The experts will correct me if Im wrong but 109 being unbalanced will freeflow at depth if paired up with one of the overbalanced first stages like mK17 an apeks or zeagle and tuned sensitively at the surface.
I don't think the MK17 has the same 'overbalanced' feature that some other sealed diaphragms have. Overbalanced is not really the right term, those regs are actually 'over-depth-compensating" in that they increase IP above changes in ambient pressure. It is true that unbalanced downstream 2nds will be more susceptible to free flow as the IP increases, but I'm not sure how much of an IP change actually occurs. Even balanced 2nds have some downstream bias, usually quite a bit, so with enough of an IP rise they'd flow too.
I never heard of the MK17 having that feature, but the MK25 is sometimes called 'overbalanced' and in that case, it is a more accurate term. There's a slight flare at the bottom of the piston shaft (supposedly, I never noticed it) which allows IP to theoretically rise as tank pressure falls. But, I never noticed a MK25 (or MK20 with the composite piston) having higher IP at 300PSI than with a full tank. Probably the friction decrease at the HP o-ring with lower tank pressure offsets any change in downstream bias caused by the flare.
---------- Post added ----------
We used a MK 17 as the first stage. His was yoke, mine was DIN not that I thought it mattered much.
I would really think that a MK17 1st stage with a 109 2nd would be about as freeze resistant a set up as you can find. I have noticed, though, that if I set my 109s (well, actually balanced/adjustables) too lightly, they're pretty sensitive to case fault geometry flows, and that flow has occurred at some depth, which is counterintuitive. Case fault geometry is basically the difference in relative depth between the diaphragm and exhaust valve; if the pressure differences are enough, exhaust valve will essentially open at a pressure low enough to not allow enough pressure to stay in the reg body to keep the diaphragm from collapsing a bit and opening the lever. Since pressure differences between a set distance (about an inch in this case) are greatest at shallow depths, you'd expect this to happen shallow, but not deep. My guess is that with increased flow at greater depth, there's an increase in venturi effect and that's what causes the slight flow. That's a guess, though.
I've also found, unlike a few of the other fans of the 109 on this forum, that mine are a little more stable converted to balanced/adjustable, and overall seem to hold a tune better. It could be something as simple as the conversion process results in a new spring, which might be more consistent. I'd be interested in trying a 109 with a new unbalanced spring. Anyhow, if your friend is having some flow problems with the 109, it's worth trying to convert it with the s-wing poppet and G250 spring. And make sure it gets set to 1" or higher cracking effort, and maybe even try a slight turn on the breathing adjustment at depth.
