Rudimentry sensor mV monitoring

Gilles

Reading through the following thread

http://www.rebreatherworld.com/megal...edures-13.html

makes me realize how potentially dangerous our reliance on sensor readings is. This danger IS NOT meg specific, but relevant to all systems that are expected to respond to O2 sensor readings. I'd like to create a summary of the value created by the above-detailed thread. Please add to this as you see fit.

1) Calibration should be made at a PO2 greater than the expected range of operation. A 1 or 2 point calibration (@ 0.209 & 0.98) doesn't cut the cake when you need to have a reliable 1.1-1.4 readings. The extrapolation that we do when diving to these apparent PO2's is dangerous.

2) Descending to 6m with an O2 flushed loop (and checking readings and behaviours on way down) goes a long way to address 1). My new procedure!

3) Those of us in remote areas won't have the means to build calibrations kits with gauges and analyzers. There's possibly a market for a better sensor calibration and performance checking kit. Following is hopefully a means to help identify a delinquent sensor whilst we're stuck with 2 point calibrations.

I attach an .xls for logging sensor readings (in air and O2) and comparing with perfect linearity. If my understanding is correct, a higher mV than theoretical (positive drift) corresponds to a higher risk of DCS (versus Oxtox). Once the drift reduces and becomes negative, you start having a positive sign of current limiting. One can maybe anticipate the onset of sensor failure by logging these readings over time and checking the trend of the drift error.

For a given O2 purity, as the drift reduces, your getting less mV for a given O2 and you're getting an initial indication of current limitation.

Hope this makes some sense.
G

Attached Files

sensor calibration.zip (4.6 KB, 169 views)

__________________
Gilles
http://www.dirrebreather.com

Gilles

After another 15 hours or so of Meg, with the new worry about whether my sensors are telling the truth or not, I made a startling discovery.

I am the only CCR diver at my locale, so whilst a , I'm forced to dive profiles of single tank OC air divers. This implies doing 2-3 45-70 minute dives anywhere between 12m and 40m in a day. In the past in between dives, I noticed how my sensors were reading .18-.19 in air, and that I could never get close to 1.6 when flushing to O2 at 6m at the end of the dive. I changed all sensors, and all did the same thing again. So up to a few days ago I was always worried about my PO2's being higher than indicated.

When I made this post I started logging the sensor mV's. By charting the data (slide 1 in attached), I could illustrate to myself how linear each of the cells were and monitor any onset of non-linearity. This in combo with going OC and reverse flushing with O2 at 6m, confirmed the cells were fine at the start. At end of dive however, 1.3-1.4 was the best I could achieve. Something definitely wrong! Slide 2 in the attached illustrates a comparison of the individual sensor mV's pre and post dive. When wet (or post-dive), the response remains linear (to PO2 of 1 anyway), but there is a dangerous bias toward under-estimating PO2! (slide 3).


Since then, I've been calibrating end of dive day, rather than before the beginning! At start of day, Air is .24-.25 and O2 is ~1.1, but I do get a very comforting 1.6 at end of dive (@6m O2 reverse flush). So at some point, (my) Meg sensors appear to get affected during the 1st dive. Everything remains great on subsequent plunges.

I dive in the South China Sea in temps of 27-30 degC. The scrubber gets up to 34 degC at times. There's never been excess water in the can (absorbant cloths slightly damp), but there are condensation droplets on the sensors and carriage.

G

PS: I hugged Meg today

Attached Files

Sensor Cal.zip (13.6 KB, 77 views)

__________________
Gilles
http://www.dirrebreather.com

schford

If you are calibrating accuratley then your sensors will be reading the correct PO2 I just got back from Innerspace where I played a bunch with calibration then checking the results in the pot. I am really confident that you need to way way screw up calibration to risk getting dangerous variances.

If you feel confident that you calibrated correctly then it has to be an external factor - I note that you say that you had moisture on the face of your cells - I bet thats what is causing it to underead....

I have never had moisture on my cells other than when I dive warm(HOT) water for long 3 hour plus dives...

Then I did notice that when I did some o2 flushes at 6m the cells were only reading 1.4

By calibrating at the end of the day - I guess that you are calibrating with moisture on the cell faces...

It would be good to know if other people diving in that temp of water for your profiles get droplets on the cell or if it is just you - I wonder if it is worth a chat to Leon about?

Stuart

__________________
Bailout and Kit Management account for Murphy's Law

The only bad question is one you did not ask and later wish you had.

Use of Rebreather World is subject to the Rebreather World Terms & Conditions of Use

Jordan

When I did dived with Caver 95 at Peacock (shallow dive in 72 F water) he had to allow his sensors to dry between dives on his meg. I took a look and they were quite moist. We did our 2nd dive at Cow and afterwards, he was amazed to see my optima sensors bone dry. I sure am glad I can leave my unit together for a few days in between dives and not have to let the sensors dry out.

__________________
Is it clear? No. Well, let's go anyways.

"Those who would sacrifice a little freedom for a little order, will lose both, and deserve neither." Thomas Jefferson

Sutty

I get condensation droplets on my sensor faces on my classic Kiss, it is a bit of a problem on a several day trip. On the recent rebreatherworld trip I was drying them off at least after every two dives, and resorted to the gentle application of the edge of a clean piece of tissue paper to mop up the bulk of it before air-drying. Cell 2 seems the worst affected.
Once the droplets cover the bulk of a sensor face it under-reads significantly.
I wish the ADV blasted dil at the faces to blow away the droplets - then a dil flush would fix the problem rather than just confirming an under-reading cell.
Anyone got any other solutions?

__________________
Never forget that life is a finite resource.

Gilles

One thing worth noting however, is that my wet cells are maintaining their linearity. Hence, there's no value added by leaving them to dry. I completed 9 short dives today on a post dive calibration performed 4 days ago. Again all is fine at surface and at 6m. The only discrepancy is before the 1st dive of the day when the cells have been exposed to air during the night. (air = .24 O2 = 1.1 & 1.8 @ 6m). I now have an indication that it takes about a half hour for the cells to reach their wet state.

Aknowledged is the risk of DCS, but someone on this forum mentions that one can get unbent, not un-dead.

__________________
Gilles
http://www.dirrebreather.com

Jordan

Gotcha. I will tell my buddy of this and maybe he won't have to break open his loop between dives anymore.

__________________
Is it clear? No. Well, let's go anyways.

"Those who would sacrifice a little freedom for a little order, will lose both, and deserve neither." Thomas Jefferson

MikeH

That's just because your a very dry(pun intended) person....

Mike

Gilles

I think I will follow your suggestion and discuss with Leon & Steve (perhaps Cedric as well).

I am calibrating with moisture on the cell faces. It so far seems that the effect (of moisture) remains constant. As stated in an above post, I performed 9 short dives on a 4 day old post-dive calibration. Yesterday morning I did a dil flush @30m (with EAN30) and got 1.2's all across (30 minutes into the dive). I now want to determine how long it takes to get from dry to wet state (in the above manner I should observe the dry sensors starting ~1.5 and leveling out @1.2 with dil).

The cell linearity remains very good (within 0.05 ATA) in wet state up to 1.6. In this case, I don't think wet cells are a problem providing they are calibrated in the state they're expected to operate in.

I also just finished putting a drop of water on dry cells. No effect (behaving as dry cells). I suspect maybe the hydrophobic membrane is effective at repelling liquid water, but water vapour gets through whilst diving and subsequently condenses on the sensor elements (potentially a function of the temperature I'm diving in).

Has anyone else gathered any pre/post dive mV data?

G

__________________
Gilles
http://www.dirrebreather.com

UWSojourner

Hey Gilles,

Can you give us an idea of your calibration routine before you went with this "wet sensor calibration." Are your sensors stored in hot humid conditions? Or are you calibrating in an air conditioned building where humidity is kept much lower? I assume you are on Apecs? How do you perform the steps for calibration of that system?

Anything you can pass on with regard to temperature and humidity of calibration location and location where your meg is stored might be helpful.

Thanks.