tdon ZWO 120mm and ZWO 30F4 guide scope.
Are you trying to use multi-star guiding? The two things you mentioned above is probably the worst combination for multi-star guiding.
With multi-star guiding. The entire field needs to be flat (some of us use guide scopes with field flatteners for that reason). Additionally, you need enough gain to find 12 stars and then add yet another 10 dB (10 dB is equal to a ZWO gain of 100) to remove the brighter stars from hogging the centroid averaging, since ASIAIR uses SNR-weighted centroid averaging (very big mistake). If you don't remove the brightest stars, the 12 stars will only be equivalent to using just 2 or 3 stars, since all the weight is spent on those 1 to 3 stars.
I don't think your camera/guide scope combination is capable of that (the ZWO 30F4 produces especially nasty star shapes; if you can still do it, return it).
The EQ6-R is a very smooth mount. If you are stuck with using ASIAIR, why don't you try 3 to 4 second guide exposures, and use one-star guiding?
Multi-star guiding is really only ideal when you cannot use long guide exposure times. It is for people who have no choice, due to the mounts that they use. With your smooth mount, you can use longer exposure times to reduce atmospheric turbulence, instead of using multiple stars with shorter exposure. If your mount can keep within 0.5" over a 3 second period, just use single star guiding with 3 second exposures.
You also may want to recompute the amount of RA and Dec correction that is needed, to avoid centroid computing errors from driving the mount into large spikes. The max pulse duration should not be larger than the worst-case first derivative of your mount's PE curve multiplied by the guide exposure time. I suspect that your mount's PE derivative is much, much smaller than the 2000 millisecond correction that you are allowing ASIAIR to correct by. Even with my strain wave geared mount (which are notorious for large PE first derivatives), I don't need more than 100ms to 150ms max pulse durations.
Chen
