What's new
What's new

What's inside a gas mixer?

Tenson

Plastic
Joined
Jul 25, 2017
Hi,

I'm a total noob welder and like many I started to think how nice it would be to mix my own gas with a Y splitter to allow an 98\2 Argon CO2 mix from the separate bottles for doing stainless. Then I read a few forum posts and realised it was not so easy.

Now my question is: what's actually inside a shop bought gas mixer?

I'm just wondering if 3D printing might allow a DIY version some success where more simple DIY methods have failed given it can produce complex geometry and software like SimScale can even simulate fluid dynamics.

An open source 3D printable gas mixer would be pretty cool and I'm sure there are some big brains here that might have the combined knowledge to make it happen :scratchchin:
 
Hi,

I'm a total noob welder and like many I started to think how nice it would be to mix my own gas with a Y splitter to allow an 98\2 Argon CO2 mix from the separate bottles for doing stainless. Then I read a few forum posts and realised it was not so easy.

Now my question is: what's actually inside a shop bought gas mixer?

I'm just wondering if 3D printing might allow a DIY version some success where more simple DIY methods have failed given it can produce complex geometry and software like SimScale can even simulate fluid dynamics.

An open source 3D printable gas mixer would be pretty cool and I'm sure there are some big brains here that might have the combined knowledge to make it happen :scratchchin:

Ed Craig at weldreality detailed doing just that.

I used to use 2 regulators (with flowmeters) and a Tee fitting set-up
that way, but now I have a smith gas mixer.
 
Ed Craig at weldreality detailed doing just that.

I used to use 2 regulators (with flowmeters) and a Tee fitting set-up
that way, but now I have a smith gas mixer.

That sounds great! Any chance you can dig up a link for it? I sent a little while looking but couldn't find such a project, although lots of advise on gas mixtures.
 
That sounds great! Any chance you can dig up a link for it? I sent a little while looking but couldn't find such a project, although lots of advise on gas mixtures.

It's qty (2) regulators (with floating ball flow meters) and a Tee fitting.
No links, videos, nor drawings needed.
 
It's qty (2) regulators (with floating ball flow meters) and a Tee fitting.
No links, videos, nor drawings needed.

Reading forums people say in practice that doesn't work well. They say problems are 1) flow gauges are usually not accurate enough for 1-5% CO2 mix and mixing of gasses is not good enough in a T or Y fitting.

I was thinking:



Both regulators set at the same psi feeding a small accumulator. Those then get dosed to a lower pressure mix chamber via solenoid valves that open with a time value equal to the gas mix ratio desired. As the mix chamber is at a lower pressure than the accumulators there will always be flow at a similar rate when the valves open.

The mix chamber could be a simple geometric design like the two part adhesive mixers or it could include a brushless fan to help fully mix the gasses.

I did consider a feedback loop to control the dosage valve using a CO2 sensor, but most affordable ones only range to about 5000ppm / 0.5% CO2 and response time can be slow and can be affected by other gases (so I've read). Suitable sensors seem to be at least $100 to start (e.g. SprintIR) which kind of destroys the affordable DIY agenda!

EDIT: Oh you'd probably need some non-return valves.
 

Attachments

  • Gas Mixer.jpg
    Gas Mixer.jpg
    11.5 KB · Views: 26
It would be easier to use a 2p3w solenoid valve with each gas connected to it and the output feeding your mixing chamber / accumulator. Use a simple drive it with a pwm signal representing your mixture ratio. It would be annoying to listen to though.

I’d just use two flow meters personally, each of the required precision.
 
My Smith mixer is as follows-

From memory- there are a few regulators. One regulates the pressure of one of the incoming lines. Then there is another regulator that uses the pressure of the first gas as a reference to match the pressure of the second gas to the first. Then the gasses each pass through a separate valve. These two valves are geared together so one opens as the other closes. This seems like a nice design.


I found a post on the miller forum with pictures
https://forum.millerwelds.com/forum/welding-discussions/23602-low-cost-high-precision-gas-mixer-the-story-continues
 
How much do suitable flow meters cost that could dose say a 1-5% CO2 mix? (in my reading it would need to be a flow-meter variable orifice, not a flow-gauge variable pressure). Also it would still need a mixing chamber from what I've read of peoples experiments simply joining two outputs via the lines.

I think my design would come to about $50 in parts.

I realised a few issues with my originally posted design so have updated it in the earlier post.

I don't think you can rely on a PWM for the valves, because it's hard to be sure the applied voltage will be linear with orifice size and flow rate. You'd also still want an accumulator on each becasue the bottle pressure regulators may not respond fast enough to keep up with the dosing valves. Or maybe it would be fine, but 3D printed it costs basically nothing to do it.
 
Reading forums people say in practice that doesn't work well. They say problems are 1) flow gauges are usually not accurate enough for 1-5% CO2 mix and mixing of gasses is not good enough in a T or Y fitting.

I was thinking:



Both regulators set at the same psi feeding a small accumulator. Those then get dosed to a lower pressure mix chamber via solenoid valves that open with a time value equal to the gas mix ratio desired. As the mix chamber is at a lower pressure than the accumulators there will always be flow at a similar rate when the valves open.

The mix chamber could be a simple geometric design like the two part adhesive mixers or it could include a brushless fan to help fully mix the gasses.

I did consider a feedback loop to control the dosage valve using a CO2 sensor, but most affordable ones only range to about 5000ppm / 0.5% CO2 and response time can be slow and can be affected by other gases (so I've read). Suitable sensors seem to be at least $100 to start (e.g. SprintIR) which kind of destroys the affordable DIY agenda!

EDIT: Oh you'd probably need some non-return valves.

What forums ?
Scientific apparatus people trying to run mass spectrometers at home ?

give-me-a-break.

I did it, your just ready forums.
 
Gas mixing is the way toward blending gases for a particular reason where the synthesis of the subsequent combination is indicated and controlled. Volumetric gas division changes over inconsequentially to incomplete pressing factor proportion, observing Dalton's law of halfway pressing factors.
 
What forums ?
Scientific apparatus people trying to run mass spectrometers at home ?

give-me-a-break.

I did it, your just ready forums.

Which flow meters are you using? One person commented such a setup might be okay for 50/50 or 75/25 mixes but down at 98/2 much less likely to work.

This is the thread that made me dount I could do it affordably with good results. It's a good read.

Making your own mix of argon and CO2 -

Weld Talk Message Boards
 
I have a very old (1950's) if I were to guess "Victor" mixer. Inside it looks like one of those epoxy mixing nozzles, followed by a sintered bronze screen. I've used it to mix welding gasses. I have no doubt it does a good job of mixing gasses; but that is the trivial part of a proper mixer.

The real magic is getting the partial pressures of the gasses right at any flow or temp; thats the part you pay the big bucks for not the swirl tube.
 
They are pretty common on ebay. I know some people like to build stuff just because, but I got mine on ebay. Mine is an odd gas combo, not what I want but it was cheap. I don’t think it will be an issue to get it dialed to where I want.
 
I've got a few mass flow control units (MKS IIRC) if anyone wants to follow the DIY path

Would need to come up with a small adjustable power supply to adjust the flow rates. ;-)
 
I was at a sale put on by a guy who owned a welding supply store back in the '90s. Among the things I bought was a NOS gas mixer called a "MIG Master". It isn't adjustable like more complicated tools. You plumb 30 psi gas from each of argon, CO2 and helium cylinders to the mixer. The mixer has 3 inputs. Depending on how you connect the hoses, you get different mixes. With 30 psi pressures on all 3 gas lines you can get:

Ar69%-He30%-1%CO2 ;; pulse spray on stainless steels or nickel based alloys
98.5%Ar - 1.5%CO2 ;; spray mode for steels, similar to 98-2
70%Ar - 30%He ;; never used this one
88%Ar - 12%CO2 ;; general purpose MIG
92.5%He - 7.5%Ar - 0.5%C02 ;; regular MIG for stainless

I have found that if I want to vary the mix I can play with the pressures going into the mixer. I have never done that, though.

It appears to all be done with precisely machined orifices. This is not explained anywhere but is hinted at in the patent document:

Pulsed arc welding method, apparatus and shielding gas composition

metalmagpie
 
The 70% Argon 30% Helium mix is great for spool-gun/push-pull aluminum applications. Also great for emptying your wallet, these days, with the price of helium. I believe it was Praxair that used to offer a similar blend, they called it “Stargold” or something similar.

Years ago I worked in a trade show/exhibit booth fabrication shop and they ran that gas on their Miller 350P’s. That machine, with a bit of an assist from the gas, made mediocre welders look like geniuses. The deposition rates were incredible and the welds looked beautiful (with a modicum of effort invested).

It’s my understanding that helium burns a lot hotter, so using it in an Ar/He blend allows a bit better penetration.





Be safe



Jeremy
 








 
Back
Top