O2 Cell linearity

Northern Monkey

I noticed while I owned the Boris a difference between the back-up handset readings and the main computer.

If the back-up handset was calibrated with oxygen the readings between computer and back-up would be differ on deeper diver, then match on the deco.
If the back-up handset was calibrated in air, the readings between the computer and back-up would start to differ on deco.

Slightly concerned over this I found if you calibrated the back-up handset in a pre known gas such as air or O2 it would not show linearity with the other gas O2 or air.

The unit contains 3 oxygen cells, all three are hardwired to the back-up display and are calibrated on this handset by 3 individual pots.

The main computer calibrates like a VR3.

Does your Boris do the same?

__________________
Death is not an option!

Genesis

How far off is it?

__________________
"A venturesome minority will always be eager to get off on their own, and no obstacles should be placed in their path; let them take risks for Godsake, let them get lost, sunburnt, stranded, drowned, eaten by bears, buried alive under avalanches - that is the right and privilege of any free American."
http://www.denninger.net
http://www.diversunion.org/liability.htm - Fix the Diving Cert racket

Northern Monkey

If you calibrated the back-up handset in air ie 0.21, when the unit was flushed with pure oxygen you would get a PPO2 readout of 1.05

The backup handset didn't seem to track the oxygen with any linearity.

I am curious now to know if most rebreathers with electronic oxygen control only have one reference point for calibration?

We all know that these rebreathers have 3 cells to allow the computer to make comparisons of cell outputs and make judgements. Some rebreather computers track these outputs every time the unit is recalibrated, to warn of forthcoming cell failure.

But do any rebreathers have two calibration references and need both air and oxygen to calibrate?

Any thoughts

__________________
Death is not an option!

Genesis

There are two thought processes on calibration:

1. Single point. The zero point (0% O2) produces a 0 reading and 0 voltage (theoretically), so you establish the slope of the response curve with one point (either air or pure oxygen)

2. Two-point. Usually done on both air and O2.

Both have advantages and disadvantages. Unless handled very carefully in the firmware two-point can be extremely dangerous, as it can set a slope that does not actually intercept zero! This is not a good thing at all, unless its compensating solely for an OFFSET (not non-linearity!) in the amplifier/ADC channels.

The problem is verifying that this is the case.

A two-point that is actually one calibration + a validation would arguably be better, but I know of no unit at present that takes that approach.

Errors are cumulative; the cell itself is typically specified as having a 1% error full-scale, the trim resistors used for any amplifier channel have errors (1% is the MOST that should be accepted; 0.25% are even better but quite a bit more expensive) and the ADC itself may have an error (or non-linearity) associated with it as well. If using a typical op-amp type circuit there are two gain resistors involved so the errors are additive there as well.

Finally, you have uncertainty error in the calibration, and if integer math is used there's uncertainty there as well (e.g. if, in air, the raw ADC value is, say, 20, then you have a potential error of 5%, since one count is a 5% deviation.)

Calibrating in oxygen results in a larger reading on the ADC and reduces that error, since you're reading a larger percentage of full-scale and a larger number of raw counts. It does not, however, affect any of the other error sources (e.g. amplifier trim ressitors, etc)

So a 5% deviation is not beyond the realm of reason when all of the uncertainties are added up - although its higher than I'd like to see. Many units calibrate in oxygen to reduce the count error but that comes with its own risks, most particularly if the O2 used is not pure or the flush is not complete.

__________________
"A venturesome minority will always be eager to get off on their own, and no obstacles should be placed in their path; let them take risks for Godsake, let them get lost, sunburnt, stranded, drowned, eaten by bears, buried alive under avalanches - that is the right and privilege of any free American."
http://www.denninger.net
http://www.diversunion.org/liability.htm - Fix the Diving Cert racket