DIY gas sensors?

[CuNanoTube]

Here is a flow diagram I was thinking up.
The Purge system for the CO chamber would probably be the hardest part to plumb, so if the chamber could open the power of one's breath might suffice. I included it because it seems the sensor you found requires an almost stationary air flow which means it might take a long time for all the CO to be pushed out after calibration, especially if the chamber is big. It seems you could greatly reduce the size of the CO analyzer you chose by using a different housing.

 

[gsk3]

Thanks Henrik.

And nice diagram Hashime!

I'm not dead-set on that panel meter for sure; it was just the first I saw.

I like the idea of a PVC enclosure for the whole CO meter assembly, but I agree it's going to waste a lot of gas. It would be better to either break it apart to expose the sensor (and just enclose that with an o-ring), or to build one from scratch so the sensor could be better isolated.

I agree about the check valve. LP check valves shouldn't be too expensive.

The dwell chamber I was thinking of would T off the main line from the reg and would have a purge or ball valve at the end of it. So you could purge at the reasonably high flow the reg was providing to equilibrate the chamber, then stop the purge and let the gas sit around the sensor for 15+ secs while the reading stabilizes. The lower the volume of the chamber, the less reference and tank gas you'd use, so that becomes the overriding concern.

The only standalone CO sensors I could find where at Digikey, and they start at 20ppm. Anyone have a source for better ones?

 

[CuNanoTube]

Here is an interesting article I found on CO sensors.
I also found some electrochemical type sensors from a company called Figaro Sensors. They do not post prices unfortunately, which means they will probably want a bulk order. I have found in the past saying you want a sensor for evaluation tend to work in these situations.

 

[Lobzilla]

Hi Ari,

I would test for CO first as hashime outlined and then O2 and He in series.
The test is over anyway if you have CO.
After finding sensitive and affordable sensors, the rest is easy, especially if you do calibration and user interface via a microprocessor (I can help there).
I will open my Nuvair Trimix analyzer tonight and try to find out where the He sensor came from.
That leaves us with the question what CO sensor would be a good candidate for a prototype.

OTOH, Amprobe CO meters cost around $130 and a Chinese knock-off can be had on fleabay for $100. Is it even worth duplicating this DIY style?
O2 and He analyzers are a different story as their manufacturers milk you pretty badly due to the small market volume.

 

[Lobzilla]

Another idea.

The inexpensive CO sensor are not sensitive enough (>20ppm) for our needs but that is at 1ATA. I wonder how they react if we perform the test at 10ATA (the typical IP of a first stage). All you need to do is to close the outlet of the test chamber in hashime's diagram to find out.

 

[CuNanoTube]

Another idea.

The inexpensive CO sensor are not sensitive enough (>20ppm) for our needs but that is at 1ATA. I wonder how they react if we perform the test at 10ATA (the typical IP of a first stage). All you need to do is to close the outlet of the test chamber in hashime's diagram to find out.

Well, CPVC can be good up to 500PSI, or one could just get SS piping to make the chamber. The only thing we will need to figure out is how the sensor senses. There could be more inference from other gasses at higher pressures. Good idea though. If reg LP is enough to measure we could do away with the flow measuring system (flow meters are only good for low pressures anyways) and hook directly to the tank. Then we would need a pressure gauge and a conversion table. Hmm... to google!

EDIT: Well, I googled electrochemical sensors, and they work by reducing or oxidizing the gas and measuring voltage from that. That means pressure will change the reaction equilibrium, but since I don't know precisely what other factors are involved in the system, and I don't know the exact reaction, increasing pressure may not change the equilibrium in a way we would expect. I don't know enough about electrochemistry to figure that out yet (I believe it is a 2nd year course in my program).
Here are some articles, they have useful stoicheometric equations, but no rate equations. Without knowing the reaction kinetics (something pressure affects) it is hard to figure out what pressure will do.
http://www.intlsensor.com/pdf/electrochemical.pdf
http://www.delphian.com/electrochemical%2sensors.htm

 

[Lobzilla]

Well, CPVC can be good up to 500PSI, or one could just get SS piping to make the chamber. The only thing we will need to figure out is how the sensor senses. There could be more inference from other gasses at higher pressures. Good idea though. If reg LP is enough to measure we could do away with the flow measuring system (flow meters are only good for low pressures anyways) and hook directly to the tank. Then we would need a pressure gauge and a conversion table. Hmm... to google!

EDIT: Well, I googled electrochemical sensors, and they work by reducing or oxidizing the gas and measuring voltage from that. That means pressure will change the reaction equilibrium, but since I don't know precisely what other factors are involved in the system, and I don't know the exact reaction, increasing pressure may not change the equilibrium in a way we would expect. I don't know enough about electrochemistry to figure that out yet (I believe it is a 2nd year course in my program).
Here are some articles, they have useful stoicheometric equations, but no rate equations. Without knowing the reaction kinetics (something pressure affects) it is hard to figure out what pressure will do.
http://www.intlsensor.com/pdf/electrochemical.pdf
http://www.delphian.com/electrochemical%2sensors.htm

The >20ppm sensors are so cheap that just trying it may be the best approach. Obtaining the calibration gas is a different story.

 

[Lobzilla]

Bad news on the He detector. I opened the Nuvair Trimix analyzer and the He sensor is not a discrete off the shelf component. The whole test chamber is custom made with the detector buried inside and the (off the shelf) O2 cell is screwed in from the side. The circuit board is made by Teledyne Analytical Instruments and it looks like the He test chamber comes from the same source.

 

[lucca brassi]

some idea ?

Ultrasonic gas measuring device - Google Patents

 

[gsk3]

The teledyne and most of the others I believe use thermal conductivity compared to the reference gas. There's another model whose brand I forget which uses ultrasonics to measure the speed of sound in the gas. More likely to find an ultrasonic sensor as a complete unit, but probably harder to actually incorporate. Thermal stuff is easier to understand.

As for CO, the TDI unit would be great for building into a box, but the key will be getting it plumbed up so you don't have to bag each sample...