Analox analyzer

[NAM001]

@tarponchik i keep bringing up the sensors used in rebreathers because they are literally identical to those used in analyzers. In fact, those that change their cells out early ~12 months, will usually rotate those old cells into analyzers. These cells have already experienced say 100-200 hours of use at a ppO2 of 1.2 ish, and then go into use as O2 analyzers.

Sensors die when they either become voltage limited, or behave in a non-linear fashion. You don't know that it is suffering from either of those conditions unless you test it both below, and above the circumstances of it's use. In this case, you test at 20%, and at 100% and use between 30-40%. You do'nt know by testing in air if the cell is limited because you have no way to validate the linearity of the cell.

Regarding O2. Allowing customers to calibrate the analyzer is frankly more dangerous than keeping an O2 tank for shop personnel to calibrate with since most people don't know how to properly calibrate, and if they aren't using a cal-gas for it, there is no guarantee that the sensor has been fully purged. O2 is a regular item in dive shops that are filling nitrox, so it's not like it's not already in the building

Let me add this. I use an el cheapo as long as you put the sensor in a zip bag when not in use they will last for years. my first went over 5 years. continous exposure to O2 burns then out.

 

[tbone1004]

@KWS correct, you are going to wear them out faster by not putting them in a sealed container *my Cootwo got an upgrade and moved into a Pelican case* than you do by calibrating with O2.

emphasis needs to be made though that calibration does not have to be done on a daily basis, ESPECIALLY when it is being used daily

 

[tarponchik]

I can double check, but the last recommendation I've seen was to calibrate analyzer "once per session". This makes sense cause if you check 5 tanks at once, why should you calibrate in between? However, If you check the tanks in the morning when it is chilly and at noon when the tanks warmed up, this makes some difference. Temperature drift will give you larger error than the above mentioned humidity. So the way they do it in, say, Sand $ in Bonaire also makes sense. You pick up a Nitrox tank or 2, and this is a session. You do it late in the evening the same day for your night dive, and this will be another session, etc.

Sensors die because every once in a while some O2 molecules bind irreversibly to the surface and it gradually gets oxidized. Ideally, one should calibrate with 2 standards every time because you need at least 2 reference points. You can establish "zero point" with one standard but you need 2 to check the linearity and the slope of your linear calibration graph, just like you said. This is the way pH meters work. They use the same principle, and users run calibrations daily with at least 2 standards. With O2 analyzers, though, we run calibration with a single standard. This can only mean one thing, that the aging of the sensor has been somehow accounted for or even pre-programmed into the analyzer, based on some tests the manufacturers ran. You tell them that you calibrate with O2 and they will freak out because they did not run tests for this. You can run double standard cals at your own risk, of course, but I will still use a neutral gas, N2, Ar or He as the 2nd ref. Just in case.

 

[tbone1004]

I can double check, but the last recommendation I've seen was to calibrate analyzer "once per session". This makes sense cause if you check 5 tanks at once, why should you calibrate in between? However, If you check the tanks in the morning when it is chilly and at noon when the tanks warmed up, this makes some difference. Temperature drift will give you larger error than the above mentioned humidity. So the way they do it in, say, Sand $ in Bonaire also makes sense. You pick up a Nitrox tank or 2, and this is a session. You do it late in the evening the same day for your night dive, and this will be another session, etc.

Sensors die because every once in a while some O2 molecules bind irreversibly to the surface and it gradually gets oxidized. Ideally, one should calibrate with 2 standards every time because you need at least 2 reference points. You can establish "zero point" with one standard but you need 2 to check the linearity and the slope of your linear calibration graph, just like you said. This is the way pH meters work. They use the same principle, and users run calibrations daily with at least 2 standards. With O2 analyzers, though, we run calibration with a single standard. This can only mean one thing, that the aging of the sensor has been somehow accounted for or even pre-programmed into the analyzer, based on some tests the manufacturers ran. You tell them that you calibrate with O2 and they will freak out because they did not run tests for this. You can run double standard cals at your own risk, of course, but I will still use a neutral gas, N2, Ar or He as the 2nd ref. Just in case.

you put far too much faith in the analyzer manufacturers. All it is, literally all it is is a voltmeter. It gets a voltage from the cell, and reports a number back. The calibration is to tell it what voltage equates to what %o2. i.e. 14mV=20.1 vs 13.7mV=20.1
That's it, nothing else. a neutral calibration gas does nothing for the sensors to tell you anything about linearity because 0=0 no matter what it is calibrated to. 0mV=0, always, you can calibrate it however you want, but 0mV will always equal 0 for the readout. You NEED a known positive value of O2 to calibrate these things.

 

[tarponchik]

you put far too much faith in the analyzer manufacturers. All it is, literally all it is is a voltmeter. The calibration is to tell it what voltage equates to what %o2. that's it, nothing else. a neutral calibration gas does nothing for the sensors to tell you anything about linearity because 0=0 and does nothing for calibration

You are dead wrong on this one because you get a zero O2 reading. "0" is a number too, like 1,2,3...

 

[tbone1004]

You are dead wrong on this one because you get a zero O2 reading. "0" is a number too, like 1,2,3...

do you actually understand how an analyzer works and what calibration does? First thing is first. 0 is not a number, it is the absence of a number.

how do you calibrate to 0? you can screw the calibration knob or whatever to any extreme and it will always read 0 because the voltage reading is 0.

say at atmospheric conditions the sensor outputs 13mV. You know that is supposed to be 20.9 so the calibration knob changes the multiplier. In this case it would be 1.6. The analyzer takes the mV reading, multiplies it by 1.6, and that is the output. The next week, the sensor is outputting 12.6mV at atmospheric so the multiple is adjusted to 1.66.
If you calibrate to 0 O2% in whatever gas you choose, the output is 0mV so no matter if the multiple is 1.6 or 1600, the reading is always 0. Ergo, calibrating to a 0% O2 gas, is 100% useless.

The reason for the 100% O2 calibration is to verify linearity. If the cell outputs 12.6mV at atmospheric, and we subject it to 100% O2, the multiple is 1.66, and we expect a reading of 60.2mV. If the cell is only putting out 59mV, then the cell is only 98% linear. The two point calibration gives you that value of 98% so we can adjust the 1.66 multiple to 1.63.

Is this significant in analyzing nitrox tanks. Quite possibly. If the is only 90% linear, then for EAN32, you would expect 19.3mV, only get 17.4mV, and at the multiplier of 1.66, your EAN32 is now reading as EAN29. There is 0 way to verify that by calibrating against atmospheric air, and it is impossible to verify that linearity by checking with a 0% O2 reference gas. Any reference gas used to calibrate an O2 cell, has to be a positive value of O2.

 

[RainPilot]

You are dead wrong on this one because you get a zero O2 reading. "0" is a number too, like 1,2,3...

To add to what tbone said, if the cell is totally kaput and giving off zero mV, then it will seem to be fine with a zero O2 gas. On my Analox it has one knob to calibrate. Thus doing multipoint calibration is pretty tricky.

I calibrate to atmosphere with a small correction if it is a very hot and humid day. If i have a 100% tank around, I might cross check against that to be sure, if it reads 100.3 % then I will dial it back to 100.

 

[victorzamora]

One reason to do a two-point calibration with a high FO2 gas is to make sure the cell isn't limited. I have an O2 cell at the house that will read air just fine, but will only read 27% when exposed to EAN32. There's nothing I can do (well, without a soldering iron ) to make the analyzer read EAN32.

One note: Cells read PO2 not FO2. Since these are all being used at the surface, it's approximately PO2=FO2....but that's not really the case at altitude and it can be fairly different even at sea level (Depending on barometric pressure). Not that that's a factor, but I've seen mention of cells and analyzers reading FO2.....which isn't strictly accurate.

 

[tarponchik]

1. It is one thing to use O2 as the 2nd ref when you want to check if your sensor is still any good, and it is completely different thing to use it daily in calibrations. But you do not really "calibrate" your O2 analyzer with a 2nd ref and you can't because you only adjust your 20.9% point, not the calibration slope.

2. 0 (zero) is a number with a physical meaning here. You do not necessary have 0 mV output at 0% O2 from your sensor; it can be a small positive value. But you adjust your scale to zero because you know you have 0% in your N2 or He tank.

3. Funny thing is, the same guy who does not know WHY analyzers are calibrated in the 1st place, asks me if I know how calibrating a device works

 

[victorzamora]

1. It is one thing to use O2 as the 2nd ref when you want to check if your sensor is still any good, and it is completely different thing to use it daily in calibrations. But you do not really "calibrate" your O2 analyzer with a 2nd ref and you can't because you only adjust your 20.9% point, not the calibration slope.

This isn't necessarily true. With the Analox analyzer it is, but at least with the DiveNav COOTwo there's a two-point calibration method using both atmospheric air and pure oxygen. This takes both points into account to calibrate the linearity and slope.

2. 0 (zero) is a number with a physical meaning here. You do not necessary have 0 mV output at 0% O2 from your sensor; it can be a small positive value. But you adjust your scale to zero because you know you have 0% in your N2 or He tank.

The way that these sensors work is there will be zero voltage produced by the galvanic cell in the presence of zero oxygen. Let's pretend for a moment that that's not the case. You said above O2 is useless for calibration and is only good for "checking" as you can only adjust your 20.9% point (which isn't true, even with the Analox analyzer). If that is true, then using pure N2 would be the same. Without a more in-depth calibration ability than the analox analyzer gives you, there's no benefit to checking at anywhere other than a known-point and verifying at oxygen.