In other words, something has to move. Just what is moved determines the focus mechanism.
Imaging Surface - This might be called the "refractor method" and consists of a manual mechanism for moving the camera, often a rack and pinion style focuser.
No magnification factor present (see next section).
Can be used with a position readout gauge (such as a machinist's gauge).
It is very hard to make small moves manually (did you ever try to turn a knob 1/200 turn?).
Sag and image shift can occur with heavy cameras or less than beefy focusers.
Primary Mirror - This might be called the "SCT method" and consists of a manual mechanism for moving the primary mirror, which is the standard method for a stock SCT.
It is simple and cheap since this is the way most SCTs come.
Position measurement can be added (but beware of backlash - see below).
Simple and rigid camera mounting is possible since the camera does not need to be moved to focus.
Backlash! Since many of the mechanisms used in these systems are known for their backlash, it is necessary to approach focus from a direction that will be consistant and minimize the effects of this. In most SCTs, this means approaching focus with a CCW knob movement. In some systems, this may still not be effective.
As with the manual movement of the focal plane, small moves can be tough when turning a knob with your hand.
In any system where a curved mirror (primary or secondary) is being moved to focus the system, the amount that the focal plane is moved will be greater than the amount the mirror itself is moved. The difference is determined by the magnification factor of the mirror, often 2X to 3X. This makes the problem of small moves that much more serious than when you are moving the focal plane (camera).
Flop! SCTs are noted for mirror flop when moving the mirror or the OTA itself. Some are worse than others.
I refer to the next set of mechanisms as "manual-electric" instead of "electric" since they are not totally automated. They still have a human in the loop to push the correct button and make the required moves of the focus mechanism. These come in the same two types as the manual versions mentioned in the last section plus a third version where the secondary mirror is the component that is moved to accomplish focus.
Imaging Surface - In this case an electical mechanism is used to move the camera to focus. The JMI DRO is one example, the RoboFocus is another.
As with the manual version, no magnification factor is present.
These often come with a digital position readout for determining where you are at.
Since the mechanism is electrical, small, well-controlled moves are much easier to make than with a manual system.
Mechanism must be beefy to avoid sag and consequent image shift.
Backlash in the gearing system may allow less than accurate readout of position.
The extra length introduced by the mechanism may cause problems with vignetting.
Primary Mirror - A variation of the standard "SCT method", this consists of an electric mechanism for moving the primary mirror. An example is the standard add-on electric focuser.
Smaller moves are possible than with manual movement of the primary.
Camera mounting is simple and rigid.
Magnification is still a potential problem but somewhat less so since the electric mechanism allows for smaller moves.
Focus shift and image shift are still possible problems.
Backlash is still an issue, and as with the electic movement of the imaging surface, may present a problem if a digital readout is part of the system. If the digital readout counts focus motor steps (as opposed to actual position), it will keep right on counting steps during the backlash reversal which can lead to an inaccurate perception of focuser position.
Secondary Mirror - Mostly present in higher end systems, this consists of an electric mechanism for moving the secondary mirror. This is the type used in my system.
Small moves are possible since the mechanism is electric.
Remote digital operation and readout are commonly available.
Rigid camera mounting is possible. With my system I have eliminated the rear focuser completely and use a bolt-on system.
Backlash and consequent error in position readout are still a possibility. Mostly they are less common with electric secondaries as they are often part of more expensive systems which usually have more precise mechanisms.
The secondary is almost always a magnifying mirror so as with the primary, smaller steps will be needed to compensate for this. Since it is an electical system, this should be possible.
Mirror and image shift are a possibilty as well with any system where the mirror is moveable. Since the secondary mirror is much lighter than the primary, these are less likely to be a problem. Still, this method does require a heavy and often expensive secondary mechanism.
Why have a human in the loop at all? The ideal system would look at a focus star and move the electric focuser to perfect focus without any human help! In fact, many of the professional adaptive optics systems do this in real time as part of the AO system. A more mundane system would allow "focus on command" and give a precise result. Software became available in late 2001, using a new parameter called "Half Flux Diameter", to do just that. Called FOCUSMAX, this product of two amateurs, works with several motorized focusers to archive accurate automated focus. I will describe this new and exciting method at the end of this discussion.