Initial alignment

Start with the telescope perfectly level and roughly pointed North. If you have a 'polar axis finder', by all means try using that to 'spot' Polaris = it's virtually impossible to get you head around a mounted telescope Optical Tube Assembly (OTA), so it's only really useful for aligning the tripod before mounting the tube.

Then mount the OTA and adjust the balance.
To ensure that the gears inside the mount always mesh perfectly, the mount should always slowly lift its load up, and not slowly lower it down. Pivot the OTA / counter-weight arm so the OTA is on the East side (left, when standing behind the mount, looking North) and the counterweight on the West (right). Add the spotter scope, camera, co-mounted guide scope etc.. Then adjust the counterweight so that side is lightly heavier (i.e so the OTA assembly slowly rises)
Swing the arm so the main telescope is aligned parallel to the RA (North pointing) axis, then clamp it. If you look through the main tube, you should see Polaris.
If the 'polar finder' IS pointing at Polaris, but you can't see Polaris through the main OTA, then it's not parralel with the polar axis - loosen the clamp and move the OTA. If you don't have a 'polar axis finder', and are sure the OTA is correctly aligned with the mount axis, then you have to move the mount head using the left/right micro-adjust clamps and the lattitude adjust screw until Polaris is centered in the main OTA field of view
With Polaris is centered in the main OTA field of view, adjust the OTA finder scope so Polaris is centered in the cross hairs.

If you now run through your GoTo controller's '2 star' or '3 star' alignment sequence, the 'guide' stars it chooses should be perfectly centered in the main scope (and finder scope).

This complets imnitial set-up. The next step will be a 'fine adjust' using the Drift Alignment method

NOTE from now on, the OTA must remain clamped to the mount - if you release it, the GoTo set-up will have to be repeated 'from Polaris' (although the drift alignment will not)

Drift Alignment

Based on material I found here

Usinga 'movie' camera

You need a camera that will take a series of images so you can judge if stars in the field of view are 'drifting'. The obvious choice is a web-cam (or your guide camera) - a DSLR (even with 'live view') is less suitable as it's hard to see any but the brightest of stars (even at the maximium ISO)

Enable the camera and move the whole OTA to point at a bright star about half way up the sky.

With the image of the star displayed onto the video feed, twist the guide camera as necessary so east-west direction corresponds to the horizontal, and the north-south direction corresponds to the vertical (and with the camera’s base toward the south, with its top towards the north).
To check this, use the hand-controller to move the telescope in an east-west direction (the guide star should move only horizontally i.e side to side). Then use the hand-controller to move the telescope in a north-south direction (the guide star should move only vertically).
Make usre the mount is in 'tracking mode' (i.e. so it's tracking the star as best it can).
An error in tracking may cause the star to drift along the east-west line (i.e. horizontal, left-right) while an error in polar alignment will cause the star to drift along the north-south line (i.e. vertical, up-down). It's more important to adjust any north-south drift as east-west drift is easily corrected by the autoguider. The drift-alignment method requires us to observe the drift of 2 stars: one in the zenith and one on the horizon. It works in a rather simple manner: 1. For a star in the zenith: a drift along the north-south line (red path) means that the mount’s polar axis needs to be moved horizontally, to the left or to the right (azimuth). 2. For a star on the eastern or western horizon: a drift along the north-south line (red path) means that the mount’s polar axis needs to be moved vertically, higher or lower (altitude ). By observing the drift with each adjustment made, it is possible to determine if the most recent adjustment helps correct the drift or not. The image below shows how the mount’s polar axis may be adjusted horizontally (azimuth) and vertically (altitude). Screws A, B, and C allow most equatorial mounts to move the polar axis horizontally (A and B), and vertically (C) in an altitude-azimuth manner. To illustrate, let me give an example. Suppose we are currently pointed to a star near the zenith and we have observed that it drifts vertically (either upward or downward, it doesn’t matter which direction). Since the star is near the zenith, it means we need to move the polar axis horizontally, to the left or to the right, but we do not know yet which of the two directions (left or right) will lessen the drift. We arbitrarily chose to move the polar axis, say, to the left, and observe if it corrects the drift. If yes, then we continue to move it to the same direction (in this case, to the left), otherwise, if the drift has worsen, then we move it instead to the opposite direction (in this case, to the right), and continue adjusting until the drift is finally corrected. We now point the telescope to a star on the eastern or western horizon, and we have also observed that it drifts vertically (again, either upward or downward, it doesn’t matter which direction). Since the star is on the eastern or western horizon, it means we need to adjust the polar axis vertically, pointing it higher or lower, but we do not know yet which of the two adjustments (higher or lower) will lessen the drift. Again, we arbitrarily chose to point the polar axis, say, a bit higher, and observe if it corrects the drift. If yes, then we continue to move it to the same direction (in this case, moving the polar axis a bit higher), otherwise, if the drift has worsen, then we move it instead to the opposite direction (in this case, moving the polar axis a bit lower), and continue adjusting until the drift is finally corrected. Better polar alignment is achieved by repeating this method several times. Permanent observatories are polar-aligned in this manner. At first it may seem difficult, but through practice, it is possible to drift-align in less than 10 minutes. As soon as an acceptable polar alignment is achieved (no drift within 5 minutes), we are now ready to start with the actual guiding operation. We now turn our attention to the main imaging scope (with the DSLR attached). Point the imaging scope to the area of the sky you wish to photograph, frame it properly, adjust focus, then begin tracking. To avoid complexities, I strongly suggest (if this is your first time to do this) that you try to image targets located to the east of the meridian. For objects located to the west of the meridian requiring what is called the ‘meridian flip’ (i.e., flipping the telescope 180 degrees along the declination axis upon reaching the meridian), you must invert the RA signals by clicking ‘Telescope‘ and ‘InvertRA‘. Now look for the nearest bright star (mag 5 or 6 perhaps) which could serve as a guidestar. Point the guidescope to this guidestar. At this point, the main imaging scope is now pointed to the target you intend to image while the guidescope is pointed to the guidestar. With the guidestar at the center of the field, click on the ‘Guide’ button (Note: The ‘Guide’ button is clickable only after the ‘Focus’ button has been clicked twice; click the ‘Focus’ button once to allow focusing, adjust focus if necessary, then click it again to deactivate it and allow guiding.).