Star clusters have an aspect that many beginner astroimagers seem not to have noticed. If you are not chasing faint nebulae within the cluster, like you would when shooting the Pleaides (M45), but you are primarily chasing images/patterns/colours of the stars themselves, be it in globular clusters or other types of interesting clusters, then the stars become fair game at heavily light-polluted sites, almost like for planets. You may well be amazed at how short you can make those sub-exposures used for stacking. At short enough subs you may well be able to ignore guiding, even at very long focal lengths, SCT territory. Why? Because the stars of interest are very much brighter than the sky background. Typically, a heavily light-polluted suburb (not yet the outer fringes of a large city, but closer in) may have a skyfog at about Visual Limiting Magnitude 3.5 = Mag 17.5 per square arc-sec. But most of the stars we normally wish to record in these clusters are brighter than Mag 15 to 13. I.e. they are at least 10 times brighter than that awfully bright skyfog :-) On the other hand those faint tendrils in nebulae are typically just a few percent the intensity of the skyfog. In this latter case the quantum statistical noise in the skyfog is relatively huge compared to the nebula signal we wish to dig out, but is fairly minor compared to the brightness of the stars in a typical cluster.
How do the minimal sub-exposure lengths click in? Let us say we wish to shoot at f10 in a VisLimMag 3.5 skyfog (Mag 17.5/sq arc-sec). We know from http://www.samirkharusi.net/sub-exposures.html that a minimal-length sub for efficient stacking using a Canon DSLR at a very dark site, Mag 21.5/sq arc-sec, will be approximately 1-minute at f2.8, or 13 minutes at f10. With a skyfog that is 4 Mag brighter the minimal-length sub for efficient stacking will be 2.512*2.512*2.512*2.512 times shorter, or 40x shorter = 20seconds :-) Hence we can go right ahead and use 20sec subs, confident that the stars that are of interest to us, brighter than Mag 15, will be exposed well enough to stack quite nicely, despite the bright skyfog. Can we go even shorter? Why not just try it out? Some of the examples I offer below used subs shorter than 20seconds, but in a DSLR that had clear glass over the sensor, sensitive to IR, hence almost twice as sensitive as a DSLR that has an IR blocker. When I compared things, my C14 does not seem to bloat stars unduly by including IR. In fact I even get sharper planetary images that way. Just keep on shortening those subs until, say, 80% show little or no trailing. No need to bother with autoguiding. If you are shooting using 2000+mm focal length, e.g. with an SCT, you may well be chasing seeing, rather than smoother, slower, mount errors. Below I am offering a few examples shot at a focal length of about 4100mm, unguided, using a C14, text beneath each slide, using modded and unmodded Canon DSLRs. To go deeper, just stack more and more of those short subs. This strategy works remarkably well for stars, but not for nebulae. For nebulae the numbers required in a stack, to make up for that very high skyfog, quickly get out of hand, into the thousands. But for stars in a cluster you will most likely find that less than a hundred subs can deliver a satisfying image.