With that in mind, I thought I would start a new thread so that everyone hear can chime in with their own experiences.
David - I wrote a response to your last lengthy dissertation to help as much as possible, but it got lost in cyberspace somewhere. Probably someone on the bbq or sheepshearing newsgroup is screeching to me right now about using OT in the header. Since you seem up against it, here is another shot at it.
No one can help you get this to 100% guaranteed solution without being there with you. All advice (certainly mine included) about what is going on is based on personal experience and your description of the problems coupled with you attempts to resolve them yourself. So as they say, YMMV. I use solvent based finishes by choice due to their predictability in application (at least in my hands), the results they give, and my familiarity with that type of finish. I use water borne when required, all the principals are the same as solvent, and I know the products you are using are top shelf. So is your application equipment.
I don't know how long you have been doing finish work, and don't know how familiar you are with turbine powered HVLP. If you are an old high pressure sprayer, good HVLP requires a brand new skill set, one that comes only with experimentation and practice, practice, practice. I don't know if you are doing this professionally, part time, finishing this at your own home (with the a/c climatizing the house and affecting humidity greatly), or on a remodel job (a/c working intermittently doing humidity voodoo dance) or on a jobsite that is has the a/c heater turned off completely.
All that being said, I would start by doing this: buy two gallons of your finish and the recommended thinner for it. Call MLC and ask them for the MSDS sheet on this product. It usually has all the safety info on it, but many also have recommended application protocols. These are good place to start.
For water borne, I use distilled water as a thinner, never bothered with flow agents. My 4 stage (and your 3 stage) have plenty of power to push the material out of the gun. Buy a good set of stainless steel measuring spoons and download the measurement conversion tables on the net so that you will know that 1 tsp = 15 ml. Go to the builder recycling thrift store or a lumberyard and buy a piece of 1/8 (or whatever is cheap - door skins will work) plywood. Take it home and clean it up and sand it. This is your test bed. When evaluating or learning about a product, I usually take 2" masking tape and tape off squares 12" high and 2' across. You need the two feet across so you can mimic your spray motion, so don't go 12" X 12". If you test on both sides, sand one of them off.
Mix up 1/2 quarts at a time of finish, and write down your mixes, and spray your plywood in a clean, dry place. Observe the results. Make your notes. Adjust the pressure on your gun, vary your mixes, add the flow enhancer and leave it out. This must be done on a stable weather day as you will learn nothing if it is 55 degrees in the morning and 75 in the afternoon with light showers.
You may not get it done with just a couple of gallons. But I can't put enough emphasis on you taking this to the woodshed. You just need to get in there and hammer this out. I have said this a million times, finishing is like any other tool, you have to learn how to do it. That means an investment in time and money. The finish I am using now in my Fuji took me about a month to get all the way up to speed, and I burned about $350 of material learning to use it. But now... confidence in the material and application no matter what the weather is doing or the jobsite conditions, is high.
Thoughts on the finish: First, call MLC. Tell them what you are doing and what is happening. Personally, from your description I think you are needing to thin more, but that may be incorrect in this case. As them what the MAX amount of thinner you can put into the sealer would be, and then go from there.
That leads to the turbine gun itself. If you have the three stage, you probably have the adjustable rpm model. Put this to the highest rpm, and epoxy the knob in that place. The Turbinaire gun is a very nice gun, I have used them and they were my second favorite to use. But that gun that comes in the 3 stage package has been around for a long time (several years), and was designed to be used with a high volume output turbine. It was NOT made for someone to turn down and adjust volume flow. You can restrict the flow with the valve, but that doesn't restrict the amount of air available to the gun. When you restrict the output of the turbine, you take away the available amount of air to the gun.
Why is this differnce important? You HVLP not only pushes the air at the aircap to atomize the material, but it siphons off 5 lbs of pressure to pressurize the paint pot on the gun. Less available air to gun means less immediate pressure to the pot when you pull the trigger. Inconsistencies in pressure mean inconsitent finishes. I was really suspicious when I saw the variable speed turbine, and didn't know why it was made or how it was supposed to work. But a call to a completely honest Turbinaire distributor confirmed everything I just wrote. Results? Keep the knob all the way over unless you are spraying tanning juice on your girlfriend.
Your pressure at the valve or on the gun should be turned down quite low. With HVLP you are NOT atomizing the finish. If you are getting anything remotely close to traditional high pressure overspray you need to cut down the flow at the gun. You should be spraying droplets with your gun like a gentle mist. I have literally sprayed clear onto front doors in garages where I just put down a 9X12 dropcloth and didn't have ovespray problems. When you are practicing on your plywood, turn the gun (again at the valve) down until it just sputters. Now open the valve the gun 1/4 turn at a time, until you can just get the finish out in a clear fine mist with droplets of consistent size. If you start to get a lot of bounce back from the surface (remember... 8" from the surface) you need to cut back the pressure and thin your material. If the bounce back comes back to the gun (probably anything more than about 4 inches) the pressure is up to too high. Thin.
I am suspecting that your bubbles are the results of too much MLC flow material, not enough thinning, and the pressure too high. This is the practice part you need adjust your gun to the MLC material.
Two other things contribute to bubbles. Incorrect mixing technique and incorrect application thickness. If you pour all your stuff into the sealer and stir it vigorously you will introduce bubbles. Since HVLP does not blast the finish apart then redirect it with air, it will allow some find bubbles to stay in the sealer when exiting the gun. In the good old days of high pressure, I was more worried about solids on the bottom of the cans of finish (paint), so I would add the thinner at the paint store and have them SHAKE the damn can up. Since your HP equipment blasted the finish anyway, it took care of the bubbles. Now with you should add your thinners/retarders/hardeners and mix slowly for a long time. I mix with a clean stir stick.
The other thing that will cause bubbles is to have the pressure a little too high, and put on the finish too thick. You are getting closer if most of your bubbles are working themselves out. Going beyond how the bubbles got in the finsh (spray pressure too high or improper mixing) the bubbles get trapped because the finish is starting to catalyse where exposed to air. The finish "skins over", not allowing the bubbles to escape. If the finish is too thick, it cannot outgas the solvent before it skins over, so everything is screwed up. Bubbles are trapped, witness lines become apparent, and varying degrees of gloss are seen. When the finish catalyses it has to follow its own precise formula of not curing too much on the top before the project side is well into the same process. The bubbles in your finish that you see disappear are actually the result of the contraction of the finish (when the solvents are leaving) squeezing out the bubbles. So in order for that squeeze to take place, you have to keep the top surface wet as long as needed. See above for practice advice.
When the finish is applied, is important to follow the manufacturer's recommedations so that you can have a good starting point before developing your own formulas. I >think< that the recommended thickness of coat for the MLC product is 3 mil. If you don't have a gauge, go to the auto paint store and get one as they will really help when you are trying to see the correct appearance of your coat when properly applied.
To the last point. If you are doing this professionally, you will starve to death if you don't master the "off the gun" finish. Polishing and rubbing out have to be done on occasion, but more often than not that kind of finish is simply not called for. So that puts you into competition with the guys that will spray and walk out.
To get the finish to look right, I apply as many coats as needed to seal the wood. For my favorite finish, I apply three - four coats of finish, and when fully cured it shrinks down to about 3 mil. I sand lightly to remove anything I don't like. I never sand between coats unless I have screwed something up. It doesn't matter what kind of finish I am putting on.. I never do. On the last coat before top coat, I very lightly sand with 220. If you sand one part, be committed to sanding it all for the same appearance of gloss.
Get the correct gloss from the manufacturer. Don't try to knock down gloss to satin with steel wool, oils, waxes etc. Remember, you aren't making furniture.
For the final finish coat, I thin as much as I can get away with (sometimes as much as 50%) and spray a water thin coat over the sanded and cleaned finish to top coat it. This will provide the proper sheen for the finish and as thin as it is will dry rapidly helping to prevent dust contamination.
Good luck. Let us know how you are doing.