calibrating on multiple days diving

[rddvet]

the risk of a bad O2 flush and miscalibrating. Since very few divers, fewer instructors, and even fewer instructor trainers acknowledge that linear deviation is a thing, they usually don't calculate predicted mV output for calibration. .

That's pretty scary that it's still a thing. I will say that I have miscalibrated once, but was able to quicly catch it.

 

[tbone1004]

That's pretty scary that it's still a thing. I will say that I have miscalibrated once, but was able to quicly catch it.

it is SCARY easy to do on the old Megs that had 2 point calibration, much harder with a single point, but the linear deviation thing is even scarier, but you'd be surprised how few divers/instructors/instructor trainers truly understand it, but even scarier how many of them don't believe in it.

 

[rsingler]

And just to argue the point, that's the nice thing about the JJ-CCR and recalibration with every scrubber change (along with comparison of the recorded hi-O2 mV with both previous calibrations and predicted from air mV - but that's not in the factory checklist).
1) It's 1-point calibration
2) It generates numbers on the Shearwater based on the previous calibration which you can watch as it plateaus. If it plateaus at 99-100% and you hit confirm, you have not only calibrated but confirmed that it hasn't changed. If the mV are as expected you have also helped confirm 1.0 atm linearity.
3) As an eCCR, the flush is controlled by the electronics and the solenoid. A slow (incomplete) flush due to partial solenoid blockage or malfunction will also be revealed each time you change scrubber. A small added benefit.

YMMV, but I felt it was worth pointing out, even if nothing ever changes before it's cell replacement time.

 

[Asheron]

Okay, SCR diver here and have a question about calibrating CCR.

On my SCR I calibrate with air and validate/span to the gas in my bottles. Bottles were checked for mix with another sensor during pickup from shop. So it takes 2 different inatruments(shop and scr computer) to agree on what is in my units loop during checks. If the shop read doesn't match computer read I investigate.

Why wouldn't a CCR be calibrated in air and then the loop flushed with O2 to verify span? We know exactly what air is, where we may not know exactly what is in the loop due to a bad flush, therefore skewing our calibration.

 

[tbone1004]

Okay, SCR diver here and have a question about calibrating CCR.

On my SCR I calibrate with air and validate/span to the gas in my bottles. Bottles were checked for mix with another sensor during pickup from shop. So it takes 2 different inatruments(shop and scr computer) to agree on what is in my units loop during checks. If the shop read doesn't match computer read I investigate.

Why wouldn't a CCR be calibrated in air and then the loop flushed with O2 to verify span? We know exactly what air is, where we may not know exactly what is in the loop due to a bad flush, therefore skewing our calibration.

ccr operates around 1.0 during normal use so there is no benefit to calibrating in air due to linear deviation unless the unit has multi-point calibration capabilities

 

[Asheron]

ccr operates around 1.0 during normal use so there is no benefit to calibrating in air due to linear deviation unless the unit has multi-point calibration capabilities

Could you please clarify linear deviation? Are you talking about:
1)points not on the expected "Slope" line with each end at the expected point

OR

2)one end of the slope starts on an expected point and it trends off the entire run?

 

[Asheron]

Also,

I dont think my petrel will let me calibrate under 90 something % O2 when in CCR mode....so yeah I'd be stuck using O2 anyways.

 

[Jack Hammer]

the risk of a bad O2 flush and miscalibrating. Since very few divers, fewer instructors, and even fewer instructor trainers acknowledge that linear deviation is a thing, they usually don't calculate predicted mV output for calibration. This means you have no actual way of knowing if the system was purged properly and you are getting an accurate calibration on MOST machines *things like the APD, Poseidon, and Divesoft are less prone to a botched calibration*. So you only get a 90% purge on your unit, and because you're an ignorant f*ckhead and don't believe in linear deviation your cells are only 90% linear, so now you have a compounding issue where by the time you hit your 70ft stop and think you're breathing at 1.6, you're actually breathing at 1.6/.9/.9=2.0 and you do the funky chicken and oxtox. I don't care what the manufacturers say, they're trying to CYOA, it isn't because it's best practice, it's just easiest practice to not have to fully teach about how cells works and just tell the diver to calibrate on setup.

@rjack321 in general industry, devices are only calibrated when necessary and against known standards, typically annually. Their calibration is typically validated on a much more frequent period of either hourly, shift, daily, etc. but the calibration is verified, the unit is typically not recalibrated. This at least applies in the plant that I run which has over 1000 devices that are required to be calibrated for our ISO9001, AS9100C, and several customer specific requirements. Most devices are calibrated annually, some biannually, and they are verified at some more frequent interval. With the O2 cells I take this same approach. I calibrate in a known controlled environment, typically directly out of the O2 T-bottle, and then I validate calibration when I assemble. If the validation fails, then I will figure out why and determine proper course of action.

That helps and makes sense, Thanks.

This is the top of the assembly checklist sticker on my unit right now. You can see MV in air, expected, & actual. I'll recheck before diving tomorrow (about 60 seconds to do) and verify it still matches what's I expect to see. Sticker stays on unit until broke down and new one filled out and applied upon reassembly.

 

[nadwidny]

@rjack321 in general industry, devices are only calibrated when necessary and against known standards, typically annually. Their calibration is typically validated on a much more frequent period of either hourly, shift, daily, etc. but the calibration is verified, the unit is typically not recalibrated. This at least applies in the plant that I run which has over 1000 devices that are required to be calibrated for our ISO9001, AS9100C, and several customer specific requirements. Most devices are calibrated annually, some biannually, and they are verified at some more frequent interval.

Your experience may be typical for your work place however don't try and extend it outside of that environment. I worked in a lab that had to meet ISO17025 (if we are going to name drop standards). Some of our instruments had to be calibrated annually, some monthly, some had to be calibrated daily. Calibrated, not validated.

 

[rjack321]

Your experience may be typical for your work place however don't try and extend it outside of that environment. I worked in a lab that had to meet ISO17025 (if we are going to name drop standards). Some of our instruments had to be calibrated annually, some monthly, some had to be calibrated daily. Calibrated, not validated.

Most of our field instruments are required to have daily calibration, pH, DO, turbidity, free chlorine and such. These sensors are electrochemical sensors that are surprisingly similar to O2 sensors in a lot of ways and they all experience more drift than you'd expect.

Lab instruments like a GC-MS don't really have a "calibration" like were are talking about here. The curve is established for every batch.