Airspeed Press is proud to have been the first to come out with plans for a cheap, easily-built DIY Oxygen Analyzer. Back in the dark ages before the OXYGEN HACKER'S COMPANION , O2 analyzers cost $600 and up, being made (and overpriced) for the medical market and, as a result, many techies and even diveshops didn't have their own analyzers - not a good situation! So when we came up with the OXYHACKER <$100 ANALYZER, in the 2nd edition of this book, it was a real breakthrough for the tech diving community.

The key to the OXYHACKER ANALYZER, and all its imitators, is to, rather than trying to build the electronics necessary to read the sensor from scratch, just use an inexpensive off-the-shelf digital panel meter so all the electronics are in one unit, and it is only necesary to add leads for the sensor, battery, and calibration pot. Sounds easy, but the trick was to figure out how to trick the meter into reading directly in percent of oxygen, and (since the manufactures of these panel meters don't give out the schematics for them) we had to go by trial and error and destroyed a lot of panel meters figuring out how.

It was not until five years later that lower-cost "consumer-grade" units made specifically for recreational divers began arriving on the market. However, these did not make the OXYHACKER ANALYZER obsolete! The main problem with the cheaper consumer-grade analyzers is that, since the sensor is the most expensive part of the analyzer, and these units are built to a price point, they skimp on the sensor, using shorter-life sensors. The builder of the OXYHACKER ANALYZER, by saving money on the "box", can use a top-grade. long-life industrial or medical sensor like the Teledyne or Maxtec, and still spend considerably less than a commercial unit would cost - typically it costs about $100 to build an OXYHACKER ANALYZER, and this includes a top quality. long-life sensor!

Also, when the sensor needs replacement some years down the road, the buyer of a commercial unit is stuck buying the manfacturer's dedicated sensor for that analyzer (assuming they are replaceable, or even still available) even though, since sensor technology keeps improving, much better ones may then be available. Since the OXYHACKER ANALYZER can be easily recalibrated to read any sensor, the owner has his or her choice of any sensor on the market, and can shop around to get the best sensor at the best price. Also, if anything goes wrong with the analyzer, it is easily fixed - replacing all the electronics inside the box will cost only around $12! And finally, as most hard core techies are aware, you can't have too many analyzers. You need one for your dive box, one for the workroom, one or two for your continuous mixer, a spare, and maybe a few more for your buddies. That can add up to a lot of money, unless you DIY.

The only problem we've had with the plans in the book is that the method of adapting any panel meter is very specific to that model, and the manufacturers keep coming out with new models or redesigning the old ones. Often when they do they list the new one under the old part number. Or a vendor will sell another make of meter under the same part number when they run out of the genuine one. When this happens we start getting complaints from people who buy a panel meter that is supposedly the one mentioned in the book and find it will not work. We try to keep ahead of the changes, updating the directions in the book and on this website, but still hear complaints from people who order something advertised as a PM-128A panel meter from a supplier, and get something completely different.

What you want is the PM128A LCD 3-1/2 digit meter made by Colluck in China and usually labled "C&C" - see the picture of the box below. We tried selling them for a while, but found they were very slow movers and so we no longer do. Also, there are several vendors now selling them on Ebay with shipping included for less than we could. HINT: if searching for these on Ebay, try variations such as "PM 128", or "128 meter" etc., since they will not always be listed by the full name.

Ask the vendor if they are genuine PM128A's before you order, and if possible confirm that they are by comparing a picture showing the back of one with the pictures below.



Common problems:

- Using a 100k ohm rather than a 100 ohm pot. This is real easy to do since 100 ohm pots are pretty rare, and if you are not an electronics person it is easy to overlook the signficance of the "K".

- Check all the wiring. Is the "+" side of the battery connected to the "+" marked solder pad? Are you sure you haven't mixed up the battery and the sensor leads?

- Low batteries. These gauges go through batteries faster than you'd think, and can start reading erratically once they get too low (which is when I usually go "borrow" the one from our smoke detector, always a bad move, burning to death aside, as they are usually almost worn-out too).

- Shorting where the wire lead is soldered on the pad where R3 was. It's pretty crowded there, and little, almost invisible whiskers or bits of solder can make unwanted connections between the pads. Taking an Xacto or other small sharp knife and and gently scraping the area in between the pads can prevent this.

- Cheap battery clips making bad connection. Sometimes these need a little squeezing with pliers to tighten so they grip the battery well.

- Recheck all joints. Did you use flux (in addition to the stuff in the solder)? It makes it much easier to do a nice job and avoid false joints.

- Bad sensor. They do occasionally come dead from the factory. The easiest way to doublecheck sensor function is by reading the sensor directly with a DVM or multi-meter as explained in the book. A functioning sensor should read 7 to 17 mv. on air. Voltage will drop off a bit over the life of the sensor, which is why there is a calbration pot on the analyzer. However, once it drops below a certain point, readings may become erratic, even though the sensor can still be calibrated. One tip off will be that the sensor will calibrate, but not hold that calibration for long. Note though that a low battery can mimic a bad sensor - always suspect the battery first since it is a lot easier and cheaper to fix!

- Bad Meter. It's always possible that an individual meter is defective, but they are very rugged and quality control seems very good. In the dozens we have modified ourselves here we have yet to find a bad one. The problem is,there really isn't any way to test the meter itself unless you have a spare one to cross-check it with, and if the meter isn't working, there is no way to tell if it was defective to begin with, or damaged or improperly wired while being modified. Since the latter is far more likely (which is why we don't warranty them), it's always a good idea to recheck everything carefully several times before giving up on the meter.

- Gauge reads minus voltage. Reverse the leads connecting the sensor to the meter (usually, the tip of the plug that connects to the sensor should go to the Gnd pad.

- Decimal Point missing or in wrong place. Install a jumper on the upper right of board as explained in the directions that come in the meter. Ordinarily, it will go in the upper spot.

Calibration pot turns in the wrong direction. Remove the wire or resistor from on the outer solder leg of the calibration pot, and solder it to the one on the side, leaving the middle one untouched.

- Obsolete Plans. A lot of our readers, for some reason, whenever they build one of the gadgets in our books, are so proud of it that they immediately put up a webpage showing how they did it, and give away all our secrets in the process! If they are nice about it, they will at least credit us and give us a link. If they aren't, they won't, since they often want to pretend they came up with it all by themselves. We don't worry about it to much, because we know they are unlikely to update their plans, and their site will quickly become obsolete. But it is worth mentioning, that if you have built an analyzer based on plans off the net, they may very well be obsolete, and no longer work with the current crop of meters or sensors.



- When a battery gets low, the meter will start acting up and showing odd or inconsistant readings which are impossible to miss. However, a week but almost-good battery may cause erratic or inaccurate readings which will not be immediately obvious When this happens, the usual solution is to find a fresher battery and see if it works better. Here's a cute way to add a battery tester to your analyzer to eliminate this problem. All it takes is a resistor and an intermittant-contact NC (normally closed, or "on") push button switch. The push botton is placed in series with the power on-off button, and the resistor soldered across the button terminals, so that when the button is not pushed the electricity is free to take the line of least resistance and bypasses the resistor, but when the button is pushed the electricity must go through the resistor. The extra resistance from the resistor puts just enough extra load on the circuit so that a weak battery will be easy to spot. To use it, calibrate the analyzer as usual, observe the reading, then push the button. If the reading changes, the battery is very near the end of its life.

- We had a customer the other day who wanted to put two meters into one box, and wanted to be able to run on either batteries or a 9V "wall wart" plug-in transformer. With a single meter, or two meters running off one battery, the easiest way to do this is to use an earphone-style jack, which has a built-in switch to cut out the power to the speaker when the earphone is in use, only wired so it disconnects the battery when the power is connected. Only problem with this, is that the sensor uses the same style plug, so it would be possible to plug the power in to the sensor port, possibily damaging the meter. So perhaps a better method would be to use a 2 way switch that would allow switching between the battery and the power supply. These switches come in two styles, On-On and On-Off-On; if the latter is used, which has an off postion in the middle, then it can also serve as the power switch.

With a double meter setup, the trick is to use a DPDT (double pole double throw) switch. This allows both of the meters to run off a single power supply when plugged in to it, and to run off their own separate batteries the rest of the time, so as to provide a measure of redundancy.

It should also be possible to use nicad batteries, wired so they are connected all the time, so they would recharge when the power supply was connected, however we have not tried this. It would require using a power supply that is well matched to the battery voltage, so it won't under or overcharge, and the analyzer case should be labled "Use Rechargeable Batteries Only" because if non-rechargeables were used they could overheat or explode. This is one application where nicads are probably a better choice than NiMHs, since they are much more tolerant of overcharging.


Back to Airspeed Press main home page