Analyze nitrox after compressor

[Björn Jakobsson]

Hi,
I have read many threads and pages about making your own Nitrox stick. All of them puts the analyzer before compressor intake. If I want to use my current analyzer (Dive Soft Solo) i will need a small sampling pump to force some mixed gas into the analyzer.

If, however, I put it after the compressor, it will automatically do its magic due do higher pressure. I have a flow limiter and was considering adding a 5/7” connection to plug it in between compressor and tanks.

Is this poor practice to analyze after compressor? If yes, why?
As long as I can guarantee I am not mixing more than 40% of course since that is max for my compressor

 

[rolanddiver]

"As long as I can guarantee I am not mixing more than 40% of course since that is max for my compressor"

Exactly. If you analyze downstream of the compressor, how to do you know what is going into the intake of it? Should your regulator on the O2 supply bottle fail, you will only know that you exceeded 40% when your compressor flies thru the garage ceiling. I would just buy some cheap ($3 IIRC) small 3volt sampling pumps to draw gas from the intake when filling. Also it is crucial to have the shortest tube between the nitrox stick sampling port and the analyzer itself to allow for faster detection of O2 changes. Older compressors (if not all) tend to pump slower when reaching higher pressures, but your O2 flow remains the same, meaning that you are pumping nitrox with a higher O2 percentage than desired. When filling Nitrox 36 it's only 4% away from the mfg recommended maximum for most compressors.
What I did is that I installed a 3.5mm jack socket on my Dive soft analyzer. It is the type of socket that stereo equipment uses when you plug in the headphone into the amp it turns off the speakers. When I plug in the external O2 sensor that is connected to the nitrox stick, it switches from the internal sensor to the external one. If your analyzer is still under warranty I don't recommend this.

Just my two cents worth.

Roland

 

[tbone1004]

I wouldn't for the last sentence in your post. O2 analyzers are cheap, you can build one for like $20 plus the cost of the sensor itself and just screw it straight into the intake.

 

[Couv]

Analyzing after compression in no way helps you prevent too high a concentration going into the compressor. As @tbone1004 says, put the sensor after the mixing chamber but before the intake. You have to analyze each tank after filling anyway.

Note the clear line allows metered O2 to enter just under the bug/dust filter. Just before the compressor intake (near the bottom) is the sensor.

 

[rjack321]

Analyzing afterwards and you are measuring what went in 5+ minutes ago. Good luck adjusting the O2 flow to make your desired mix. Every little change will take so long to show up in the post compression analyzer that your tank will be full of random nitrox before you get it all dialed into 32%

Nevermind the oops boom factor

 

[tmassey]

Do what I do: meter before *and* after!

You *have* to meter before for all of the reasons already stated. Metering O2 after really doesn’t help — it’s too late to change it. But doing so, I can echo @rjack321: it takes minutes for changes at the source to catch up at the output, because of all of the gas in your filter stack, moisture separators, etc. (Think mathematical asymptote.)

So why do I have metering on the output? Continuous CO monitoring. *That* can only be done at the output. So I threw an O2 sensor there as well. But like I said, it’s pretty useless.

ETA: don’t waste your very nice tank analyzer as a Nitrox stick monitor. Buy an O2 monitor better suited for the stick as tbone mentioned. I like the Miniox 3000: you can get them on eBay pretty easily. Sure, you’ll need another O2 sensor at $80, but it’ll last multiple years, and if you can’t afford an extra O2 sensor, SCUBA compressor ownership is likely not for you...

 

[Celt]

What’s coming out has to be the same as what’s going in, if it’s 32% O2 at the tank then it’s 32% at the intake, I use a needle valve at the intake and monitor the high pressure side works perfectly and before anyone starts jumping up and down saying your compressor is going to blow up, no it’s not.

 

[Björn Jakobsson]

Thank you all for the replies. First of all, it was never about the cost, but convience. The boom factor is not my biggest consern, but the “5 minutes to see changes” really made the difference here. You are absolutely right and that is why I am going to place the sensor on the intake

please post more here, I’m always eager to learn!

 

[Björn Jakobsson]

Do what I do: meter before *and* after!

You *have* to meter before for all of the reasons already stated. Metering O2 after really doesn’t help — it’s too late to change it. But doing so, I can echo @rjack321: it takes minutes for changes at the source to catch up at the output, because of all of the gas in your filter stack, moisture separators, etc. (Think mathematical asymptote.)

So why do I have metering on the output? Continuous CO monitoring. *That* can only be done at the output. So I threw an O2 sensor there as well. But like I said, it’s pretty useless.

ETA: don’t waste your very nice tank analyzer as a Nitrox stick monitor. Buy an O2 monitor better suited for the stick as tbone mentioned. I like the Miniox 3000: you can get them on eBay pretty easily. Sure, you’ll need another O2 sensor at $80, but it’ll last multiple years, and if you can’t afford an extra O2 sensor, SCUBA compressor ownership is likely not for you...

Is it a standalone sensor that can be screwed right into the mixer meaning no pump needed?

 

[DiveClimbRide]

What’s coming out has to be the same as what’s going in, if it’s 32% O2 at the tank then it’s 32% at the intake, I use a needle valve at the intake and monitor the high pressure side works perfectly and before anyone starts jumping up and down saying your compressor is going to blow up, no it’s not.

How long does it take to get the mix right initially? Do you find there's much of a delay? Why measure at the hp side? Convenience? It seems that an intake sensor would be easier and have less latency.