CORRECTIONS, UPDATES AND ADDITIONS - click on title to select book, or scroll down to find

SCUBA REGULATOR MAINTENANCE AND REPAIR

IMPROVISED AND LOW COST HP GAS BOOSTERS

THE DIVELIGHT COMPANION

THE OXYGEN HACKER'S COMPANION

 

SCUBA REGULATOR MAINTENANCE AND REPAIR

SOURCES

Page 198 APEKS

Divers have for some time noted that the US/Seaquest Apeks TX 2nds breathe better than the Euro versions. The US versions, it turns out, are fitted with a slightly shorter balance chamber so the spring isn't as compressed, which reduces cracking pressure in the easiest breathing position. This makes a noticable difference in how easy the reg breathes, on the non-adjustable models expecially. The reason for the difference seems to be that the Euro regs must meet the very picky and restrictive European CE standards; the higher flow of the US model put it in violation of some cold-water standard.

Euro version TXs are easily converted to the easier-breathing US-specs simply by replacing the little plastic balance chamber part AP2038 - the US and Euro versions are completely interchangeable, to the point where Apeks uses the same part # for both. Seaquest approves the mod, and has an Aug 2000 dealer's bulletin detailing it.

The Euro version has a little tail on the end, is (usually) black and measures .886" (22.50mm) overall. The US one has no stub and is (usually) beige and measures .795" (20.20mm).

The two chambers appear identical other than the tail, and many divers seem to have simply modified their Euro chambers by cutting or grinding off the tail - several shops report routinely doing this to any Euro TXs that come in for service, as well as older Zeagle Apeks which used the Euro setup, and another shop reports their Apeks rep suggesting doing so (incrementally, using a utility knife). Care should, obviously, be taken not shorten it more than .795" (20.20mm) or to remove any of the body itself, since the pressure chamber comes very close to the end.

Back to TOP

IMPROVISED AND LOW COST HP GAS BOOSTERS

There have been a couple incidents recently involving homemade and converted boosters, that anyone thinking about building a booster should give some serious attention to. One involved 100%, and the other 50/50. While our book doesn't advocate using homemade boosters for boosting high FO2 mixes, we suspect enough of our customers may be tempted to do so that we have added a section discussing the two incidents and the risks involved in boosting high FO2s. You can download it in PDF format (360 K). If you have an earlier copy of the book we suggest printing it out and adding it to your book.

SEALS
A serious booster will usually require proper springloaded lip or "U" seals like Bal Seals or Parker's FlexiSeals, in a low friction compound (usually filled PTFE) suitable for dry running. McMaster-Carr lists some similar "Spring Loaded Teflon Seals". While the specs are not as good as the Bal or Flexi seals, their better availability may make them worth consideration. Most of the seals you'll find in the catalogs are not suitable for pneumatic use, and it can be hard to tell from the descriptions in the catalog since they are often mainly concerned with fluid applications. Most seal manufacturers and larger distributors will have engineers or tech reps who can help select the right product though it may take a few calls to find the right person. Keep in mind that, while the HP seal will usually be the most highly loaded, the drive cylinder on a gas driven booster will need low friction seals too. Another alternative are the Sirvon rings made by CR (Chicago Rawhide), or M-C's similar 5000 psi "Teflon Seals" (part 9420K**. these are no longer in the catalog but M-C can still supply them). These are low friction square section PTFE piston rings made for pneumatic applications, suitable for dry running, and good up to 5-8000 psi (500 bar) and 500°F (260°C). They are normally used with an O-ring "energizer" beneath them to provide extra tension, however, if slight leakage is acceptable (as on the drive side of a booster) they can be used alone. The M-C description only mentions oil and fuel, but the manufacturer (Allegheny-York) says they are suitable for air and O2, and we've had several users report good results with the M-C version in converted hydraulic cylinders.

There's a major catch to be aware of when you start going through the seal manufacturer's catalogs. Most of the seals listed in the seal manufacturers catalogs don't really exist! They will make them up if someone orders enough, but they aren't sitting on a shelf somewhere ready for immediate shipment. There are also some big seal supply houses that stock seals that even the manufacturers don't bother to stock, so it can be worth calling around. This is what makes the M-C seals so useful - while they may not be perfect, they are available off-the-shelf, in small quantities, at reasonable prices

ALTERNATIVES
Anyone thinking about buying a Haskel should be aware that there's another company, Hydraulics International in Chatsworth, CA that makes near-identical copies of out-of-patent Haskel models. Their AGT15/30 clone lists for about $4300 compared to $5800 for a genuine Haskel; however John Allen at Northeast SCUBA Supply sells them for even less!

And Jetsam, who make the lovely Micro Booster, has just come out with a new $1200 air driven booster, the Baby Booster. It's pretty much a scaled-down Haskel, intended, like the Micro Booster, for boosting relatively small volumes of gas into rebreather and deco bottles, and compact enough to take on the diveboat with you. Hydraulics International has a similar mini, which can also be run (slowly!) manually.

SURPLUS BOOSTERS
HP boosters turn up regularly used or surplus, on Ebay and from dealers such as Groban. While if you are very, very lucky you may stumble onto a bargain surplus O2 Haskel, most of them are snapped up by dealers and tend to go for top dollar (though a fellow recently posted about buying an AGD-30 at a dive flea market for $100!). Most of the used boosters one will encounter are non-O2 models, and the real bargains are mostly nitrogen boosters and amplifiers since these are used and surplused in great numbers by the military. Since these often can be had for $100-$400, they can be a very attractive alternative to a homebuilt booster, as long as one understands their limitations.

What's the difference? In addition to the lack of multi-seal separation, nitrogen boosters and amplifiers tend to be more cheaply and simply built, with aluminum heads instead of brass, steel or SS, and may not include cooling for the HP cylinder. The gas passages in the head and valving may not be engineered with adiabatic concerns in mind. The big issue, though, is separation. With any gas-driven booster there's always the danger that leaking seals could allow the boost and drive gases to mix. If an O2 booster being driven by dirty shop drive air allows it to come in contact with HP O2, bad things can (and do!) happen. Proper O2 boosters are built with multiple seals and a vented gap separating the two sides so that any leakage will be vented out rather than being forced past the seals into the other side. Nitrogen boosters/amplifiers usually aren't, and the drive and boost sides may be separated by only the HP piston seal. Even a good seal can leak in this situation, if the drive side is pressurized first, since the seals used tend to be directional.

Seal materials, interestingly enough, do not seem to be a problem. The O2 Haskels and their gas cousins mostly use the same PTFE dynamic seals in the HP section, and the same butyl valve seats.

O2 CLEANING HASKELS
Another question that keeps coming up is what is involved in O2 cleaning a standard (non-O2) Haskel. Having it done by Haskel or an authorized dealer usually means doing a complete, expensive, rebuild but if one has a Haskel that is in good operating condition but is not O2 clean, one might wonder how much difference there really is between the O2 and the regular models. Not much - comparing the parts lists for the O2 clean and regular versions will find that only a few parts, all in the HP cylinder, are actually different - in an AGT-15/30, just the 568012 valve O-rings and the HP piston bearings 51372 and 51365. This being the case, it should be possible for a reasonably competent individual to O2 clean a standard Haskel simply by doing a careful cleaning, as per normal scuba practice, and replacing just those parts. Only thing is, many Haskel dealers will not want to do this for you, for liability reason, and will insist on doing a full overhaul anytime they O2 clean a machine.

Back to TOP


THE DIVELIGHT COMPANION

Page 3 The CANISTER
With the advent of NiHM. canisters suddenly gotten a whole lot smaller. 2" PVC plumbing or conduit PVC (that's I.D., plumbing and conduit are described by the I.D. not the O.D.) works great for a 4.5Ah, and 2 1/2" for a 9Ah. However 2 1/2" is an oddball size in plumbing PVC and while it exists can be hard to find. Not to worry. - 2 1/2" is a standard size conduit PVC (as is 3 1/2") and available from electrical supply houses and many hardware stores. Higher strength schedule 80 is also available though not as easily, and might be preferable for a a light that will be going deep.
Delrin, whichgets only a passing reference in the book, because it is too flexible and prohibitively expensive for larger canisters, becomes well worth considering for smaller ones if one has the ability to bore it out - since Delrin cannot be bonded, usual practice is to make a one-piece canister from sold. This raises costs but results in an extremely rugged canister. Enough Delrin rod to make a 4.5Ah or 9Ah canister costs $20-30.

Page 4 THE LATCHES
Halcyon has recently started using a modified canister latch which has a built-in locking tab so it won't come undone accidentally. Halcyon seems to be using the HC833142LALBSS, but McMaster-Carr doesn't stock it and it is available from Neilsen-Sessions only in quantities of 100. There's a similar latch, the HC83314-42LALBSS which is available in small quantities via the on-line orderform on the NS website or from McMaster-Carr, part #1794A55, the only difference being that the bottom of the lever curves out instead of in and will have to be rebent for UW use.

Page 10 THE WATERTIGHT CORD GLAND
Originally the book recommended using teflon tape to seal the threads where the gland screws into the canister top. The problem with this is that when teflon tape is used it is very easy for someone not experienced at working with acrylic to over tighten the gland,which can cause hairline cracks or worse, and it's just about impossible for a book to explain just how tight it should be. So we are now suggesting a slightly different approach, using silicone sealant on the threads, and tightening the gland only enough to snug it firmly in place enough that it won't come loose with normal use - say, 1/2 turn past finger tight. Some of the newer lights are using straight-threaded glands with O-ring seals, such as the Agro gland mentioned below.

The elegant strain relief-less SS gland Halcyon was using does not seem to have worked out well in the real world, and Halcyon has switched to a SS gland with a built in strain relief that looks very similar to the Agro glands used by Salvo. Agro glands are available from Maryland Metrics in the US, #1017.52 seeming to be the most popular for divelights.

Page 12 The CORD
Pirelli cord does not seem to be available anymore. Dive Rite seems to have switched from the rugged but stiff Oflex Classic 100 "gray wire" to a much more flexible reddish silicone cord, SILFLEX SIHF CE. Other divelite companies seem to be using Ozflex FD 890, a high-flexibility PVC cord. Both are made by Lapp and can be ordered from electrical suppliers.
UPDATE the red DR cord turned out to be a disaster being too soft and flexible. It tended to break where it goes into the glands (DR does not use strain reliefs!) and puncture elsewhere. DR first tried to fix the problem by going to a larger gland with an integral strain relief, then switched plain ol' hardware store SJOOW, which Salvo has been using all along.

Page 15 THE REFLECTOR
The Lowel Light part number given for a 5" video reflector, part #01-15 is wrong - LL doesn't seem to make a suitable reflector. A better choice is Lumedyne reflector 091W (the W is for wide - there's a narrower beam 091 too) which lists for about $23 and can be ordered from B&H Photo (big NYC/Inet photoshop) for about $16.
Recently rumors began circulating of a new super-reflector, with a tight, clean, laser-like beam originating with the Euro EKPP, and sold by the DIR Zone in Germany. The superior beam seemed to be due to two factors, an extended shroud that extended 1" (25 mm) or so beyond the reflector bowl to reduce back-scatter, and a slightly larger, more precision reflector. Most US divelight manufacturer quickly lengthened the shroud on their #3 Super Spot goodman handles to match, and one or two even sold them to the unsuspecting as "EKPP reflectors". The genuine EKPP goes for about $100 without the handle, and appears to use an off-the-shelf Dev Pein #230-938 reflector cup.

Page 16 DIFFUSERS new text
Diffusers, usually made of translucent or textured plastic, are sometimes used in front of a light to widen and even out the beam. They waste a lot of light, but do a great job of smoothing out irrgularities in the beam, and especially filiment shadows, and are easily and cheaply added to an existing light, making them idea for a dual-purpose or temporary setup.

The easiest way to make them is from the prismatic acrylic diffusor panels used with fluorescent light fixtures, and available at most electrical and building suppliers for about $6. A scrap of this material, duct taped to the lighthead, can magically turn a #3 Super Spot or spot beam MR16 into an acceptable video light. For a more professional job a PVC collar can be made that slips over the lighthead, and the diffuser cemented to it.

Page 18 LAMP COVERS new text
We've finally found a direct source for test tube lamp covers - Reeves Glass in Florida. 1" x 3" TTs as used on most 18 to 24W HIDs run about $6 each, and larger sizes, like 2" x 3-1/2" for an auto HID or 200W metal halide about $25, with a minimum order of only $35.

FUSES new text
Some readers have wondered why we don't mention fuses. The reason is, a fuse is far more likely to cause a problem, by cutting out when it shouldn't, than to avert one. Also, the situations when fuses can be helpful tend to happen on the surface where the results, while irritating and expensive, they are not life-threatening, where fuses seem to like to blow at the deepest, darkest point of the dive.

Since canisters lights have traditionally been used by serious divers, cavers and techies, doing critical dives, they have generally avoided fuses seeing them as just one more possible failure point. As more divers begin to use canister lights for recreational dives, or video lighting, where a fuse blowing is only a minor inconvenience, we'll probably see more fuses turning up on divelights. But its worth noting that most of the problems a fuse might avert can be equally well prevented by a little extra care in maintaining and handling the light. Circuit breakers are a particular problem, as many of them are not happy in high moisture environments.

Page 32 OTHER BATTERIES
NiMH batteries have been dropping in price and becoming more available in larger capacities, making them increasingly worth consideration for divelight use. They are very similar to nicads, and come in the same sizes and configurations, but with two important differences: they hold about 50 to 100% more electricity than a nicad of the same size - from 1800 mAH for AAs, up to 9000 mAH for Ds - and are pretty much immune to the memory effect. This means they can be partially discharged or partially recharged with impunity, a very useful capability in a divelight. As one UW photographer put it ³NiMHs are everything nicads were supposed to be, but arenıt.² NiMH batteries are so superior to nicads that, in my opinion, thereıs no reason to consider nicads anymore unless one gets a deal on surplus/used nicads.

Anyone thinking of using NiMHs should know that there are some oddball industrial size NiMH cells that work much better than the standard flashlight sizes. One of the most useful is the 4/3 A (or 4/3 Fat A which are just tiny bit wider). These look like oversized AAs, and can be found in capacities of 3800 to 4500 mAH, for $5-7 per cell with solder tabs for assembling into multi-cell packs. Most of the commercial divelight manufacturers are using them as the building blocks for all their NiMH canisters. The usual configuration is in a series of 10 or 11 cells to give 12 or 13.2V depending on the bulb/ballast being used. For higher capacities several series are paralleled together - Halcyon (as EE lights are now called) uses two stings of 10 in its 9 AH NiMH and 3 in its 13.5 AH - and the 9 AH canister is only about 2 1/2" (60 mm) wide by 11" (275mm) long!
However it's always a good idea not to use two cells if one will do, and as larger cells come down in price it's likely we will see more divelight manufacturers using fewer, larger, cells for their higher capacity canisters.

NiHM batteries need the right charger to perform their best, preferably a smart charger that senses full charge and automatically switches into a "float" mode to maintain it, and finding suitable chargers that can handle a multi-cell pack has been a problem. One off-the-shelf solution that works well for smaller packs is the Maha 777 which can charge 2, 4, 8, 10 and 12 volt packs at 600 to 800 mA. If you hurry you may find a bargain in the Radio Shack 23-250, which is almost identical to the 777. RS has discontinued it, but some stores still have them on the shelves, at an unbeatable mark-down price of $15. If higher charging rates are desired its also possible to build a charger, using a cheap wall transformer and a smart charger IC chip like the Maxim 712/713 or the TI/Unitrode BQ2004E/H. These require soldering up a circuit board with the other necessary components to make them work. A shortcut is to buy an "evaluation board" from the manufacturer, for $50 (TI) to $240 (Maxim), which comes assembled with all the other necessary bits and requires adding only a power supply and connectors. Both boards can be configured to work with a thermistor in the battery pack for detecting full charge.

 

Lithium Ion batteries have been dismissed in the past as unsuitable for UW use as they tend to combust or explode if they become wet. Nevermattter, though - the cave diver's lust for more amp hours is greater than his fear of explosion, and several manufactures have started using them in their canister lights.
lithium ions are VERY fussy and must be carefully monitorer.s

Page 52 BONDING ACRYLIC
For solvent bonding of acrylic the right stuff to use is either ethylene dichloride or a solvent glue (usually ethylene dichloride-based) made specifically for acrylic. Some sources say to use acetone, but it is harder to use and doesn't do nearly as nice a job. Solvent glue is most easily available from plastic supply houses or display companies; the all-purpose cements available from industrial suppliers are usually too viscous.

Page 75 TEST TUBE LIGHTHEADS
There's a dumb goof on older books on page 75. The directions for figuring O-ring groove depth say "To figure out how deep to cut the groove, just measure the OD of the tube and subtract .156". That should, of course, read ID not OD. The previous paragraph has it right, which may be why no one has complained about it yet.

Page 77 PVC PIPE FITTING TT REFLECTOR HOUSING new text
It's also possible to make a very servicable reflector housing for the Lowel Light reflectors out of a 2 1/2" PVC pipe cap and a scrap of 1" PVC pipe. The cap should be a fairly deep one with a domed end - there are some short flat-ended ones that are too shallow to work. A standard 1" slug is a sloppy but usable fit in 1" PVC, but the cap must be modified to accept a standard 2 7/8² reflector. This can be done by boring out the end of the cap to fit, or by heat forming the cap to enlarge it. Boring is easiest done in a lathe, but can be done with hand tools or a rotary grinder.

It's also possible to slightly cut down the reflector so it will go in an unmodified pipe cap. Whatever method is used the pipe cap should be shortened first to fit the reflector. To heat form the cap, heat it until it is soft and then force it down over a suitable rounded object - I use a tin measuring cup - and hold it there until it cools. Actually, since there's a limit to how much you can bore out the cap, or how precisely you can expand it, it may be easiest to combine methods and expand the cap somewhat, then trim the reflector to fine tune the fit.

Once the reflector has been fitted to the cap, or the cap fitted to the reflector, bore or cut a hole in the end of the cap just big enough to fit the pipe stub, and cement it in place. A short reinforcing collar cut from a 1" pipe cap or union and cemented on where the stub meets the cap will make the joint considerably stronger.

1" pipe is a little on thin side to hold machine screws for mounting a goodman handle, so it may be safer to use a clip as for a broomhandle or MR16, Or cement a doubler on to gain a little more thickness. The doubler can be cut from a 1" pipe coupling, cap or adaptor, or made from a scrap of 1" or 1 1/2" pipe heat formed if necessary for a good fit. The doubler can then be ground/filed/sanded flat to provide a good mating surface.

By the way, when a diffused beam is desired, a white PVC cap makes a pretty good reflector all by itself - the Benthos RUV (remote underwater vehicle) people use a similar PVC fitting for their reflectors. Fit it with a diffusor and it makes an acceptable video light. And if you should ever lose the reflector cup, a white PVC housing will make a much better substitute than black acetal or gray hardcoated aluminum.

Chapter 9 INSTANT DIVELIGHTS - additional text
One particularly nice light for this kind of conversion is the Mag-Lite. Mag-Lites are ruggedly made from anodized aluminum with O-ring seals throughout, sell for $15-20, and accept MR-16 reflector bulbs as if they were made for them. The flimsy lens must be replaced with something heavier, but there's room - just barely - for a 3/16" (5 mm) one. The two weak points are the switch cover, which is a simple snap-in rubber cap and not able to handle pressure, and the end cap, which is a bit thin to hold the cable gland. Usual practice is to make a plug for the switch hole, epoxy it in, fill the recess with more epoxy, then cement a sleeve made from a section of bicycle inner tube or other tubing over the hole to provide redundant sealing. Using a good anaerobic sealant on the gland threads, and maybe a nut on the inside as well, takes care of the endcap, but one is still left with an ackwardly long handle.

Marshall Reymard and his accomplices from the South Shore Neptunes have come up with an even better solution - throw away the stock handle/battery compartment and replace it with a Delrin stub threaded to fit the head. This eliminates the switch hole, provides plenty of meat for the cord gland, and allows making the handle any length you please. The stub has to be bored for the socket, and threaded to screw into the head, so it's stretching the definition of ³instant², but it's a straightforward job for anyone with access to a lathe. The C-cell Mag-Lites are preferable to the Ds because they have the O-ring groove in the head rather than the handle, so less machining is required, and the thinner handle is more comfortable to use or adapt to a goodman handle. The lighthead has been down 200' in the pressure pot without problems, and the Delrin seems to hold up fine with a 50W bulb. (see the GALLERY for a picture).

Page 120-121 10W HID LIGHTHEADS
Halcyon has gone over to a completely new 10W design, which puts the reflector and bulb is a dry, focusable housing in order to eliminate the little "extension cord" that the old version used to allow separating the ballast from the bulb. Apparently, over the course of production Halcyon found wide variation in bulb/ballast performance, with batches of them coming though that would not reliably strike with the bulb not plugged directly into the ballast. WA says the QC problems have been solved and Dive Rite continues to sell a test tube 10W similar to the old Halcyon, but Halcyon apparently got burned badly enough that they decided to take no chances and came up with the new design. The old 10Ws either work or don't work, so if you have a good one don't worry, but if you are thinking of building a 10W that separates the bulb from the ballast keep the distance as short as possible, and it might be a good idea to test the bulb, ballast and whatever wiring you plan to use, before you build the whole lighthead.

Sartek has also come out with a focusable 10W HID that, like the Halcyon, puts the reflector inside the dry housing.

 

UPDATE Welch Allyn originally had 4 different 10W ballasts to suit different battery configuration. Now they have come out recently come out with a single #B10R001 ballast which is regulated and usable with a wide range of input voltages so it replaced all the previous versions! They also raised the wattage a little, to 14W, so puts out noticably more light. It can be used with both the new M10P003 bulb, or any of the older 10W variants. W-A now has online ordering for alll the HID bulbs, though with a $100 minimum, the new ballast is $30 and the bulb $50.


There's a promising new 21W HID bulb and ballast from Brightstar in China that Dive Rite and Salvo are now using, which comes in both barebulb and MR16 versions. It uses a very compact bulb very similar to a halogen bipin but with an integral base, and is extremely rugged. Most lights using it do not bother to pot the ballast, which considerably simplifies things. It's also much cheaper than the W-A bulb.

Even more interesting, Sartek has found that the Brightstar will run fine on a W-A ballast, and has come out with a $125 upgrade for the 18W Sartek HIDs, and, possibly, similar lightheads. UPDATE there is a new version of the Brightstar 21/25W bulb which will NOT work on a WA ballast!
Brightstar also makes a 10W bulb and regulated ballast, but with a barebulb and MR16 version available, which should be perfect for a low end DIY HIDs. They also make a 10W that plugs into the ballast which is almost indentical with the W-A but not interchangeable. Brightstar components can be ordered directly as samples from Brightstar (it helps to have a business name) though they require payment by cable which can be complicated. There have also been some group buys, on the HID divelight list and elsewhere, and the components have been showing up on eBay too.

Trail Tech is a maker of lighting for motorcycles and ATVs who sells several HID light assemblies at such good prices that many divelight builders have found it cheaper to buy and cannabalize one them than try to buy individual parts elsewhere. Their 10W uses the usual Welch Allyn parts, but they have a 30W (sometimes described as a 35) which uses an ballast of unknown manufacture along with a bipin HID bulb similar to the Brightstar. Unfortunately, the 30W ballast seems to run hot enough as to make it difficult to cool adequately in a UW lighthead - in Delrin heads especially the thermal protection seems to shut the ballast off after 30-40 minutes of use, and similar problems have been reported in aluminum lightheads.

Higher-wattage ballasts are becoming available that overdrive HID auto bulbs - Auerswald in Germany has one that can be switched between 35 and 50W.

Halcyon has also redesigned its killer 200W HMI, making it considerably smaller, mainly by switching to NiMH batteries. They've also immersed the ballast in an exotic, expensive and hard-to-find heat transfer fluid that keeps the ballast cool enough that it won't shut down after 45 minutes the way the old ones did.

Former EE and Halcyon light guru Barry Miller has now gone out on his own and is selling top quality divelights under the Salvo Diving label, including an updated, reliable version of the old test tube 10W. Barry has always been DIY-friendly, and Salvo has good prices on parts. UPDATE Salvo is no more, Barrry having closed shop and vanished.

LEDS and

The King is dead, long live the King! LEDs continue to evolve like fruit flies, and the Luxeon Stars we mention in our LED update have already been displaced by the SSC P7 12W emitter. It's a single emitter, and can be used with a more-or-less standard reflector rather than the dedicated optics the Luxeons required, and puts out a bit more lumens than the10W HIDs, and just a bit less than the 18/21Ws. Like the Luxeon, it needs an insulated heat sink, and either a driver circuit or a 3 to 3.6 V battery pack.

 

 

Back to TOP

THE OXYGEN HACKER'S COMPANION

UPDATES

PM 128 Digital Panel Meter availability The PM 128 is one of the two digital panel meters recommended in the book for making a simple, low-cost oxygen analyzer, and sold by a number of mail order electronics suppliers. The PM128 mentioned in the text has, however, been superseded by a new version, the PM128-A, which uses tiny SMD (surface mount) resistors. The board is also laid out slightly differently so the instructions in the book for the PM128 won't work for the PM128A.

The little SMDs are a bit trickier to work with, being smaller, but since they are soldered to the surface, rather having a leads running through a hole, are actually easier to remove (the easiest way is to use two soldering irons, one on each side, and just flip the resistor off; otherwise, work from side to side heating the solder and prying the resistor up gently with a little screwdriver as you do).

On the Original 128A resistor R3 is the one to remove, as shown on the LH diagram above. One lead goes to the bottom solder pad where the resistor was and the other to the top solder pad of the trimmer pot to the left of the resistors (the trimmer pot is left in place) as shown above. It's much easier if you tin (heat and coat with solder) the ends of the leads before you try to solder them to the tiny pads. Other than that the conversion is essentially the same as described in the book, though it may require a little more fiddling to get the right combination of resistors. Using a 100 ohm linear pot and a 1k 15 trimmer pot for a series resistor instead of the audio pot and fixed resistors originally specified makes calibrating much easier . The DMD3500, which the book recommends over the PM 128, does not seem to be available anymore.

NOTE:
There seems to be a new, slightly revised version of the PM 128A. The only difference is that the trimmer pot has been turned around so two legs are at the top and one on the bottom instead of one on the top and two on the bottom as before and the resistors have been rearranged. It is converted just like the old one except that the lead is soldered to the top left leg of the trimmer pot, that is to say, the one closest to the upper corner of the board, and while R3 is still the resistor to be removed, it is now the 2nd over from the on-board trimmer pot rather than the 3rd over. It is labeled as R3, but a clip partially covers the writing. Both changes are shown on the left hand diagram above.

NOTE 2: We've had a few reports of readers who have bought what was described as a PM128A, from various vendors including the one listed in the book, All Electronics, but have gotten something entirely different so that the directions in the book or the ones on this website don't work. These appear to be panel meters from other manufacturers that some US suppliers substitute for PM 128A, and in some cases, actually re-label as such. This can be very confusing when you try to convert it and nothing is where it is supposed to be! if you are shopping for a PM 128A be sure to compare the pictures of the PM 128 in the book with the one you are thinking of buying to be sure it is the same.

To try and simpify things, we are now stocking the "real" PM128A so readers can order a known-good one and avoid any surprises. See our Oxy Analyzer page for details

 

The DMD 3500

Early editions of the book recommended the DMD 3500 panel meter by Circtest in Canada. It's actually easier to convert, but harder to find and more expensive than the PM128 so we don't cover it anymore. However, should you have one, it is easily converted by removing the 24k (red and yellow striped) resistor located just below the trimmer pot on the board and soldering on two wires connected to a 100K or so pot. A resistor (or several resistors) in series with the pot works even better, since it allows fine-tuning the range. Itıs hard to give exact values, since different sensors may require different ranges, but a 15K resistor and a 50K pot in series worked well on the prototype. A 30K pot with a 20-30K resistor would narrow the range and improve resolution, but the 50K is much easier to find. Increasing the value of the fixed resistor moves the range upward, and decreasing it moves the range downward.

Oxygen Sensors:

Teledyne has always made good sensors, but at the time the OXY HACKER was being written they were quite a bit more expensive and harder to find in small quantities than the CAG 250 specified in the book. Since then Teledyne has made several significant improvements in their sensors, and the price/availability situation has improved (thanks mainly to Oxycheq!) a great deal, giving the Teledyne R-17 MED an edge over the Maxtec (formerly Ceramatec) MAX/CAG 250. Teledyne sensors are the ONLY choice for rebreathers, since they have a hydrophobic membrane that protects the sensor from water - the Maxtec sensors are not happy in high pressure or high humidity situations.

Stainless Steel and HP O2:

For metal parts, brass in definitely the metal of choice when it comes to HP O2. It doesnıt spark and doesnıt (for all practical purposes) burn.

Aluminum and titanium are definitely out - they ignite easily and burn violently in the presence of HP O2. Steel is not a good choice either - it burns, as anyone who has ever used a cutting torch knows. But the big problem with steel is that it sparks so readily, making it a prime potential source of ignition. SS falls somewhere in between. While it doesn't spark or burn nearly as easily as steel, it is not completely immune. The catch is, even good brass fittings are often rated only for only 1500-2000 psi, (100-133 bar), and rarely for more than 3000 psi or so (200 bar). For higher pressure systems, then, there really isn't any affordable choice but stainless steel, and just about everyone uses it.

It's hard to get a definitive answer on how suitable SS really is for HP O2. The first editions of this book cautioned against it, based on recommendations in several NASA publications, but dig a little deeper and you find that for every one that says not to use it there seems to be another that says it's OK. While the NASA Glenn Safety Manual says that only iconel or monel are suitable for high pressure O2, the NASA Safety Standard For Oxygen Systems NSS 1740-15 says that "Stainless steels are used extensively in HP O2 systems" and that "few problems are experienced with the use of SS storage tanks and lines" but that "ignitions have occurred in SS components such as valves at high pressures and high flow rates", though they are probably talking about pressures much greater than the average nitrox whip sees (both documents are available on the web). And the CGA specifies that valves for O2 tanks should be made of either brass or SS.

http://www.hq.nasa.gov/office/codeq/doctree/871915.htm

http://osat-ext.grc.nasa.gov/gso/manual/chapter_05.pdf

Continuous Mixing
Page 108 4th Edition

Bauer seems to have changed its mind about using Bauer compressors with continuous mixers, and is labeling most of its compressors as suitable only for air. This repudiates an earlier letter from Bauer Norfolk which was the basis for the information in the book, that Bauer said it was OK for up to 40% if a proper oil was used, the design speed not exceeded, and the output pressure kept below 3600 psi. Note that this doesn't necesssarily mean continuous mixing with a Bauer is dangerous or will harm the compressor - there are probably hundreds if not thousands of Bauer and similar compressors all over the world being used with mixers and membrane separators - just that Bauer no longer approves it.
Coltri still permits it, and is even marketing membrane separators now, that pump nitrox through the compressor in the same way as a continuous mixer does.
Airetex, Alkin's American distributor says "Continuous blending is safe with any high quality compressor in good working condition as long as the mixture is kept below 40%. All compressors for breathing air (nitrox or not) should use good high quality synthetic breathing air compressor oil.We do not void the warranty for using Nitrox."

 

CORRECTIONS

Early copies of the OXY HACKER may have certain errors in them. If you have one, you ought to check for them and update as required. Most of these seem to be the result of glitches that crept in when we redid some of the tables and graphics so we could print the book from disk on a Docutex printer rather than scanning hard copy.

Page 112
On the OTU table on page 112 in some early copies of the 3rd Edition of the OXYGEN HACKER'S COMPANION and the Upgrade, the 100% column is the same as the 80% column, and the figures in it are the ones for 80% rather than 100%. We are terribly embarassed that we missed that (and that none of our readers picked it up!) , as the error is blatantly obvious when you note that, according to the numbers, 33 feet at 100% gives a PO2 of 1.6 rather than 2! Here are the proper numbers. If you've got one of the bad copies, please write them in.

100%. OTU PO2
2 1.08 1.05
3 1.16 1.1
5 1.24 1.15
7 1.32 1.2
8 1.4 1.25
10 1.48 1.3
12 1.55 1.35
13 1.63 1.4
15 1.7 1.45
16 1.78 1.5
18 1.85 1.55
20 1.92 1.6
21 2 1.65
23 2.07 1.7
25 2.14 1.75
26 2.21 1.8
28 2.28 1.85
30 2.35 1.9
31 2.42 1.95
33 2.49 2
100%. OTU PO2

Page 45
There's another fairly obvious typo on page 45 of some copies of the 3rd Edition, on line 4, where after doing the math to conclude that a 17% mix is necessary, in the final line, the mix is given as 19/45/38, which not only doesn't agree with the rest of the text, but doesn't add up either - it's supposed to read "the answer is a 17/45/38 trimix (or a 17/45 as such mixes are often referred to, omitting the FN2). "

Page 46
On the Trimix Calculations worksheet , the fudge factor has mysteriously become a plus rather than a minus. The 10% fudge factor should be subtracted from - not added to - to total O2 to be added to the mix. The operation is correctly described in the text, though. The text at the bottom should read "Minus compressibilty fudge factor", and the final O2 figure be 175 psi.

Page 98
The analyzer schematic on page 98 of early 3rd Editions lost a couple parts. Missing are two of the three solder tabs on the calibration pot. The missing tabs should be on either side of the single tab shown, the upper one so it connects to the wire from the gauge and the jumper resistor lead , and the bottom one is unused. If they are missing from your copy, we suggest simply drawing them in.


Corrected Schematic from page 98

BTW, it doesn't really matter which two tabs are used, as long as one of them is the center tab. Which of the outer two are used will detirmine in which direction the pot works; I usually just hook it up at random then change it if I don't like the results.

There's also been a little confusion about the jumper resistor shown in the same diagram which runs between the two tabs on the pot - this is normally required only with the PM128, as explained in the text.

THE OXYGEN HACKER'S COMPANION 4th Edition:

Page 40
We've gotten a few emails from readers who want to know why GIII's article calls for a fudge factor of 17% while the text says 10%. The reason is simple - it's a fudge factor! If we could tell you exactly what percent to use, then it wouldn't be a fudge factor anymore. The exact percent depends on a whole bunch of factors quite beyond our control, such as what order you add the gases, how fast you add them and how long you let them cool before checking the pressures. That may sounds frighteningly inexact, and may put you off trying to mix, but it's not nearly as critical as it might seem - the exact % of He really isn't that important, and as long as the FO2 is coming out OK you're going to be OK. If it isn't, adjust the fudge factor until it does. Once you get mixing you'll quickly learn what works for you.

Page 89-95 WHIP and FILL VALVES
Graingers no longer carries teflon/braided SS hose, and M-C has changed the part # for theirs - it's now 4468K202. M-C also has "U-Build" bulk hose and user installable JIC and NPT fittings, handy if you want a hose with mixed ends. Either way, don't get seduced by the SS end fittings - brass ends are better since they act as heat sinks, and a lot cheaper. The Graingers/Sun needle valve mentioned is no longer listed either - in its place we suggest M-C's #7833K74, an $18, 3000 psi brass needle valve. Actually, we are buying most of our valves on eBay now, lovely used and suprlus SS Swagelok and Parker for pennies on the dollar, but that's not much help if one is in a hurry.

Page 96 QR FITTINGS
Parker ST-series quick release fittings are the QRs of choice for many gas mixers, being cheap, readily available from McMaster-Carr, and having unconvoluted straight-thru bores. The only catch is that for the male plugs have been available only in steel, which is unsuitable for O2, and SS, which is expensive. It turns out Parker does make a brass male fitting - it's just that M-C doesn't list them, so you'll have to find a "Parker Store" to buy them from, although a few people say they've been able to buy them from M-C even though they aren't in the catalog.

Page 127 HELIUM ANALYZERS
The He analyzing game has just heated up (a good pun, if you know how these things work) with the arrival of Oxycheq's new helium analyzer. It performs similarly to the Atomox/DiveRite unit, but is housed in the same tiny case as the Oxycheq Expedition, making it by far the most portable unit around.

 

New Text POOR MAN'S COMPRESSOR
One of our readers recently came up with an interesting alternative to a compressor, of stunning simplicity. He found that his gas supplier had nitrogen in 580 cf/ 6000 psi bottles for about $18 a fill, and simply added a N2 cascade to his O2 and He cascades so he is, in effect, PP-mixing his air.

This seems a bit round-about at first, but when you look at the number it starts making sense - figuring on 8¢/cf for the O2 and 3¢ for the N2, it would cost him about $4 to fill a 100 cf tank with 21/79 nitrox, aka air, this way - about the same as a dive store fill. With nitrox and trimix the "air" percentage and hence the cost goes still lower so it's actually cheaper than having a dive shop top up the tank.

So why isn't everyone already doing this? It takes high pressure gas to be really practical - I've know people who tried it in the past, but with lower pressure N2, and found they were always having supply pressure problems and often had to settle for reduced fills. So the 6000 psi nitrogen is the key since it provides enough headroom to allow getting full pressure (or better) fills without needing an inordinately large cascade. It's also impractical to make gas this way if you aren't already mixing since you have to go through the whole PP rigamarole, including analysis, just to make air. But if it's nitrox or especially trimix you want it's no extra work at all.
Up-front costs are very low because the most expensive components - the tanks - are leased, though a reliable regulator is a necessity to protect against extreme overfills.

This might also be a good dodge for anyone living far from the nearest dive shop, not for full time use, but just as a backup to allow banking enough gas to do a couple fills anytime one wants to do a dive but can't make it to the dive shop. It also makes an interesting solution to the where-to-get-your-homebrew-topped-up problem if one doesn't have a homebrew-friendly diveshop nearby.

Practicallity if very dependant on availability- when I called up my gas supplier to ask about HP nitrogen I was quoted over $200 a tank!

Back to TOP

Return to Airspeed Press main page