Here are some things to look into:
1) make sure your blast orifice is only about 1/4" open on your blast end (because that will push open to 1/2" when the air pressure hits it), and 1/2" on the other end. If you have it too far open in any place, this will cause imbalance in the system, and thus dust.
2) Make sure your uptake and downflow tubes are un-restricted. Even though it would seem unecessary to check the down-tubes (since no debri travels through there), you should still check it. I have seen wet mist slowly carry enough sediments, that it starts to coat the sides of the down-tubes as well, eventually causing constriction, and thus imbalance, and thus dust
3) Make sure all your tubes are fit into the sleeves where the connect. Sometimes this is deceiving, because a tube may look
like it's connected well, but back behind where you can't see as well, may be sagging off the elbow hooks that it is supposed to fit into.
4) Make sure there are no rips or tears in the tubs, and that they are hose-clamped on tightly, so no air is escaping.
5) Make sure the part of the hopper where the dirt comes in, is not caked with solidified mud. I have seen where tubes may be free of jams and caking, BUT the part of the hopper where the air curls into the hopper has caked up with mud, effectively reducing and constricting the airflow, which in turn, causes dust d/t the imbalance. When checking for this one, you have to look carefully, because in poor light (or even good light), the caked debri can be so smooth, and the same color as the metal, so as to appear to be the actual metal. Ie.: it is worn so smooth and even, you may have to poke with an ice-pick to really tell if you're looking at mud cake, or the actual metal surface.
6) Make sure your screens are clean and clear. I used to not worry if perhaps 2 or 3 of the 9 panels were clogged (leave or mud or whatever), because I looked at the remaining 6 or 7 panels and figured there was still plenty of breathing room before I had to stop and hassle with cleaning them. But I soon found out that if, for example, 3 of the panels were plugged out of the 9, you have basically dropped 1/3 efficiency of balance, and thus will be more susceptible to dust. Watch also for screens that may not be fully clogged, ... like where they all have "daylight" showing through, yet the individual holes are not completely free of caking. Certain types of debri are more prone to making this happen, like asphalt crumbly stuff, or cement washout stuff, where it is sticky and adheres.
7) Don't depend on your in-tank water sprayers to hit and arrest the incoming debri. We just disconnected ours and never use it (they're a hassle always getting in the way, tips clogging, etc...). Instead, we just do this: When you start your job, fill your hopper (like with a garden hose or whatever) with about 1/4 to 1/3 of water. What will happen is that this bathtub of water sloshing around in there as you drive, will create a vapor. The in-rushing air hits that water, and makes a heavy vapor/mist, which accomplishes the same thing, if not better, than in-tank water sprayers. As you're sweeping, the water will eventually be totally absorbed into the debri, and then you can just stop and add more water, to bring the water table up to a level above your debri again. Of course, you can't fill above the level of the peak door, lest you get a lapful of water when you go to look inside
So with this method, you'll have to dump and re-fill with water more often.
Make sure your flaps are good, and touching the ground all the way around, and make sure your skids are flat on the ground.
9) To test to see where your air flow is leaking, to cause imbalances: close your leaf bleeder, park on a flat spot, and turn on the system. Then walk around with a piece of paper around all the portions where air flows (tube, head, doors, etc...) putting it next to any place you suspect is a weak spot. If the paper blows away, or sucks in (blow side or vacuum side weak spot), then that means you've found a spot where air is escaping. We used this test to find out that the rubber weatherstripping wasn't working well. A cotton ball or other such light-weight test item would adhere to the seam where the peak doors close, showing us that air was being sucked in at that point, and we need to re-weather-strip that. (this is just an example, a week spot can be multiple places)