[PSUBS-MAILIST] O2 Sensing

Mon Aug 19 23:57:56 EDT 2019

A minimum indicating accuracy of +/- 0.015 bar corresponds to 0.014804 standard atmospheres. It is odd that they specify this in terms of pressure instead of percentage for analysis, but I guess you could have a cabin pressure which is not standard. Anyway, you are correct in that is is ~1.5% of a standard atmosphere, but that seems coarse for an oxygen analysis. If I'm more than 1% off of target when mixing gases for diving, I dump the gas and start over. Of course, the consequences are less severe in a 1 atm space. My analyzer reads to 0.1% ( ~0.001 bar), but sufficient accuracy shouldn't be difficult to hit with any electronic system. Even an 8-bit ADC will give you 0.4% resolution. 12 bit gets you 0.025%.

Where does your 18% figure for the low oxygen limit come from? I wouldn't be comfortable with allowing the oxygen to drop that low - particularly with sensor error that could be as much as 2% FS. I'd probably set my low alarm at 19%, giving normoxic +/- 2% as the alarm points, and might set caution alerts at 20 and 22.

As far as calibrating the readout, at minimum you require a two-point calibration in order to set zero (offset) and span (gain). I use pure helium, but argon, nitrogen etc. also work. With a 0% oxygen gas flow, adjust the zero so that the readout is zero. Then flow air (20.95%) and adjust the span accordingly. You then need to double check the zero point again to ensure that the offset hasn't moved with the new gain. If it has, just repeat the procedure. More complicated calibrations with greater numbers of points are possible (linear regression, linearization table, etc.), but likely not worth the effort in this application.

Sesn

-------- Original Message --------
On Aug. 19, 2019, 20:50, Alan via Personal_Submersibles wrote:

> Hank,
> I agree with that; if you get to 25% & can't get the percentage down,
> then you would abort your dive. The only problem I have with the 25% range
> of the optical sensor is that DNV-GL require 100% range & I have been trying
> to build to their specs. They are also saying " the system for the analyses of
> oxygen shall have a minimum indicating accuracy of +/- 0.015 bar". I make that
> 1.5% ( could someone confirm). Most galvanic O2 sensors are saying +/- 2%.
> for accuracy. I see the sensor Cliff referred to is +/- 1%. ABS are just asking for a
> system that keeps O2 between 18 & 23%.
> With regard to checking the O2 meter & calibrating it, some of us are just buying
> the sensors & writing code for them in to an on board computer, like Cliff's system.
> So it is impracticable to take it outside of the hull to check. I would think that any
> Imbalance between the outside & inside of the hull would equalise pretty quickly
> with the hatch open.
> For our purposes where we should not be going outside the 18-23 % O2 range,
> & are staying close to 1atm; surely a 1 point calibration or check against air at sea
> level would be good enough?
> Alan
>
> On 20/08/2019, at 9:22 AM, hank pronk via Personal_Submersibles <personal_submersibles at psubs.org> wrote:
>
>> I think you guys are over thinking this O2 reading beyond 25 percent.   If you are diving and you reach an alarm of 25 percent, you go to the surface. There is no reason to go beyond 25 percent reading because you will abort your dive. Calibrating is an other issue if your sub is not well vented prior to diving. I take the O2 sensor outside the sub to calibrate.   A large blower is my solution also.
>> Hank
>>
>> Sent from my iPhone
>>
>> On Aug 18, 2019, at 2:04 PM, Brian Cox via Personal_Submersibles <personal_submersibles at psubs.org> wrote:
>>
>>> Alan,          The shape of the hull and overall length is for good sea keeping ability because I plan on traveling some distance on the surface.  Unlike the common Kittridge designs which do not handle surface transversing very well.  Also the other drawback with the Kittridge designs is their lack of freeboard, which is ok if you're in a lake but in the ocean, not so good, especially if you have to open the hatch.
>>>
>>> Brian
>>>
>>> --- personal_submersibles at psubs.org wrote:
>>>
>>> From: Alan via Personal_Submersibles <personal_submersibles at psubs.org>
>>> To: Personal Submersibles General Discussion <personal_submersibles at psubs.org>
>>> Subject: Re: [PSUBS-MAILIST] sub test
>>> Date: Mon, 19 Aug 2019 07:36:14 +1200
>>>
>>> Hi Brian,
>>> congratulations on getting under the water.
>>> I thought you might have been up on the conning tower rocking her around
>>> to check the stability.
>>> DNV-GL have rules on stability & procedures for testing. They might be a good
>>> Guide line to see if the stability is acceptable.
>>> Maybe you could get rid of 10ft of the front of your sub. I can't see it's purpose.
>>> You could always put a small deck on the sphere starting at the base of the conning
>>> tower, & have either syntactic foam, floats or a ballast tank under it. Then add lead as needed to the inside of your hull.
>>> That huge conning tower is going to destabilise you on the surface until you descend
>>> to the point that it's weight is negated by it's positive buoyancy.
>>> Alan
>>>
>>> On 19/08/2019, at 6:12 AM, Brian Cox via Personal_Submersibles <personal_submersibles at psubs.org> wrote:
>>>
>>>> Hi All,
>>>>                   Well I removed a massive section of concrete from the front top area of my sub to lighten the overall weight and especially the weight up front.   I wasn't really sure how this would effect things so I didn't try to anticipate anything, so I just put her back in the water to see what would happen.  I did move some weight forward because I though maybe that would be necessary.   So it looks like I have an imbalance and I appear to be a little top heavy when in the surfaced state, and the nose will need some added buoyancy.  It looks like I'm going to have to add some flotation pontoons to the sides back towards the sphere so I don't tip over.  Also I have some material that I can remove up top that is not necessary  , I built on a support structure under my deck that is way over built , so I can shed a little weight there.
>>>>
>>>> Anyway here is a video of my latest dunk test:
>>>>
>>>> https://www.youtube.com/watch?v=-S3CZU66mGs
>>>>
>>>> Brian
>>>
>>>> _______________________________________________
>>>> Personal_Submersibles mailing list
>>>> Personal_Submersibles at psubs.org
>>>> http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
>>>
>>> _______________________________________________ Personal_Submersibles mailing list [Personal_Submersibles at psubs.org](/eonapps/ft/wm/page/compose?send_to=Personal_Submersibles%40psubs.org) http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
>>
>>> _______________________________________________
>>> Personal_Submersibles mailing list
>>> Personal_Submersibles at psubs.org
>>> http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
>
>> _______________________________________________
>> Personal_Submersibles mailing list
>> Personal_Submersibles at psubs.org
>> http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.whoweb.com/pipermail/personal_submersibles/attachments/20190820/c5613d5d/attachment.html>