On Thu, 19 Aug 1999, Joe & Gladys Koontz wrote:

> Sounds like you were building a closed loop rebreather!  I agree, just
> build the rebreather and monitor the scrubber.

Gold Star for Joe!  Yes, I've done several PIC-based controllers
for closed-circuit rebreathers.  Of course, the absolute volume on
a rebreather isn't fixed, so you *do* need an O2 sensor.

Regards,

-Will


> good luck,
> joe
>
>
> "William M. Smithers" wrote:
> >
> > On Thu, 19 Aug 1999, Russell McMahon wrote:
> >
> > > I tried using a Zinc-Air cell as an O2 sensor REALLY roughly (seemed like
a
> > > cheap way to make a sensor) and cell voltage seemed to be very insensitive
> > > to Oxygen concentration (as you suggest it would be). Stored O2 appears to
> > > be a major problem.  I'll be trying it again in due course with  a little
> > > more finesses (hopefully). We bought a batch of "real" Oxygen sensors but
> > > even in 100 volume (I think it was) they cost around $US45 AFAIR. These DO
> > > produce a linear pO2 versus Voltage response. Being an electrochemical cel
l
> > > they have a finite lifetime and they drift severely so you need to
> > > calibrate them before every use. Fortunately there is an easy cheap and
> > > fairly accurate source of 20.8%-Oxygen calibration gas available. (Or, if
> > > you want to argue, 20.95 +/- 0.01% - yep, the ranges don't coincide-
> > > depends who you listen to :-)).
> >
> > Before I address the O2 sensor issue, let me first ask if you
> > really need one.  If you're looking to do life support
> > in a fixed volume container, all you have to do is blow
> > a muffin fan through a can of Calcium hydroxide (Sofnolime
> > brand is good), and have a demand regulator hooked up
> > to a source of 100% O2.  The Co2 goes into the scrubber,
> > which drops the pressure inside the environment, and
> > causes the demand reg to pump in the appropriate amount
> > of O2.  This is what they do in mini-subs.
> >
> > For the sensors - I did some extensive hyperbaric tests using
> > zinc-air cells as oxygen sensors.  My interests were a little
> > different, as I was mostly interested in measuring higher
> > partial pressures (0-2.0 ATA PPO2).  At any rate, the upshot
> > is that they have horrible drift and are not particularly linear.
> > They also have a very short life in hyperoxic environments, as
> > you might expect.
> > If you want to do some playing around with them, use a *small*
> > cell (Duracell DA13, as I recall), and put a few hundred ohm load on
> > them.  The smaller cells react faster and store less gas internally.
> >
> > Galvanics are the way to go for O2 sensing.  If you want small,
> > Teledyne makes a sensor called the K-1, which is about 3/4 inch
> > in diameter and 3/4 inch long.  The calibration issue is real - you
> > do need to recalibrate regularly if you want accuracy.
> >
> > There are also fourescent dye based sensors, which apparently
> > work extremely well, but I don't have any experience with those.
> >
> > Finally, there's paramagnetic O2 sensors.  This is really the
> > ultimate O2 sensor, as they never need to be recalibrated and
> > never wear out.  Unfortunately, the ones that have no moving
> > parts are the size of a Coke can and draw major current.  They're
> > also expensive ($2500 as I recall).  The current draw is because
> > they've got an internal heater, as apparently the paramagnetic
> > effect is very temperature sensitive.
> >
> > Along these lines, I screwed around with using Hall Effect
> > and GMR devices to brew up a cheap paramagnetic O2 sensor,
> > but never got too far with that project.
> >
> > -Will
> >
> > > decent
> > > >pO2 sensors.
> > > >
> > > >Just looking at their voltage wouldn't work, of course.
> > > >The voltage would be something like the log of pO2, a weak function, and
> > > >subject to stored O2 every time it got a good exposure.
> > > >
> > > >But the sensor was run as a current source into a virtual-ground meter
> > > input,
> > > >with a small diffusive air leak,
> > > >the current might be proportional to pO2 with a response time of under a
> > > >minute.
> > > >
> > > >This might be useful for monitoring inert-gas vs oxygen.
> > > >I never got around to trying it.
> > > >Has anybody tried this or heard of it?
> > > >
> > > >Jim S
> > >
>