Nitrox stick O2 sensor placement

[Diver-Drex]

I have an MCH 16, which I’m estimating an average cfm of 10cfm over the operating pressure during filling. This would give me 39lpm of O2 flow for EAN 32, and 67lpm for EAN 40.

I’m going to use a typical O2 welding reg and a 10-100lpm flow meter. 2” PVC for the mixing chamber. My compressor intake is 1”.

My question is placement of the O2 sensor. If I tap into the 2” the sensor would be seeing 5cfm flow. Is this going to cause an issue with readings because of high flow rate? If I use a 2x1 inch tee to hold the sensor out of the main flow, would this ‘correct’ the high flow rate impact or possibly not give enough flow? Or tee off a 1” section just prior to the intake hose?

Thanks,
DD

 

[rjack321]

The flow rate issue comes into play when the gas is flowing "against" the sensor as it creates a relatively pressurized membrane. If there is a T in intake line with the cell off to the side it's not an issue. Also you calibrate with the analyzer with the compressor running and no O2 flowing. So any minor pressure differentials skewing the sensor up or down are accounted for under air (20.9%) conditions anyway.

 

[tbone1004]

doesn't particularly matter so long as it is after the mixing chambers. The closer to the intake head the better. It is very important to note that depending on what you use for the mixing baffles that you are not creating too much negative pressure in the intake. This matters for two reasons, first is it can cause the pump to work harder and less efficiently which is not good for it, but also if you calibrate the meter with the compressor off you'll end up with a skewed reading on the sensor since partial pressure is directly related to ambient pressure. As @rjack321 said if you calibrate with the compressor running and O2 off then you'll be fine but this can throw the final mix off depending on how large the filtration system is so you do want to be very quick with the calibration.

I personally would use a reducing T to go from the 2" to the 1" and put the sensor there

 

[Diver-Drex]

Thanks. I highly respect input from the both of you.

 

[tmassey]

It is very important to note that depending on what you use for the mixing baffles that you are not creating too much negative pressure in the intake.

Any suggestions on what we should be targeting for back pressure?

I just rebuilt my trimix stick. For my 6 CFM compressor I’m using approximately 6 feet of 2 inch PVC with the following layout: 02 injection, then approximately 24 inches of bio balls, then O2 sensor, then helium injection, then approximately 24 inches of bio balls, then O2 sensor. Approximately 4 inches between each injection spot, bio balls, sensor, etc. which uses up the extra 2 feet of length.

Measuring at the final O2 sensor spot, a magnahelic measures approximately 4 1/2 inches (water) of vacuum. Unfortunately that doesn’t include about 8 feet of 1 inch vinyl tubing that goes between my trimix stick and the compressor itself.

I’m not using any kind of tee for the O2 sensors: I simply drilled a hole in the PVC and stuck the O2 sampling nipple into the flow. I have paid attention to the difference in calibration from still air and with the compressor running. Probably varies about 0.3% between the compressor running and not running.

I’ve looked up compressor specifications for supply piping length and number of elbows and I’m well, well under that length. But I’m assuming they’re not packing their tubing full of bio balls, though. So I’m just trying to get a general consensus on what is a reasonable amount of pressure drop on the intake.

From a blending standpoint, my setup is 100% fine: I start my compressor, let everything stabilize for maybe 15 seconds, then hit the calibrate button on my O2 sensors. They calibrate, and stay nice and stable. But from a loading my compressor standpoint I have no idea. That’s actually the part I’m most worried about. That’s primarily what’s motivating my question, but I’d take any information anyone could supply.

 

[TravisD]

Dangit, tmassey, now you've got me browsing eBay for a cheap Magnahelic so I cat test my stick.

I thought that "any" vacuum on the intake was considered bad?

A 2" stick for that size compressor seems small. For the OP, I'm running a similar size compressor as you (11cfm charging rate Bauer). My stick is made out of 4 feet of 4" PVC for the mixing chamber (Bioballs), that reduces down to 2" on the output which is where I drilled and tapped (M16x1.0) for the O2 sensor to screw into. I have about 5' of 1.5" tubing that connects to the compressor inlet. The sensor seems to do fine here - as others mentioned, I start the compressor and calibrate the sensor with the pump running (sending the output to my air bank), then add in the O2 slowly. Once it's stable and runs for a minute or two to make it's way to the output I switch over to the Nitrox bank.

I don't have a flow meter, just a welding reg feeding a 2nd low pressure regulator with a needle valve on it's outlet to give it something to work against. Adjustments are done based on what the sensor sees, so I haven't felt a need to add a flow meter yet. Most of the adjustment happens at the 2nd regulator - touching the needle valve makes bigger changes than I usually want. Still a work in progress (what fill station isn't?) since I"ve only had to bank Nitrox a few times in the couple of years I've owned the fill station.

 

[tmassey]

Dangit, tmassey, now you've got me browsing eBay for a cheap Magnahelic so I cat test my stick.

Well, then, my work here is done.

I thought that "any" vacuum on the intake was considered bad?

I think us “shade tree“ compressor owners tend to be a lot pickier than commercial applications. I think we tend to obsess a lot more and overdo things.

For example, here’s a link to a document that outlines pipe lengths for compressors of different size. Notice that the length allowed for a 6 CFM compressor using 2 inch PVC is over 700 feet…

https://www.powerexinc.com/wp-content/uploads/2019/10/Compressor-System-Intake-Pipe-Size.pdf

Yes, things like elbows do reduce that, but still: 700 feet.…. Like I said: my worry is 4 feet of bio balls, not the length of 2 inch piping I have.

As a quick and dirty stand-in for a vacuum measurement: how does your O2 sensor reading change on air with the compressor running and the compressor not running?

 

[TravisD]

As a quick and dirty stand-in for a vacuum measurement: how does your O2 sensor reading change on air with the compressor running and the compressor not running?

Sensor installed in pipe, compressor not running (and hasn't run for a few weeks), calibrated to 20.9%

Start compressor, start building pressure, putting air in the bank - 20.8%

Stop compressor, wait a minute or so, analyzer reads 20.9% again.

Is there a formula to derive pressure drop from these readings?

As an aside - when I'm not making nitrox I cover the hole in the pipe where the O2 sensor goes with a piece of masking, er, tank marking tape (blue or yellow depending how I feel. Tape doesn't seem to suffer at all from this so I'm assuming that ΔP is relatively low

Oh, and don't forget that in addition to the mixing media, most of us run an intake air filter on the end of the stick to keep crap out of there. That's going to add a little bit of resistance too.

 

[rjack321]

I thought that "any" vacuum on the intake was considered bad?

You have to have *some* pressure gradient between the intake and the actual compressor inlet to have gas actually flow. I have a 2" stick and 1.25" hose with a 3cfm compressor. For a 10cfm unit I would probably be up around 4" for the stick itself with baffles and the hose would be the 2" part.

 

[Diver-Drex]

The flow meter contains a needle valve, so that is covered. I just like the idea of having the visual.

Looking at what is available commercially, Envirodive, 1” works for up to 13cfm. Based on that, 2” is overkill for my compressor. I have plenty of room for bigger pipe so that isn’t an issue. Is everyone upsizing because you’re adding more mixing media?