the guiding within 16arc". I've noticed -during calibration the dec does not respond the same way as Ra. Bare in mind the mount works perfectly with CdC and sync with astrotortilla it slew, goto - every time spot on the cross-hair. I'm using the standard SW 9x50 finder as a guide and either icx098 or ICX414 as a CCD and for this setup the PHD calculates calibration pulse >1000ms. I've tried manual guiding and when the guiding pulse is too big and overlap the DEC motor start grinding in place. when I use short pulses separated - the mount moves as expected. I've put intentionally wrong data in calibration pulse calculator(Longer focal length 1000mm instead of 170mm) and then - magic :the calibration goes fine. I am not familiar how phd sends the guiding pulses, but there must be some discrepancies or the problem could be with my build ?