The Truth About Cell Calibration

This is written by Gerhard...for all those rebreather divers out there (and especially those who have been following the recent reported, alleged, near death on an inspo). If you have any questions....mail G at gerhard@tekdiver.co.za :)

First things first: cells becoming current limited is a huge reality, can happen for sure! Remember they are one of the weakest links in your re breather. So what to do....


Looking at how the unit calibrate:
When you hit the "yes" button, the display you are looking at is the milivolts divided and measured against air in Oxygen. So to give you an example , one for altitude , one for sea level.

Sea Level: assume ambient 1 bar. Air= 0.21 , O2= 100% Calibration starts and the displays read : 1.02 1.00 0.96 What the unit does now is compare the mV reading to air whilst being in O2 , thus ACTUAL mV will be the following:
1.02 will be basically 10.2mv (thus the inspo now uses a set MV for start calibration assuming about 4.7- 021/100% oxygen- as its linearity checking) so the real MV of the cell submerged in O2 will be 10.2 mV * 4.7 = 47.94 MV , and that would ASSUME the start up MV was 10.2 as well ( It assumes Cell linearity is constant) NOW if you kick the lid off and measure cell 1 with a mV meter and do NOT get a reading of 10.2mV in air.... oops then the assumed linearity is incorrect....more about that later!

The Cell should be kicked out if it falls below 8.3mv approx (39.01mV)

The Hammer boots out ANY cell that does not make 40mv actual as it is predicted to become current limited(including at altitude ,which is a
bugger)

At altitude the same applies EXCEPT that you must take the MB reading into
context:
Thus if your reading is 0.89 0.88. 0.91 at altitude with a 850mb reading then just multiply the resultant by 0.850 to take it to "altitude"
Thus cell on will be 8.9mV * 4.7 / 0.850 = 49.2mV SEA LEVEL , or 41.83 Altitude reading- good stuff.
So even if you do not see a 1.00 when calibrating at altitude it does not mean the end of the world. According to your altitude it should be fine (above 40mv) SO if you work out the actual MV and see it is below 40mV actual for the altitude regard it as suspicious

NOW what I do NOT like is the fact that it works on a constant linearity assumption when you calibrate( 100% / 21%=4.7). So if you look at an actual as in the hammerhead you have the following:

At sea level my cell mV read in air : 10 9 9 After dousing it in pure O2 for calibration they all stabilize at 39 48 43 mV respectively.
That means that the ACTUAL linearity is as follows :
Cell 1 - 3.9
Cell 2 - 4.8
Cell 1 - 4.3
Thus you can see that it is NOT an actual constant and also cell 1 that seems high did not actually make the calibration and is now booted out as a possible limited cell, but the inspo will calibrate this cell as acceptable (borderline but there) esp if you told it the O2 content was 98%......

At altitude I expose the head to air to see the PO2 shown to verify air:
The mV reads 10 9 9
The Po2 I expect to see is .21 * 0.850 mb = 0.178 PO2 actual that the handsets should display.
All cells displayed within that range since previous calibration.
I wrote down my mV in Oxygen from the previous dive which was 39 48 43
respectively.
Thus at altitude with pure O2 I expect to see a PO2 of 0.85 shown for each cell.
They all get very close to it.
You can then look at the linearity from each cell then vs now.

So cell 1 is 39/10= 3.9
Cell 2 is 48/9= 5.3
Cell 3 is 43 /9= 4.7
If you calibrate now and all values are still within that linearity level the cell is shown to be constant within calibration range. So if you DID fitness test the cell on the previous dive at 6m and it passed and now linearity check and mV seems stable the cell in theory should not be a problem as it's linearity range did not change (you can only hope!)

Also , in theory it also this tells us is that the cell showing highest linearity should be the most accurate in response range as it can read over a finer scale , no it does not mean it might not be limited, it just tells me the response range is the widest. It is like looking at a 300bar gauge vs a 400bar gauge. The 300 bar gauge will show clearer needle movement across the board as it uses a finer scale. You can see 2bar movement more clearly on the 300bar one than on a 400 bar gauge. So I would expect cell 2 to react nicely during flushes and O2 additions with cell one maybe a bit slower.

Use an approach where you can see yourself what the actual mV in air and actual PO2 is in air in regards to the previous calibration, and once the unit or cells are flooded with O2 then you can record that mV and PO2 shown BEFORE calibrating again. If the linearity is within the same limit the cell should perform in the same fashion ,esp if it is a few months old. You will at some point have 3 new cells in your unit ,it is unavoidable sometimes , but they need to be checked vigilantly esp in the first month. The hammer you can also leave flooded with O2 and come back after a while and look at the ACTUAL mV of the cells to see if there is any deterioration in the actual voltage. If there is , flood with O2 again, if the cells do not come back to previous max mV then get worried! The only way to test if it is current limited is to actually pressurize the unit on the surface or dive to 6m then spike it, deeper would be great. Problem is you do not want to start a heavy dive with an O2 spike through the lungs;-) But you can always go O/C , flush at 8 or so meters while manually manipulating the lungs to try get O2 cycled , look at the readings , then flush again and get on the loop and continue the dive.


Now this all sounds very good , but lets get back to the BUT... There are two options, one is single point calibration- so we use high o2 or 100% O2 to calibrate the cell, OR we try 2 point calibration, so we assume x linearity in air and x in O2 and draw a line, sounds good...and here is the BUT. Cells are NOT linear, they have a curve! The further away the maximum or minimum extends the worse the curve slopes, so your most accurate reading is the CLOSEST POINT TO CALIBRATION. In other words, if you calibrated at sea level- Po2 1 - then your most accurate readings should be in the 0.7-1.3 vicinity, if you calibrate at altitude- Mb 850- then all cells read Po2 0f 0.85.... hmmm...then reaching 1.3 is that little bit harder... Remember cells are usually within 2% error margin of calibrated value, so the only solution to check high end readings is to test the cell in a hyperbaric scenario. 2 Point calibration is not helpful either as I showed you with the hammerhead cells in air and O2 as the actual linearity differs, and as explained on a single point calibration you have a higher reference point from which to work, thus a smaller error margin...wow..!

Next option- Have cell verification like what was designed for the new Cis
Lunar- For those who do not know: It has only 2 cells ,one is the master and has a pure O2 and a diluent (air) injector close to the cell face. Approx every minute the handset hands control over to the slave cell and test fires a small amount of O2 at the master cell and a small amount of Air, it then looks at depth and if the cell does not show correct Mv for the depth the dive is aborted, so a constant hyperbaric test while diving, VERY cool...
except.... when testing was done on various other units ,they found that when blowing on the cell face you are faced with a few issues : If there is condensation on the cell face it sometimes helps to decrease it (yay!) , OR it actually blew moisture past the membrane into the cell causing.... yes you guessed it horribly incorrect readings. And guess which one manifested itself the most ? The latter of course.

The second fun problem (had one of those) is when you get a cell stuck. That basically gets a layer of moisture onto the cell face, it traps O2 behind the water ,and the cell only sees that O2 reading ,which is usually
setpoint. Now the computer sees 1.3 1.3 1.3 while you are happily
breathing the loop down and die embarrassed...or finally, when the O2 behind the cell face gets metabolized then it will fire the solenoid... oops, hopefully you are not an idiot and do realize when your solenoid did not work for a while, or your lung volume breathes down etc... The cool thing is if it was close enough to fire the solenoid before it got stuck it might be within range to start firing and of course will not stop firing the O2 in as the cell cannot see it... he he.

So what to do.....

IMHO ,only other possibilities :Nr 1= FLUSH YOUR UNIT OR GO O/C if you think there is a problem!!!!! The ONLY reliable piece of hardware is the thing between your ears we all assume is a brain ;-) Roll cells over (BEST), if you passed a fitness test on the cells within the previous 24hrs they should be good in theory, buy a cell checker from Narked at 90 (not really worth it-can built one way cheaper- 600 Pound Sterling!) or start diving at the setpoint where calibration passed (yeah sure!) Check your cells and record linearity if possible, but as stated that will just help you to identify if a cell seems to be performing in the same fashion as the previous dive, and where to expect a slow or fast reaction. The chances of loosing 3 cells to current limitation at the same time is VERY slim indeed! But be careful of exposing and leaving the cell in excessive heat or freezing it. Use a system that independently calibrates the cells and does not "take "readings from another. Use a 4th cell if possible.

In regards to cells, AP always claimed a better and more stringent protocol when verifying their cells, seems there might be a small hiccup. AI ( Analytical Industries) makes most of the cells. I have had as many AI cells fail as PSR cells as AP cells. Once again. Roll cells over, check linearity yourself for interest. You can clearly see how an assumed linearity can produce interesting readings. Just use bit of common sense and remember :
the job of the re breather is TO KILL YOU! Your job is to make sure IT DOES NOT!
If you weren't confused before , I hope you are now ;-) Cheers for now.