I'm working on a nitrogen generator which will take compressed air (filtered/dried) and pass it through a nitrogen separation membrane. This is pretty simple to do, but I would like to be able to display some of the metrics responsible for determining the purity of the nitrogen on a display screen.
The membrane filter I'm using is a MHN-1512A, and the concentration of the N2 in the output is a function of three things:
The temperature of the input gas
The pressure of the input gas
The flow rate of the input gas
Now the pressure and temperature are relatively easy to control and sense with sensors I already have. But the flow rate is a bit different. I know there are plenty of air flow rate sensors out there, but a lot of them don't fit this scenario. The flow rate sensor needs to meet the following criteria:
Sense inline air pressure - as opposed to something like a wind speed sensor (the plumbing is mostly 1/4 NPT)
Handle at least up to 13 bar (188 PSI), though 20 bar (290 psi) would be nice, as that's what the membrane is rated up to, but I don't think I'll be using anything that high.
Sense flow rates up to 20-25 NLPM , though 40 NLPM would be ideal.
Be at least decently accurate. Ideally have a resolution of around a tenth of a NLPM.
Not cost an arm and a leg.
Just so you have an idea of the parameters I'm looking at, here's a chart that came with the N2 sep membrane (The chart was a bit confusing initially, so I added some color along along with a legend at the top, hope it helps):
I initially thought that I could maybe use one of those inline water flow sensors with a brass body (example), I realize that these are made for water and not gas, so I figured I could kinda celebrate it by using an adjustable flow control valve (example) to correlate the pulse rate from the water flow sensor to what the air flow rates I see in the flow control valve, but after testing it out, it looks like its far too unreliable for whatever reason.
I had thought about getting a mass flow sensor or mass flow control valve, but those seem a bit overkill (and expensive) for this project.
So, does anyone have a good solution that would help with this? Any input is appreciated.
For this applications the best choices is the coriolis mass flow meter and the vortex flow meter.I dont know your variable NLPM, but maybe in mass is better for you.
The bad point here is the cost, these sensors are expensive and need to be specified by the flow and diameter pipe, because a bad reynold number calculation makes have a bad flow metering. The other way is use a multivariable or Differential pressure flow, using a wedge tube or venturi tube.
The most cheap sensor is the turbine sensor, but need to select right, and you need more info about the flow.
For give you more help, please put more info, process dti, the flow range in kg/l or volume, etc. etc.
Right, I was looking at stuff along those lines (eg: SFM3119), but as you pointed out, they're pretty expensive. If I needed something with such a wide range of flow rate and very high accuracy, I wouldn't mind spending that much on it. But I feel like that's trying to kill a fly with a sledgehammer (overkill).
Ok, I thought I did provide some of this in the chart and bullet points above.
Up to 20-25 NLPM flow rate
At least up to 13 bar (188 PSI)
What do you mean DTI? (I work for a bank, so I think Debt To Income ratio, but im assuming that's not what you mean, lol)
You do not seem to be aware of how to determine the purity of your nitrogen. Let me help you. That is determined be measuring the amount of oxygen in your gas. The nitrogen is the percent left after all oxygen is removed.. Other gases are negligible.
I'm trying to keep everything to 1/4", but not all components will be 1/4" NPT. And even if they were, since I'm going to be passing the gas through a series of filters, I can't really just use the pressure and pipe ID to calculate the flow rate.
No, I do know how to measure it. Some of the components I already have:
Way to measure the O2 purity (OOM202)
Pressure sensor
Humidity and temperature sensor (I found a few SHT-85's on Alibaba, and
But this is more of a fun project/challenge than anything else. If I can measure the flow rate going through then it would make it a bit more useful than just being able to read the O2 concentration.
I was looking at the SFM4300-20-P on Digikey, but according to the datasheet, its only meant to really work up to 7bar, with an overpressure of 10bar:
And I would like to be able to have a pressure of at least 13 barg at a flow rate between 14 to 22 NLPM, which would yield between 3 NLPM of 99.5% O2 or 12 NLPM of 95% O2 (as detailed in the spreadsheet image above).
One tried and true method of measuring flow rate is measuring the differential pressure across an orifice. You will need a DP sensor that is rated for your maximum system pressure and an orifice sized to your flow rate.
The DP is a square root function of flow rate. Since you are measuring a gas flow you will have to compensate for pressure and temperature also. Fortunately you have a computer to do all of that for you.
Thanks for the compliment This is a project im getting a little carried away with. Ultimately, for what I plan on using it for (pics here and here), knowing the exact flow rate isn't 100% necessary, but for things I would like to use it for in the future, it may be necessary. Plus, I think it would be nice to have a display with some of the live stats to give me a feel for whats going on "under the hood" (eg: if the pressure increases but the flow rate drops, then maybe the flow is slightly obstructed, etc).
Indeed, that would be nice, especially since it sounds like something that could provide the type of accuracy I'm looking for (at least with the type of sensors I would use). But I think I would need a lathe or mill to be able to make something like that :-\ Unless you know of some existing products that I could use for it? Im thinking something made out of metal should be a necessity given how important the orifice needs to be a known size (thus it changing due to wear isn't ideal) and the fact that it's also going to need to deal with some elevated pressures.
A typical industrial orifice plate produces a pressure drop of 100 inches H2O. (Sorry, I didn’t translate to SI units) . At the pressure you are working with you could use a plate with a hole drilled in the center and fit it into a pipe union or some other similar arrangement. You are not tied to that DP either, I have worked on systems that operate on DP of less than one inch.
At that pressure drop with clean, dry air there should be very little wear. The hard part is finding a reasonably priced sensor that will operate at the pressure you are running.
Im seeing that there are some UNIT brand Mass Flow Controllers (so not just sensors..)
that are configured for N2 (or "air") that are in the flow range I'm looking for, and some even with 1/4" threaded attachments - For about $60 to $70. They're mostly old and obsolete, and use those antiquated card connector things, but thats not a big deal. Here's an example one: Unit UFC-1110A Mass Flow Controller MFC, H2, 20 SLM, 1/4" Male | eBay
Anyone know how to use one of these things with an Arduino?
From my quick search, the device is powered with a dual 15 volt supply. One supply is +15 volts the other is -15 volts measured to a common pin.
The output is 0-5 volts DC measured to another common pin.
So, the only Arduino connection is the 0-5 volts to an AD pin and it's common to the Arduino ground.
By Differential pressure can measure the flow, but the line diameter is too small. In this case if can you get one use a ultrasonic flow meter for gas metering, is the best choice, can you use a chinese flow meter or a keyence FD-G series meter.
I did look into those. Most of them don't handle pressure very well. The closest one I found was able to handle barely 5 bar, or 7 bar overpressure. The ones that do are far outside of my price range.
Yeah, I found some stuff online, one thing that may be very helpful is this: DIY Mass Flow Controller Interface – Tom – Teenager, Ireland
Granted, that's for a different version of MFC (It uses a better connection than some of the ones in my price range). It also references Bens video (Applied Science) where he uses a few of them, and that's how I found out about these as well.
Another helpful paper here uses a totally different (more expensive) MFC, but it looks fairly similar. And this one includes an Arduino sketch Low_cost_MFC_control_unit_using_microcontoller (2).pdf (2.9 MB)
He still does, lol. I signed up for his Patreon, I message him every so often and he replies. Very smart guy. He's actually what inspired me to start this whole project.
Im totally getting side tracked here, but did you see his DIY Liquid nitrogen setup? He pulled the sterling cryocooler out of an old RF filter that he got for like... less than $200 I think? And found other parts for very cheap. So I figured I could do it if I was patient enough to wait for the parts to. popup online somewhere.
I waited over 3 years, looking everywhere you would think of and probably places you wouldn't (all online), and found squat. I even asked the guy who Ben purchased his from. Turns out, after Ben posted his video one or two guys went around and purchased every unit they could find. They just took the cryocoolers out of them and sell them for like 10x what they bought them for >_<
But I talked to Ben about it, and he said that even though it's very cheap (or was for him, at least, lol), it only generates like 1L of LN a day (or less), and given how quickly it boils off if you're using it for something (eg: vapour trap, cold bath, etc), I figured that I should try to get something better.
I pasted these links above, but I didn't mention what it was I was linking to.
Liquid nitrogen generator I purchased off of ebay:
Pretty excited about it. I know it works, I just need to get some Helium to recharge it and finish this N2 generator system, then ill have unlimited LN2
P.S: Just so I keep this somewhat on track... I purchased a UNIT MFC for N2 Should be perfect. It was only $50.
Just had an idea, what about an automotive mass flow sensor? Those are used to determine how much fuel to inject, and they seem pretty accurate and cheap.
