Bill of Materials:
1. Alkin 3.5cfm 220V, single phase compressor. (Damn nice unit, by the way.)
2. Backpressure regulator and check valve, plus a hex nipple and 90 degree elbow, for the compressor (to keep the filter stack pressurized and thus improve its efficiency when filling a near-empty tank), along with preventing any backflow into the compressor from a tank when it is connected.
3. Second fill whip w/valve (used to connect to the Hyperfilter)
4. Lawrence Factor Hyperfilter, 7,000 cuft capacity (to the right.) I removed the original whip that came with the compressor, O2-cleaned it, and installed it on the output of the LF unit - its a much nicer whip than the separate one! Note that this is NOT an Oxygen whip (in fact, none of these are)
5. I also added a digital pressure gauge to the whip for final pressure monitoring. It is more accurate than the one on the hyperfilter. This is a 1% gauge - not a precision 1/4% unit. Accurate enough - within 50 psi.
The reason for the hyperfilter is that I am partial-pressure Nitrox blending. Not shown is the O2 tank and whip - the whip is current complete EXCEPT for the handwheel and nipple, which, of course I need to hook up to the bottle. Should have been here today, but its not yet - probably will be here Monday.
The O2 tank itself is chained to a hand truck (which is chained to the wall when not in use, so it cannot fall over), and for filling use will be rolled out into the driveway.
Why? Two reasons:
1. IF you are going to have a problem with a partial-pressure fill, odds are it will happen when putting the O2 in the tank. If you have it outdoors and away from the building, you have the option of running. That is not an option if its in your garage and lights it on fire! (Putting an O2-enhanced fire out if you can't get to the valve is likely to be flatly impossible.)
2. The risk of a problem while adding the AIR is controlled by (1) pressurizing the whip above tank pressure BEFORE opening the tank valve (thus pure O2 never passes above the valve seat) and (2) using hyper-filtered air.
I considered continuous entrainment ("stick" nitrox mixing) and for the time being have decided not to go that way. I am just not convinced that it is safe and doesn't lead to excessive wear of the compressor hardware itself. I may revisit this if I end up dumping a lot of half-full tanks, but Lawrence Factor specifically does not recommend running enriched air through the hyperfilter, and, for that matter, neither do any of the compressor makers (although lots of people do indeed do it!)
Fill time is a bit over 20 minutes from 500 - 3000 psi; the first 2-3 minutes or so are spent pressurizing the hyperfilter stack before gas actually starts to flow into the tank - a consequence of the considerable volume of the Lawrence Filter stack and the priority valves fitted in the system. If you're filling multiple tanks that obviously only has to happen once, which helps a great deal.
The Hyperfilter and check valve on the compressor are my anal side on safety issues showing through - many people will say its not necessary, and they may be right. On the other hand even with the cost of using it a Nitrox fill will cost me about 1/4 of what I pay for them at any of the local shops.....