|
|
The Oxygen sensor cell has come a long way since its invention by NASA as a fuel cell to power its spacecraft. Its become a lot smaller and a lot more durable. There seems to be a lot of confusion as to what a cell actually does and whether it produces voltage or current (Both, cant have one without the other). There are also some controversial design aspects of the Inspiration's electronics.
What oxygen sensors does the Inspiration use?
Inspiration cell use controversies
Practical testing of your cells
|
|
The modern galvanic oxygen sensor is a miracle of micro engineering, physics and chemistry. In reality it is a battery that uses oxygen to oxidise lead which produces a potential difference. The cells are quite complex
At the top we have a PTFE membrane to keep the electrolyte fluid it. This membrane (Gore-Tex) also allows the oxygen in and keeps water out. It also happens to be hydrophobic
Next we have the Potassium Hydroxide electrolyte. This Is the ion transfer agent between the anode and cathode. It is the depletion of transfer ions in this medium which is one of the factors that limits the cell life. The electrolyte ensures maximum electrical contact between the cathode and anode
The cathode is made form a noble metal (Gold, silver, platinum, rhodium etc) with numerous perforations to permit the flow of electrolyte and its ions
The anode is made from lead. The lead is made from spherical particles and compressed to maximise surface area and is normally a dome shape covering a large area. The lead oxidises in the presence of Oxygen, which draws electrons from the electrolyte across from the cathode. The electrons gathering on the lead give it its negative charge
Below the anode is a flexible membrane (normally PTFE) again to allow for fluid volume changes (Not shown)
The electronics board is where the anode and cathode is terminated. A simple temperature compensation circuit is mounted here (temperature sensitive variable resistor), this changes the internal resistance of the cell to keep its output constant regardless of temperature. The circuit board also serves as a mounting for the molex connector. The board normally has holes in it to allow ambient pressure through to the flexible membrane. To waterproof the board it is normally covered in resin/wax (Conformal coating), this has been known to block the equalisation holes and stop the sensor pressure compensating. Always check all cells for vent hole blocking.
The cell acts as a simple fuel cell, in that the numbers of electrons at the anode are directly proportional to the oxygen availability (molecular, not percentage) across the ranges we are normally interested in. This property is called the linearity of the cell. If there is no oxygen there is no output. In air (20.9% O2) most normal cells output around 13mv when new. The useable range of output for a cell is approximately 13.5 - 7 mv. (when measured by a multimeter (infinite resistance)). The voltage is proportional to the surface area of the anode available to collect electrons (the potential difference), the current produced when there is free O2 ions is proportional to the flow of ions through the electrolyte. Hence if the electrolyte is exhausted the cell can produce no current but still produce a voltage (this is known as current limitation) and is normally the result of old age of the cell
A great way to picture a cell is as a normal AAA size battery. It produces voltage/current when connected and it has a finite life as it consumes its internal components. Just like an O2 cell, deferent loads, usage patterns etc can effect how long it lasts and if it has enough power for the application. For example a battery that is getting dim in a torch will probably power a digital clock for many more months. Just like a battery, the bigger it is the more material is likely to be in it and the longer it should last
Most cells are rated to last 18 months in air. of course we use them at much higher O2 levels than that so it is expecting a lot for them to last that long. Most people change their cells yearly. Measuring their voltage output in air is a poor test as this just tests for total failure or consumption of the anode. The most worrying condition is current limitation where the cells cannot produce the required output for the O2 level they are exposed to, The only true way to test cells is to do a 100% flush at 6m at the END of a dive (when they have been stressed for a while) and make sure they can still reach 1.6 bar PPO2. That way you know they are responding correctly at 1.3 bar
Changing cells when they physically fail or the Inspiration reports them as out of range is very dangerous as this will happen a long time after they have degraded enough to not report elevated PPO2 levels.
|
|
Most oxygen sensors are of a similar size and output and therefore its not surprising that several will actually fit the Inspiration. So what one should you use. Well officially the answer is easy. The R22-2BUD from APD is the official cell and the only cell that can be used if you want to keep the CE certification while using the unit (and your warranty). Of course if that isn't a concern then there are 2 other main cells you may be interested in. But be warned if you send a unit to APD with other cells fitted they will be removed and you will be billed for the 3 new R22-2BUD cells (I can understand why they do this, but someone needs to point out that its an illegal practice under UK law. Its the same law that stops garages charging you for extra work on your car, without getting permission first)
R22-2BUD
Made
by Teledyne for APD and sold through APD, this is the standard cell for the
Inspiration. it has undergone numerous design changes and currently has better
temperature compensation and a hydrophobic front membrane to repel water.
Earlier versions have had problems with water on the front mesh membrane,
insufficient temperature compensation, conformal coating blocking the pressure
equalisation holes (See
http://www.therebreathersite.nl/ambient_pressure_holes.htm), bad soldering
and QA problems. The R22-2BUD still seems to have QA problems judging by the
number of cell problems on the forums. I personally out of 5 cells have had one
fail totally after 2 month use and one fail with 2 hours use, A third failed
after 30 dives. A 60% failure rate
is not good and these are all 2002 cells.
By the way there shouldn't be an O ring on the cell like the picture on the left has, especially if you are using the cell in the VR3 adapter
The R22-2BUD costs £55 from APD. If you are in the UK, this is the cheapest cell and the easiest to get hold of. APD normally keep a good stock so 3 day turnarounds are normal.
If however you are outside the UK then it may be cheaper and easier to get hold of the other cells
The temperature compensation on the R22-2BUD is identical to the R22D except that the BUD's thermister is mounted on the board and hence is measuring air temperature rather than the cell temperature. The R22D's thermister is in wires and is placed closer to the rear membrane
Teledyne R22D
The
brother of the R22-2BUD. This 10mV +/- oxygen sensor has a 6 second response
time and a 36 month life expectancy in air (very questionable). This is a replacement sensor for
the Inspiration Rebreather, CCR-2000. the Uwatec Oxy2, the KISS rebreather, the OxySpy and the VR3. It has hydrophobic membrane at the inlet, has a coated
board, and uses a Molex connector.
It is sold by Oxycheque (agents for Teledyne). Its thermister is mounted in foam close to the pressure compensation membrane (the circuits are the same other than that)
The oft asked question is what is the difference between the R22D and the R22-2BUD, the following is lifted from Teledyne
The R22D falls within the millivolt operating range required by the Inspiration.
The R22D has Teledyne's new patented temperature compensation circuit board.
The R22D has the same Molex connector as the R22BUD-2.
The R22D and R22BUD-2 both have conformal coated boards.
The R22D and R22BUD-2 are both made by Teledyne.
Other than a label, the only difference is the R22D has the newer, patented temperature compensation thermister locationInside Information from a conversation with John Lemm at Teledyne. The R22D and R22BUD-2 ARE the same cell apart from the thermister (which adds about 3% extra temperature correction). BUT the R22BUD-2's under go extra rigorous cyclic testing to ensure they can cycle repeatedly between low and high PPO2's. So in fact the R22BUD-2 is actually a higher specification cell than the R22D
PSR-11-39-MD
This
cell features a clamped hydrophobic membrane with a 6 second response time. The
temperature compensation pcb is constructed with both the components and traces
on the same side of the pcb. By using this construction method there nothing on the back side of the pcb. Therefore conformal coating is a not an issue.
The PSR-11-39-MD's pcb is also physically much closer to the rear membrane of
the sensor. It's thermistor is in physical contact with the rear membrane of the
cell providing excellent temperature compensation. It's weld free cathode
connection help increase the cell life. PSR-11-39-MD sensors are also about 20%
smaller than the OEM sensor. Customers have mentioned the smaller sensor allows
an easier installation and a neater configuration. The top cover being less
crowded.
For details see here Oxygen Analyser.
|
|
OK, so we now know that the sensor is a fuel cell and hence will generate a voltage/current if a circuit is completed to it and there is oxygen at the front face. The question is how do we measure that oxygen
Assumption 1) That the current is proportional to the amount of oxygen molecules present at the face
Assumption 2) That that the increase in current is linear for the range we are interested in (i.e. if we double the O2, we double the current)
Assumption 3) That we are removing the electrons off the anode no faster than the electrolyte can re-supply them
Assumption 4) That the voltage (mv) is proportional to the amount of available cathode and anode and will not change during the time we are using the unit between calibrations
So the simplest way of measuring the O2 level is just to connect a simple Ammeter onto the circuit. Measure the output in Air, measure it in pure Oxygen and put 79 divisions between the two. We can also extrapolate the units above 100%.
Calibration
The problem is that not all cells are the same and as they age the anode oxidises and the voltage drops. So we need some way of standardising where the gauge sits (the amount of swing per extra % of O2 should stay the same). so to do this we insert a variable resistor and use this to "calibrate" the gauge at a known level (Normally in Air or pure 02).
Some systems use manual calibration where you can select the control gas (like the VR3) others like the Inspiration use an automated calibration routine. Remember that atmospheric pressure will change calibration, So those of you with rebreathers than can go months without calibration, be careful, that's only valid is the air pressure doesn't change in that time. Personally I calibrate before each dive
Amplifier circuits
Modern Sensors try to draw as little current as possible from the cells to promote longevity and rapid response times, this has mean that we need powered amplifier circuits to oomph the current up to a more easily measured range and that is why many rebreathers need battery's (that and to power the solenoid and handset displays). Many units have un-powered secondary displays that run on the low current directly supplied by the cell
|
|
Age
As the lead and electrolyte are exhausted the cell becomes slower to react and eventually becomes current limited. It will calibrate in air and too 100% but eventually it's response ceiling (the PPO2 it can respond to) will drop. Then if the PPO2 is above this ceiling the cell will only show the ceiling as its incapable of producing the current needed to read higher. This will result in a cell giving and out of range warning eventually. The danger is that 2 cells will become current limited at the same time, and that their ceiling drops below 1.3 bar. This will result in 2 cells currently saying the O2 is below the set point, they will outvote the one good cell and the unit will inject too much oxygen. This will be spotted by the 3rd cell giving an out of range warning even though it is correct. This is why the calibration routine measures the voltage of each cell and rejects those it considers to low. The best defence for this is to change your cells regularly, make sure they are of different ages, check them against another source (ie VR3) and to measure their rest voltage often and reject cells that are low or sluggish when calibrating (you do watch your cells values and speeds when calibrating don't you?)
The worse scenario is that all 3 cells are current limited. You then have no way to spot this as all 3 cells will roughly agree during the dive. It is after some time in high O2 levels that the problem becomes pronounced and will result in a hyperoxic mix towards the end of the dive or while surfacing. Many Inspiration users have reported this to me. The last was no surprise as all 3 cells were over 2 years old. But all calibrated fine, were within tolerance on the voltage. Hence no cell warnings during the dive. Fortunately the person concerned recognised the oncoming visual disturbances and bailed in time. This was then repeated on the next 2 dives. ALL cells passed all normal pre dive and calibration checks
The ONLY way to test a cell for current limitation is to flush with 100% O2 at 6m and make sure the PPO2 raises appropriately. Better still just swap all cells at 12 month intervals
Catastrophic Failure
Of course there is also catastrophic failure. This is normally a function of broken internal connections in the cell. One of my cells failed after 2 hours of use. I put this down to APD shipping it to me in a box with 2 tubs of 20Kg Sofnolime, with no real padding separating them. There is also an internet rumour that the cells of batch F had a dry solder problems (I've had 2 batch F cells fail on me)
What I'm saying is that cells can and do fail, even if replaced at good intervals and new cells may not always work. Despite their expense it is worth keeping several spare cells with you, unless you want to marshal while everyone else dives one day.
Recommendation
Dave Thompson suggests you keep a log of the values cells start at and reach during each calibration. These are their raw values and tracking these over time will give you a good guide to the cells health, watch for slowing down or reactions
|
|
During the summer of 2002 several of us noticed that with time the O2 level measured by our 3rd party sensors (VR3's normally) began to drift upwards after about 20-30 mins use. The difference wasn't great but was about 0.02 to 0.05 at the end of a one hours dive. Both units would then drop to a return correct readings in air after the dive. The question was, was the issue the Inspiration or the VR3. I set out to do some measurements of the circuitry. The issue was also clouded by many questions as to the validity of the VR3's calibration to start with.
Some R22-2BUD specs for you. The cell is designed to have a resistance of 10 ohms loaded by the measuring apparatus. The Inspiration supplies 5 ohms. The cell is designed to provide a current of 1 μ amp when in use, the Inspiration draws 6 μ amp from it. I personally believe that the Inspiration over draws from the cells and the discrepancy I see is the Inspiration returning lower values with time than it should and hence slightly injecting more O2. The VR3 is reporting the true value.
Supporting evidence for this is the fact that the discrepancy is worse as the cells get older (new cells can tolerate the high draw better) and the fact that cells in the Inspiration degrade much much faster than cells in the VR3. e.g. 4 cells opened on the same day, all at over 13mv when new. After 20 hours use in 2 months, the 3 in the Inspiration were all below 11mv where as the cell in the VR3 was still at >13mv. Placing the VR3 cell into position 3 drew it to below 12mv in 2 hours of use!!!!
Monitoring of the Hammerhead replacement electronics show they draw 1/10th the current of the Inspiration head and the VR3 agrees exactly with the HH readings after the 2 hours I dry dived to test this. I expect the cells to last a lot longer in the Hammerhead.
This problem is very controversial and some folks don't see it, others do. Some say it is a VR3 calibration issue (I see it regardless of which way I calibrate). Martin Parker gave me a very long winded and calculated explanation about why the resistance is not a worry etc. Problem was he used AC circuit impedance maths in his sums, when the Inspiration is DC so I just hope the electrical guys on his team know a little bit better than that. He cant argue why the measured (Oscilloscope) current draw is nearly 6 times the specification.
This is not really a problem in use as the drift is minor, but could be an issue for folks doing very long or very deep dives or not replacing cells at yearly intervals
I have also noticed some other very spurious electrical readings while I was
probing and would like to ask
a) Why there is such a high current spike on the cells when the solenoid fires
(I would have hoped the Solenoid and Sensor circuits were separated)
b) Why is there so much stray noise on the measurement circuit in general (low
spec wiring and soldering?)
The Inspirations set point control was tested during its CE certification and was very good (checked by spectrometer), but that test was only for 20 mins and the discrepancy normally doesn't begin to show in that time. I am still happy to dive on mine. And believe this is just one more reason why cells should be changed at 12 monthly intervals
|
|
1) Regularly measure the output voltage of each cell. It should be between 13.5 and 7mv (2 outer pins) this is a crude representation of the amount of lead anode remaining as the potential difference is roughly proportional to the area of lead anode available to the Oxygen molecules. As the cell ages and is consumed this reading will decline. Cells below 7mv are at the end of their life. Obviously long term storage in High O2 environments will degrade cells much faster than storage in air. Cells should last 18 months in air, but as we also use them in high o2 environments a lot then 12 months is a realistic time expect them to last. Using them till they fail to calibrate is not wise (Calibration current is far lower than use current)
2) When calibrating cells, make sure the lungs are fairly empty and what gas is left is air (do the calibration after the negative pressure test), Watch the reading in the cells, they should start at around 0.7-1.2 and rise as the O2 is injected. You want to watch for cells that are wildly different from the others or respond at different speeds. Hopefully the calibration will take a fair time and you will see the cells travels through a large range of readings before the calibrating message appears. Dave Thompson suggests that you log the final values before "calibrating". This test checks the cells linearity over low set point ranges
3) Once you are deeper than 4m, use the O2 inject to quickly drive the cells up to 1.4. This verifies that they can respond to higher O2 levels. As cells age they can produce a lower and lower peak current. There comes a time that that peak current represents 1.3 bar PPO2, that means that the cells are unable to show a higher PPO2. Even worse is when they cant quite produce 1.3 bar PPO2, in this case if 2 cells reach this state the unit will NEVER think you are at the high setpoint and will continue to inject O2, even if the 3rd cell says 2.5 PPO2!!!!!!. This simple check makes sure that the cells will work fully at the high setpoint
4) Keep a log of when cells are purchased and opened. Write the opened date on the cells. Do bin the cells after the time interval you deem appropriate (I believe this is 12 months). Or use software like "Your Inspiration Buddy" to track you cells usage
5) Always empty the loop of elevated O2 levels before storage
|
|
r22d.pdf. Specifications of R22D cell
diving.pdf. Teledyne Diving Sensors Availability Sheet
lauer.pdf. General Information on the workings of Oxygen Fuel cells
sensormsds.pdf. Oxygen Sensor Material Safety Data Sheet (handy if travelling)
|
|