Name of the Machine

The G1 Rebreather is named as a personal tribute to Geoff, the only diving buddy of mine who later died underwater. It is also known as the "Butter Fingers Rebreather" - because of the "Magnificent" engineering skills of its constructor, as in "Magnificent Men and their Flying Machines". What follows below is a teardown with photos; for a photoshoot with the G1 in action see pshoot.html.

Idea:

The G1 should be a very simple CCR, not too big, capable of sidemounting for cavediving purposes, no large turtle shell on ones back. This author is more interested in longer shallow dives than in deep ones and design may reflect this. MUST HAVE always visible heads up display with twin digital meters (most rebreather incidents, the victim didn't know his PO2 at the wrong moment). Only way of not looking at the meters on THIS rig, must be take the mouthpiece out of ones mouth.

Plan:

make it a complete self-build and have fun trying to do things like one-way valves. No manufactured-for-rebreather components allowed except the sensors!

Execution: 1 - the multimeters.

Multimeters 2 basic multimeters (cost UK £4 each, good thing because I flooded about 6 of them in various experiments) are frozen forever on their 200 millivolt range and potted in epoxy putty (i.e. epoxy glue mixed with dry sand to a sludgy consistency). You can see some of the sandy putty at the end of arrow 3 (see marked up photo below). The meter displays also have a layer of clear epoxy glue over them; so there is epoxy all round the meters in some form or other. A tiny patch of Blu-Tack inside keeps the epoxy away from the partly conducting foam that electrically connects each meter to its liquid crystal display (otherwise you have potted the meter but the display no longer works, I speak from experience). But apart from the Blu-Tack and the foam itself, the whole meter is then a solid lump of electrics/epoxy so should resist the water pressure quite well. As indeed it does - this pot-the-meter idea is one of the little triumphs of Butter Fingered Design.

Execution: 2 - the rest of the electronics.

numbered The meters are connected directly to two O2 sensors (no amplifiers on the G1, maybe I'll try some on the G2) and an external 9V battery to power the meters (the pods containing the sensors are at the end of arrow 1 on the marked diagram). Battery is switched on with a reed relay and so is each sensor. Three reed relays then, switched by external magnets (the relays are inside the lump of epoxy putty at the end of arrow 2 on the diagram; the silver grey band to the left is duck tape containing the magnets, which are held in place by a "snoopy loop" of cycle inner tube like the black band you can see to the right which holds the epoxy block in place). All the electrics (sensors, reed switches, battery) are potted in lumps of epoxy that are connected by phone extension cables (which contain 4 wires, about the right number. I originally used Euro-video connector cables - you can still see a bit of one, the black wire coming out of the left hand meter - but they have about 20 wires in them and it got too confusing). When you replace a battery, a lump of epoxy is thrown away with the old one and replaced; so also with the sensors. These events are expected to occur about once per year. (This minimal circuit uses minimal current so a standard 9V "PP3" type battery should last a while, and of course you will change the battery when you change the sensors). The magnets are the only things outside the epoxy and exposed to the water.

3 - building the mouthpiece assembly

The sensors are on top of the copper 22mm pipe which you can see in the picture, and which comes over the left shoulder of the diver (left hand side = upstream side on this rig since the inventor reads from left to right. If I were an Israeli or an Arab it would be another matter), turns a right angle (standard brass fitting), goes through a standard brass 22mm one-way "check" valve as used in the UK (and available at B&Q), to a tee where another right turn leads to a standard brass 22mm gate valve to protect the loop when I'm not breathing it, then an adaptor to the smaller 13mm size of pipe, the compression joint nut on which has been shaved in a bench grinder until just the right size for fitting a standard rubber mouthpiece over it. The mouthpiece is tied on with a plastic hose clamp as usual. Downstream of the tee there is another brass one-way valve followed by another right angle turn and a copper pipe leading back over the diver's right shoulder. Total weight of brass = enormous for a mouthpiece assembly, but surprisingly comfortable in the water. Gives one the appearance of a hammerhead shark.

4 - Counterlungs

Left Shoulder Before either copper pipe reaches my shoulders, it's covered with a wider (32mm) plastic pipe to which it's sealed with more epoxy to keep the water out (the photo shows what happens on my left - upstream - shoulder). The plastic pipe adapts onto plastic (easily removable and replaceable!) compression joints directly over my shoulders; turn down at that point and you enter the upstream (LHS) or downstream (RHS) counterlungs, both of which are the largest size of motor cycle inner tube I could find. Continue on and you are heading for the scrubber canister which is made out of a British army ammo cartridge container. (Convergence in evolution?? Duncan Price uses I think the same kind of container in his far more sophisticated DDDDD rebreather. Maybe we saw them in the same shop, in my case the Inglesport cafe/caving shop in Ingleton, Yorkshire - a well known haunt of grubby characters who go in caves, not usually underwater at least that's not the way most of them plan it. I'm told the British cavers who famously got stuck by flooding in a Mexican cave system came from this neck of the woods). The pipes into the scrubber have to be flexible yet not too easily crushable; mine are cycle inner tubes containing lengths of washing machine waste pipe to give the crush-resistance. At the scrubber end the waste pipes are tied to the 32mm pipe leading into the scrubber, which has holes in it for this purpose; the inner tubes go over the holes to preserve watertightness (I hope).

5 - Scrubber

Inside the cylindrical scrubber canister is a very simple arrangement, the Butter Fingers Four Sock Radial Scrubber. Two 22mm copper pipes go into the top of the cylinder, one in the middle and one at the side (see "scrubber lid" photo below). The one in the middle continues inside the cylinder to about half way down, and the exhaled air is released from it into the inside of a 32mm diameter plastic pipe with many 6mm (1/4 inch) holes in; on the far side of the pipe it is surrounded by a polyester cotton sock. Then there is a gap filled with sodalime (circa 2.5 kg of it) and then there is a cylindrical coarse wire mesh surrounded by more polycotton socks which prevent the sodalime escaping into the inhale side of the loop. The outer socks are sandwiched between two layers of chicken wire, the inner sock is sandwiched between the sodalime and the 32mm plastic pipe. The outer chicken wire is held tight by some string which you can see in the photo. Chicken Wire Top and bottom of the sodalime "cylinder" are protected by the two lids of a large can of spinach (see Spinach can lid photo below), and the sodalime is kept under modest pressure by two 4mm threaded rods (see photo) which link the two lids and are tightened after the compartment between them is filled with sodalime. The ends of the outer socks are usually secured with ducktape which I removed so as to get the photo. You can see the ends of the socks in the photo, and a bit of sodalime trying to make a bid for freedom... (the inner sock is secured to the 32mm pipe with a small circle of ducktape at one end). The other 22mm copper pipe simply ends just inside the scrubber canister. This scrubber seems to work. While the sodalime is fresh I can start my dive without bothering to pre-breathe the scrubber, and without any sensation of shortness of breath. Once it is getting used up, however, this policy causes a distinctive shortness of breath which is cured by a dose or two of diluent while the scrubber warms up. Incidentally I never carry less than a 12 litre cylinder of diluent. [NB Socks work better than mosquito netting which is too easy to put holes in]. Scrubber lid Spinach can lid

6 - Gas Injection and Gas Dump

In the left counterlung is a standard black drysuit inflator for inserting diluent. The button is just left of ones left tit. In the right counterlung is a highly nonstandard drysuit inflator (a Schraeder valve off a very old Northern Diver drysuit) for inserting O2. You can see it to the right of and below the 32mm compression joint in the photo. Come to think of it, that photo shows the right shoulder rather than the left (O2 is added on the right). The button is just right of ones right tit, and down a bit. This inflator happens to be red, appropriate enough for O2 addition. The left (upstream) counterlung also has an adjustable drysuit dump valve attached to it so that gas is released if the excess pressure becomes sufficient. The dump valve also enables one to do a standard pre-dive bubble check (close mouthpiece gate valve, tighten dump valve so it doesn't vent too easily, submerge the unit, press dil button until vented bubbles appear at the dump valve, check no bubbles elsewhere). The dump valve rests about on ones left hip when the unit is in use.

In Use: 1 - Making Sense of the Sensors

Using the rebreather: I turn it on (with the magnets) with the sensors exposed to air at 1 bar. Sensor range is 8-12.5 millivolts, in the photo you can see mine reading 8.9 and 10.4. One predicts that with pure O2 one should get 42.3 mV and 49.5 mV (pocket calc if doubtful). Now one gives the sensors some pure O2 and checks that one does. This r/b is used for shallow dives, so it makes sense to INTERPOLATE rather than EXTRAPOLATE when deciding on setpoint, so at present I aim for a PPO2 of 0.8 bar when in use - i.e. 33.9 mV (left sensor) and 39.6 (right sensor), comfortably less than the figures of 42/49 which one has already seen from the sensors. So we KNOW the sensors aren't "current limited" near the values we're aiming for. 34 and 40 are then the target millivolts, and at depths to 100 feet I can simply tell my computer I'm on air and have some extra safety margin (if your PO2 is 0.8 your FO2 is always better that 0.21 down to 95 feet of fresh water). Deeper than that - or doing a genuine deco dive - and one has to get smarter and aim for 1.0 not 0.8, and tell ones dive computer (Buddy Nexus) one's in CCR mode.

In Use - 2: Other predive checks and predive prerequisites.

  • My min configuration is 2 sidemounted tanks, one big one for diluent and open circuit backup, one smaller one for O2. More preferable though is TWO large dil/backup tanks and one small O2 tank, or just TWO large tanks one with air diluent and one with 40% Nitrox masquerading as pure O2, in which case I'm in SCR mode and will occasionally have to breathe out through my nose to maintain neutral buoyancy. This I like for shallow cavedives because you keep the number of tanks down, yet 40% is breathable down to about 100 feet so you have a lot of open circuit margin for error.
  • Undoing the plastic compression joints one listens to the noise as one breathes the large mouthpiece assembly only. Air noise should come from ones left ear - only - on inhale, from ones right ear - only - on exhale. If not the one way valves have a leak, usually because foreign matter has got in. (I once had this fault when a grain of sodalime successfully migrated round the loop).
  • Then one checks that ones open circuit sidemount gear is perfectly ready to get one OUT of whatever one is getting INTO, without any help from the rebreather at all - the rebreather is far too simple (nay crude!!) to have multiple redundancy, and therefore must not be relied on.
  • One does not enter the water without reliable backup buoyancy control, NOT just a drysuit. True, the breathing loop can be used for this purpose in normal circumstances, but lack of proper backup buoyancy control has been known to turn a loop flood into an obituary. This is especially true when combined with the problem that diluent/open circuit backup tanks are too small.

    First impressions when diving it

    The G1 Rebreather Once diving, one has to get used to how long (surprisingly long) it takes to breathe the PO2 down even by as little as 0.2 bar (~10 mV). And one uses the inflator buttons to keep the PO2 close to 0.8 (circa (34,40) mV). The hoses used to get kinked - I had to fix this as it otherwise causes excessive work of breathing. Once out of the water I have always always found about a cup of water in the right (downstream) counterlung, but scrubber canister and upstream counterlung rarely have more than a few drops. But unknown to me my mouthpiece had a small leak in it, so i hope to reduce the water in the downstream counterlung too.

    Future plans

    Scrubber bits For the shallow dives I'm now doing,, a PO2 of 0.8 is perfectly adequate and that's what I use. Later on, though, I shall use a PO2 setpoint higher than 0.8.

    To this end, I might flush with O2 at 10 (or 20) feet and get a direct reading at a PO2 of 1.3 (or 1.6), so one can verify things are linear as predicted up to and beyond a setpoint one may choose in the 0.9-1.1 (or 0.9-1.4) range. Note I would NOT consider calibrating at 1 bar as adequate preparation for holding the PO2 at 1 bar - interesting things might happen if the sensors' response fell off just above 1 bar in that case. Put not thy trust in sensors, nor in any child of man. Soon I hope I shall arrange some kind of pressure pot so I can check directly, at the surface, that the sensors are responding correctly to 1.3 bar PPO2. Then I will use 1.1 as my normal PO2 setting.

    I also intend to arrange for diluent to be blown directly onto the sensor faces at regular intervals, so they do not get covered in water and cease functioning.

    In the long term I intend more complicated electronics including amplifiers, solenoid O2 injection, CO2 detector, and so forth.

    In the meanwhile

    you can find the first 9 dives I have made recorded here . Nothing dramatic.

    AND FINALLY:- Things to Avoid - Causes of Accident in Order of Yours Truly finding numbers of accident reports on the Net.

  • KNOW THY PO2 Most frequent cause of death by rebreather - not knowing ones PO2. Perhaps an outright majority of rebreather deaths involve this error. [ I decided not to dive any rebreather without at least two digital PO2 displays always visible in front of my nose --- hence the eccentric design of the G1.]
  • MAINTAIN THY BUOYANCY Surprisingly, it seems to me from reading accident reports that buoyancy control failure is the second most frequent cause of rebreather deaths, after wrong PO2. Example: The most recent fatality on the List as I write occurred in Hemmoor, and it started when the victim had problems clearing his face mask, and his perilously small diluent tank ran out while trying to clear it. Eventual result (after confusion, panic and further errors): a violent ascent to the surface on inadequate backup gas. Two buddies very seriously injured with DCI. The victim died because he became negative at the surface and could not ditch his heavy primary light quickly enough, so he sank again and drowned. [I have decided my smallest dil tank will remain 12 litres. O2 tank as small as I please!!]
  • Errors that probably relate to the higher task loading are a part cause of some accidents. [Practise is the only solution to that I reckon, plus not doing too much too soon.]
  • Hypercapnia is an occasional problem 'specially with dodgy scrubbers like mine! [ Keep a check on scrubber hours and replace the chemical.]