This adiabatic cooling can cause ice to form in colder water. Since piston regs are usually not environmentally sealed (I have never seen one, may be wrong here), you can have ice form inside the regulator around the HP seat, which will prevent the piston from sealing properly. Since the HP seat doesn't seat, this causes an increase in the Intermediate pressure, and will force the valve open on the second stage.
Basically, what happens with a first-stage freeze is this:
1. The gas expands (the entire point of the reg is to drop the 3,000 psi of the tank, or whatever it is, to the IP for the second stage) and as the gas expands it cools (the exact opposite of what happens when you fill a tank and it gets hot. Heat is, in the gross physical sense [simplified!] the representation of the "vibration" of the molecules in a substance; at absolute zero there is none. If you double the volume of a gas, you half its pressure, and the same amount of "vibration" in the molecules now takes place in twice the space - hence, the temperature falls.)
2. The cold gas absorbs heat energy (cold is the absense of heat) from the regulator parts in an attempt to come to an equilibrium. The parts closest to the orifice (the HP seat) where the expansion takes place give up the most heat, since the gas temperature rises as it absorbs heat while it travels towards the reg. In a piston regulator, this is the piston "knife edge" where the seat is, and the bore of the piston as the gas travels down it.
(BTW, this is why, beyond tank corrosion issues, having very DRY air is important - if its NOT dry then the moisture in the AIR in the TANK can freeze in the reg - irrespective of whether you are using a piston or diaphram design! Bad news!)
3. The problem occurs because a short distance down the piston bore there is an O-ring and, on the other side of that O-ring, the water that you are diving in. This water gives up some of its heat to the piston and, in doing so, can freeze.
4. The coldest part of the piston in contact with the water is right at the high-pressure O-ring. Unfortunately, in order for the reg to work, the piston must slide through this O-ring to seal and unseal as the gas is breathed. If ice jams in the space between the piston and the O-ring, it cannot move and will not be able to seal.
5. The result of this is that the 1st stage will continue to supply gas even though the intermediate pressure has been reached. This does two bad things - it will overcome the spring pressure in your second stage, thereby causing a freeflow, AND, since gas continues to flow, the piston gets COLDER STILL! That causes yet MORE heat to be drawn from the surrounding water, and even MORE ice to form.
Once this gets started it gets worse FAST, because you are supplying yet more heat-robbing expansion of supply gas that is lost to the freeflow. The only solution is to shut down that regulator and allow it warm up. By definition if you are diving the water must be above the freezing point - otherwise it would be a block of ice and you couldn't swim in it! Unfortunately, on a single's rig used for recreational diving, you don't happen to have a second 1st stage handy to use while the first one warms up!
Ok, so what can we do about it?
Essentially, we must keep the water away from critical parts of the reg, so that if it freezes, it does not prevent the parts that move inside from doing so. If we can do that, the reg does not "freeze up" and there is no problem, even if the exterior was to become a huge icicle.
We can do this one of three ways:
1. We can encase the entire reg in air, and "seal" it, then arrange for some other material (say, an external diaphram) to transmit outside pressure to the inside, so that the balancing function (which is the point of water being allowed in the reg in the first place - to balance increased pressure as depth increases) can be done without water getting inside. This is how a diaphram reg does this job. They seal the water outside, and while it can freeze, it won't hurt anything if it does. It presses on a diaphram on the outside of the reg, which transmits that pressure to the actual regulating diaphram inside through the means of a rod or lever. At no time does water get where it can jam the mechanism if it freezes.
2. We can fill the parts of the reg near the piston with something that is (1) very viscous, so it doesn't leak out and make a mess, (2) is not compressible, as is true of most liquids and semi-solids, and (3) will not freeze. This is done with some piston regs. The cavity where water would usually go is filled with either silicone or christolube, and then a trim ring is placed over the balance chamber holes with a very small hole in it to allow water to press upon the grease. Since the silicone grease (or Christolube) will not freeze, the problem is solved. However, this is messy as all get-out when it comes time to service the reg, since you have to clean all the old grease out of there, and it CAN be messy in between if some of the grease leaks! It also has a potential secondary problem, in that the grease can trap some water in the reg when you're done using it, and that water (especially if its SALT water) can cause severe corrosion damage over time, since it cannot be rinsed out and won't dry out.
3. We can try to EITHER (1) keep water away from the parts of the reg that can bind up and freeze, and thermally insulate the cold parts of the piston such that not enough heat is transferred from the water to the cold piston to freeze the water, or (2) INCREASE the flow of water through the balance chamber so that it doesn't stay in contact with the cold piston long enough to give up sufficient heat to freeze. This is what SP and some others have attempted with their "thermal insulating" systems. The problem with this is that it is not foolproof, since you can reduce but never eliminate the transfer of heat, and in some ways insulating the piston actually makes the problem WORSE, in that the colder the piston the more heat it desires to transfer from the water. IF the integrity of these mechanisms to prevent water contact fails for any reason, you will get an instant flash-freeze, as opposed to a gradual problem building. So these systems are in some ways useful, and in other ways counterproductive (especially when they fail to work!) SP has been at this general game for a loooong time with varying degrees of success. That they keep changing it says everything you need to know about their ability to actually accomplish the goal - if they had found "the answer", they wouldn't have a reason to change it all the time, would they?
IMHO the
best solution to cold water (under 50F) is a sealed diaphram regulator. Second-best is a piston reg filled with grease. Third is attempting to keep the water from freezing but letting it in there anyway.
For those who dive in warmer water, however, piston regs have certain advantages. They have far fewer parts in them than a diaphram reg, which means they are less likely to fail and are easier to fix when they do. They also generally have the ability to deliver more gas than a diaphram, all things being equal, because the gas passages inside are both larger and more direct. Finally, because the gas path and area involved in items like seats are larger, they are more tolerant of abuse, such as accidental ingestion of water into the gas pathways. This does not mean you don't have to be careful about getting water down the inlet (you do!) but that if some DOES get down there it is less likely to cause severe and immediate damage with a piston than it is with a diaphram regulator.