Analox analyzer

[tarponchik]

I certainly having nothing more to add to the many fine explanations tbone has wasted his time typing....

In Hollywood, they call characters like you "a helpful sidekick" Jar-Jar Binks was a good example

 

[stuartv]

Get lost, finally. You have nothing to say.

This, from the person who brought us these lovely gems:

if you are so concerned about safety, diving 100 ft on Nitrox is borderline.

And maybe you still believe in Santa? Oxygen analyzers used by divers in dive shops to check tanks are so inaccurate you can use them only to tell if you got air or some kind of Nitrox, and then trust in the guy who mixed it.

I wonder how you manage this? If you have a solid grasp of what you're talking about, you do know, of course, that there is no oxygen tank where we divers pick up Nitrox.

I love to use Nitrox on 60-90 ft dives but I do not go deeper. On air I went down as deep as 180 ft

Being an anti-oxygen freak, I do not breath Nitrox unless I absolutely have to. Hey, if only you guys have seen what oxygen radicals do to our DNA like I've seen it in the lab, you'll all become anti-oxygenists!

 

[NAM001]

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

If i understand his response..... you cant say a 0 reading with no O2 is good if a 5% gas also results in a 0 reading. a 2% reading 2% O2 means more than a 0 reading 0. BECAUSE 0 IS 0

 

[NAM001]

for this discussion we have many opinions that i think are based on different situations.

low O2 contents --------------- cal ing with air is all that is needed ------------ use a tank of air to use as a dry source
high O2 contents ------------- another game all to gether. -------------- I would not use an air calibration to measure deco O2 perhaps 50% gas is fine but not pure or 80% and above. You have to use a source that you can exceed the gas being measured to trust the reading.


FOR LOWER % GASSES THE USE OF 30 or 35% test gas IS CENTRAL Range TO ALL GASSES UP TO 50%. 90% of all no air divers fit that group of 40% nitrox and less. And is just off enough from standard 32% to know that your reading is not still the test gas.

 

[victorzamora]

1. I am talking only about analyzers I know and have used.
2. You misunderstood, in fact, I said that 2 refs calibration would better than one from the start.

It's my understanding that the Analox analyzer can only be calibrated at one point, but verified at others.

Explain the process as you envision it: How would you calibrate at two points on an analyzer that only has one adjustment knob?

 

[dberry]

To quote, "Electro-galvanic fuel cells have a limited lifetime which is reduced by exposure to high concentrations of oxygen. The reaction between oxygen and lead at the anode consumes lead, which eventually results in the cell failing to sense high concentrations of oxygen." But if you are concerned about the alkaline electrolyte, why CO2 isn't the main problem?

To address your 2nd point first, because the diffusion rate of CO2 is much less than that of O2; that's why a semi-permeable membrane is used. Teledyne sells a wide variety of sensors and CO2 sensitivity is one of the parameters. As far as the lead anode being consumed with oxygen, that's a valid point. The wikipedia article - although described as pertaining to fuel cells in general - seems highly focused on O2 sensors for SCUBA and rebreathers, and the cell shown is not the same as one the in my analyzer (which is exactly the same as what we use in lab.) I'm not certain our sensors use lead anodes, but it's going to come down to what quantity of anode material is present. Very little O2 actually permeates the membrane to generate the current. In normal use to check gases in a tank I'd bet (but can't prove) that cell evaporation is more important than consuming the anode. OTOH, constant use to monitor O2 in a rebreather might be a different story, especially if the gas stream is humid (which would minimize evaporation.)

Sorry, I've let the details of the chemistry get us further off-topic.

 

[tbone1004]

They do not agree with everything you said. For example, they agree with me on this one: "However there are also safety issues to be considered as pure oxygen can increase the risk of fire in many situations."

Also, they agree with both of us that sensor output becomes non-linear at high O2 and this results in error (the yellow line on their graph proves this). They suggest to use a 2nd calibration gas like 50% O2 to get around this (which you did not). Then they immediately switch into selling mode: "However, you’d now be carrying around two calibration gases and associated accessories, when all you really need to do is to change your oxygen sensor more frequently." Bingo! Buy more and support the economy!

So their answer why using pure O2 is better is "because you are more likely to notice that your sensor is too old." I won't argue against this one, maybe you will. However, because of the non-linearity near 100% issue, your measurements will only gradually become less accurate and you will deal with sort of creeping error.

BTW if you look at their graph, zero is the most reliable point. All 3 lines meet there, no deviations or non-linearity is seen.

obviously oxygen is a fire hazard, that's a no brainer, but it's already there, and it's the most accurate calibration gas out there. The risk is essentially negligible for what we are doing with it.

they aren't carrying around cal gases, they're actually saying that that is an issue and the best gas to calibrate with is 100% O2.

what you are failing to quote from them to support your agenda is that they say you should calibrate with a gas that is of a higher O2 content than what you are analyzing.
What your 0% analysis won't tell you is the linearity of the cell because 0 is not something you can calibrate to with these types of sensors. You can't "0" them out, because again, it's like your car engine. When it's not running, it doesn't tell you anything.

The real risk? If the cell is only 90% linear, and is giving a "normal" reading of say 12mV at atmosphere, i.e. 12mV=.209pO2=20.9% O2, then at 90% linear, when trying to read EAN32, where it wants to see 18.37mV, it's only going to see 16.53 and report a pO2 back to you of 28.7% O2.

How does your 2 point O2 calibration to 0% O2 tell you anything about how to draw that line and prevent EAN32 from being reported as 28.7 and seriously disturbing someones NDL calculations? Thankfully they all fail by failing to give enough mV so you aren't going to tox someone so long as they don't readjust it, but if you're PP filling and using an analyzer with your method of calibration, you could well tox someone. By forcing it to get up to what it wants to read at 32%, you're actually going to be filling 36%, and if they're diving to the MOD's, it could tox someone.
Much worse is since you have to PP blend 50%, if you calibrate your way, and a cell is only 90% linear, it wants to see 28.7v. You PP fill and top off again and end up with a 56% mix and run the risk of seriously toxing someone at depth when they're on deco by running their PO2's way up.

How does your 0% O2 address any of the linearity issues that O2 cells are prone to? Is the risk of calibrating against 100% O2, which is already in all of these shops that are filling nitrox anyway, worth the risk of not being able to prove cell linearity which based on how the analyzers work as shown in the math above which can result in some serious deviation from the actual mix you are analyzing?

I have shown the math of how the cells actually work, and how the analyzers are calibrated proving that analysis from a mix of higher ppO2 is critical for accurate calibration of an analyzer. Please demonstrate, using math and examples of how a 0% analysis is beneficial, how you can actually calibrate to that, and why it is better. If you can do that? I'll yield and start calibrating my own analyzers with a 0% mix. What I have known from experience and validated by reading the Analox manual is that the only way to give an accurate analysis of your gas mix as well as proper function of your sensor is to give a dual point calibration with at least one of the two points at a higher ppO2 than the mix you intend on analyzing. If it is not, you do not know if it is limited, and you do not have a way of accurately determining the linearity. Please enlighten me and Analox on how we are both wrong

 

[GlennL]

Dang, I didn't mean to start such a heated debate. I have learned some good info though and think I have a good direction for "calibrating" / adjusting the set point for the analox. Thanks y'all

 

[rhwestfall]

Dang, I didn't mean to start such a heated debate

you realize this is SCUBA Board, right?

 

[tarponchik]

@tbone1004:
obviously oxygen is a fire hazard, that's a no brainer, but it's already there, and it's the most accurate calibration gas out there. The risk is essentially negligible for what we are doing with it.

Fine with me if the analyzers are used only by those who work in dive shops. But you know that everyone and their grandma is using them.

OK, I finally figured this one out. This is a Y=a*X+b equation. So when you "calibrate" with, say, 50% O2 and 100% O2, you actually calibrate using only 50% O2 (or whatever your 1st reference was) and then either the slope ((a) in the equation) must be preset in the analyzer and you adjust only the (b) parameter, or the algorithm in the sensor forces the line through the 0,0 point every time you calibrate. There is no way around these 2 options.
You use the 2nd reference only to check if the X=100% point will get onto the straight line but it has no effect on the calibration, cause it is already there. Thus, this is not a "dual point" calibration because you used 2 references; it is a dual-point calibration because the 2nd point is a mathematical 0,0 point. Whether there is a physical zero signal at X=0 we do not know since we see only the adjusted readings. The neutral gas (X=0) will work just the same if only (b) is adjusted because in this case the aging sensor will show you result not equal to zero at X=0 (in fact, it can be negative). However, the 1st possibility is unlikely because the slope (a) depends on the temperature and needs to be set through the calibration process. Thus, you were right and I was wrong, though not because the X=0 point is useless but because it is already in use by default. I suggested to use it because, based on the description given in Wikipedia, I assumed that the actual signal at X=0 may be not equal to 0. However, this cell actually works in quite different way, so your analogy with the engine not running is applicable here. My apologies and thank you for you patience.

BTW, the way the sensor works is that the current in the cell is proportional to O2 PP, not the voltage. In effect, by adding O2 you decrease the internal resistance of the galvanic cell while the voltage in the circuit remains constant. For linearity, the resistance of the external load must be low. The aging of the sensor comes not from O2 absorbing irreversibly but simply by lead depletion from the anode.

I still am sure that high O2 will kill your sensor sooner.