Flowmeter vs regulator? Interesting results.

[Mach_Zero]

Ok so I've been chasing a bad argon coverage issue on my tig welds. I had posted this thread when I first thought it was a contaminated tungsten electrode: Bad tungsten causing sparks and porosity on mild steel?

After a lot of back and forth, I got some great results from new consumables (both new stubby collet body and new stubby gas lens body worked perfectly).

However, in my search I found some info regarding how regulators tend to be less accurate than flow meters. I bought a cheap flow meter on Amazon (probably a big mistake here, but I'm cheap). It arrived today and I set out to do some testing with the primitive equipment I have on hand. I figured I'd post it here in case it's amusing to anyone or just to get your opinions on it.


I wanted to test both devices to see how accurate they are. I set both of them to 25 scfh and measured the volume of gas they dispensed over a controlled period of time. I measured the gas with some super high tech equipment: balloons around the cup of the tig torch . I pressed the pedal for 15 seconds and collected the gas in the balloons. Then I measured the diameter of the balloons and calculated the volumes.

Turns out the flow meter is off by nearly 40%! I know this is not a great test, but the regulator was almost exactly correct.





So I know I totally could have bought one of the flow meters that you pop onto the cup of the torch. But if I get a cheap one from ebay, how do I know it's correct? I'll be back to square one. As ridiculous as this method is, it actually produced some interesting results lol. What do you guys think? Guess I'll return the crappy amazon flow meter.

[Meltedmetal]

While it would be nice if the cheap "stick it over the cup" flow meter was accurate, for your test it could give you an indication of the difference in volumes as long as it was consistent. If you keep on practicing TIG you will eventually recognize inadequate gas coverage and meter/gauge be damned you'll turn it up or learn to turn it down to what is sufficient. In your current experiment you need to have your balloons calibrated to ensure that one is not holding at higher pressure than the others. Always fun to experiment though.

[Oscar]

While it would be nice if the cheap "stick it over the cup" flow meter was accurate, for your test it could give you an indication of the difference in volumes as long as it was consistent. If you keep on practicing TIG you will eventually recognize inadequate gas coverage and meter/gauge be damned you'll turn it up or learn to turn it down to what is sufficient. In your current experiment you need to have your balloons calibrated to ensure that one is not holding at higher pressure than the others. Always fun to experiment though.

Or simply run the experiment about 10x with each balloon to ensure some kind of consistency, and then swap the balloons with respect to the regulators and run another 10 trials. Then get averages and compute final figures.

[Mach_Zero]

While it would be nice if the cheap "stick it over the cup" flow meter was accurate, for your test it could give you an indication of the difference in volumes as long as it was consistent. If you keep on practicing TIG you will eventually recognize inadequate gas coverage and meter/gauge be damned you'll turn it up or learn to turn it down to what is sufficient. In your current experiment you need to have your balloons calibrated to ensure that one is not holding at higher pressure than the others. Always fun to experiment though.

This is an excellent point and I actually did do this shortly after posting lol.

I used the same balloon for both and got the same results as the first tests. Still about 40% off for the flowmeter.

Though I don't really think it would make a big difference since I'm not expecting these balloons to hold pressures above a fraction of a psi. And I believe the flowmeter operates at 50psi, so I doubt the balloon pressure will affect the results. Also the flowmeter was reading 25scfh for the duration of the test. If there was significant back-pressure from the balloon, the ball bearing would have to fall slightly in response to the reduced flow.

Either way, this flowmeter is garbage. But it's good to know that the original regulator works pretty well. I had my doubts since it's a harbor freight item, but I'm definitely impressed.

[Mach_Zero]

Or simply run the experiment about 10x with each balloon to ensure some kind of consistency, and then swap the balloons with respect to the regulators and run another 10 trials. Then get averages and compute final figures.

There's definitely a bunch of experimental error in this methodology. I could reduce the error significantly by using a large trash bag or much larger balloons. If I run the test for 60 seconds or greater, the timing error is reduced (I just don't want to waste that much argon!). But yeah I don't think the back-pressure from the balloons will have a significant effect on the test since the torch side of the regulator/flowmeter run at pretty high pressures.

But with that being said, I doubt there's that much error. And even without the balloons, I was already skeptical of one of the devices because I could tell they were very different. Holding my hand up to the torch while it was reading 25scfh was very different for each device. But I just didn't know which device was wrong. And that's what prompted the balloon test lol.

[Welder Dave]

There's always a regulator whether it's a flow gauge or a flow meter. The set outlet pressures can vary greatly and these pressure differences could throw a loop in your experiment. A simple hand held gauge on the nozzle would tell the CFH.

https://www.norco-inc.com/2381755/Pr...-32_Flowmeters

[Oscar]

There's definitely a bunch of experimental error in this methodology. I could reduce the error significantly by using a large trash bag or much larger balloons. If I run the test for 60 seconds or greater, the timing error is reduced (I just don't want to waste that much argon!). But yeah I don't think the back-pressure from the balloons will have a significant effect on the test since the torch side of the regulator/flowmeter run at pretty high pressures.

But with that being said, I doubt there's that much error. And even without the balloons, I was already skeptical of one of the devices because I could tell they were very different. Holding my hand up to the torch while it was reading 25scfh was very different for each device. But I just didn't know which device was wrong. And that's what prompted the balloon test lol.

Yes I understand, there will always be experimental error, even with the initial testing such as with a balloon not being a sphere, and neither would an inflated trash bag be; the balloon is still the better choice between those two. But from a statistical standpoint, if the variables are carefully controlled by the operator during trials, the averages help smooth things out quite a bit.

[M J D]

Personally I wouldn't worry about initial numbers. Do some test welds. If a number 7 cup takes 18 cfh instead of the recommended 14 so be it. After you confirm what works use your test gauge to verify your cfh settings

[danielplace]

Edit:
As I see Dave has already mentioned,

Your test will never match unless they are both the same pressure when not flowing gas. The initial dump from a 30 psi reg compared to one at 80 would be huge even at exact same verified flow rate settings.

[shovelon]

So the question is which one is accurate?

[Mach_Zero]

Personally I wouldn't worry about initial numbers. Do some test welds. If a number 7 cup takes 18 cfh instead of the recommended 14 so be it. After you confirm what works use your test gauge to verify your cfh settings

This is probably the best answer lmao. But as a beginner, I tend to rely on settings from recommendations. And if I see online that it says to use 10-12scfh for a #5 cup, then I'll set it there when starting out. Then when my welds are crappy, I won't know if it's a technique issue, or not enough gas. So it helps to just have a good reference. I wasn't expecting it to be so wrong on the flow meter.

Edit:
As I see Dave has already mentioned,

Your test will never match unless they are both the same pressure when not flowing gas. The initial dump from a 30 psi reg compared to one at 80 would be huge even at exact same verified flow rate settings.

I had not considered that. I'm not sure what the regulated pressure is for the regulator, but the flowmeter says 50psi. But either way I'm sure I could control for that by starting the gas before putting on the balloon and then quickly putting it on the torch and timing it like that. Good point though.

So the question is which one is accurate?

Well from my tests, the regulator is pretty damn close to the expected volume with about a 5% error. The cheap flow meter is off by about 40%.

[Mach_Zero]

Also if I wanted to get crazy about it (which I have been known to do), I could also test at different flow rates and plot the error through the range of flow rates. Maybe the flow meter is accurate at 10-15scf and not so much at higher flow rates? But at this point I don't really care. I'm satisfied with knowing that the regulator is pretty close and I'll just keep using that. From there I can just adjust the flow depending on the results of the weld. But it's nice knowing when I set it to 15scfh, it's not at some completely different flow rate.

[M J D]

You will get further with your welding by welding practice rather than science experiments.

[Mach_Zero]

You will get further with your welding by welding practice rather than science experiments.

Well I guess that's why I'm a scientist and not a welder

But anyway I thought it was interesting and felt like maybe someone else would find it interesting too.

[Kelvin]

Turns out the flow meter is off by nearly 40%!

I doubt that.

[G-ManBart]

You probably considered this, but most of the cheap flow meters have more than one scale for different gases/mixes so you have to make sure you're using the correct one. Also, some tell you to read off the bottom of the ball, and some the middle of the ball. The wrong scale and the wrong spot on the ball could really throw off your numbers...just a thought.

[Jack Ryan]

Well I guess that's why I'm a scientist and not a welder

But anyway I thought it was interesting and felt like maybe someone else would find it interesting too.

Well, it is interesting even if, in the end, it is not all that useful. It is true what others have said, you need what you need irrespective of what the flow meter says. It would help though if flow meters indicated the same flow - particularly if you are documenting a procedure or just giving helpful advice.

Two points:

1. They are both flow meters consisting of a regulator and the actual flow meter. One is not a "regulator" and the other a "flow meter".
2. A flow meter will dispense a set flow in litres per minute (or other unit) at STP. If you are using the flow to inflate a balloon, the pressure increases way beyond standard pressure.

Also, the gauge type flow meter is indicating flow - when there is flow. When the flow stops, the reading increases so it is very inaccurate then.

I am happy if my flow meters are consistent and repeatable and have a scale that covers just what I need (easer to set). If they are accurate as well then that's a bonus.

Jack

[Oscar]

Well I guess that's why I'm a scientist and not a welder

But anyway I thought it was interesting and felt like maybe someone else would find it interesting too.

I had a feeling you were somewhat like me. Just the fact that you whipped up that excel/spreadsheet told me you're familiar with data analysis and what not. Just the fact that you know which variables even exist, can vary from reality, and need to be checked says that you will learn pretty quickly. It's this kind of analysis that will eventually make you more self-aware of both the big-picture and the micro-details. The vast majority of "help me" threads are a result of the beginning user not knowing what can even go wrong in the first place; the self-analysis is not there just yet. They don't know what to check and how to check it. I have a feeling that won't be you all that much.

BTW, I'm the resident mad scientist AFAICT.

[Mach_Zero]

I doubt that.

Certainly understandable given my shoddy experimental methods lol. But I can even feel and hear the difference! Even without a balloon or anything on the torch. There's a very distinct sound difference between the two when both set at 25scfh. One of them has to be wrong.

You probably considered this, but most of the cheap flow meters have more than one scale for different gases/mixes so you have to make sure you're using the correct one. Also, some tell you to read off the bottom of the ball, and some the middle of the ball. The wrong scale and the wrong spot on the ball could really throw off your numbers...just a thought.

Great points G-Man! I actually wasn't sure how to read the ball meter so I googled it (this cheap meter didn't come with any documentation). From my research, it seems you're supposed to read to the center of the ball, but even increasing it a bit until using the bottom of the ball, there's not a big difference in the flow rate.

Also, yeah I did use the Argon scale, but the CO2 meter is barely different on this meter. Great points!

Well, it is interesting even if, in the end, it is not all that useful. It is true what others have said, you need what you need irrespective of what the flow meter says. It would help though if flow meters indicated the same flow - particularly if you are documenting a procedure or just giving helpful advice.

Two points:

1. They are both flow meters consisting of a regulator and the actual flow meter. One is not a "regulator" and the other a "flow meter".
2. A flow meter will dispense a set flow in litres per minute (or other unit) at STP. If you are using the flow to inflate a balloon, the pressure increases way beyond standard pressure.

Also, the gauge type flow meter is indicating flow - when there is flow. When the flow stops, the reading increases so it is very inaccurate then.

I am happy if my flow meters are consistent and repeatable and have a scale that covers just what I need (easer to set). If they are accurate as well then that's a bonus.

Jack

Yeah I totally agree Jack. I know that I can vary the flow based on the weld results but still it's just nice to have accurate numbers so that when I change materials or cup sizes, I can look up in the literature what the flow should be and hopefully I'll be close.

For the gauge type meter, I take the reading while there is flow. I'm aware the pressure increases when there is no flow but I set it while the pedal is pressed and the solenoid opens.

And for the ball float type, I made sure that it was reading 25scfh the entire time it was filling the balloon. I just don't buy it that the balloon increases pressure so much to alter the flow by any significant amount. It didn't even inflate the balloon halfway. These balloons can easily be inflated by humans, so I'm assuming they're way lower than 1psi.

[Mach_Zero]

I had a feeling you were somewhat like me. Just the fact that you whipped up that excel/spreadsheet told me you're familiar with data analysis and what not. Just the fact that you know which variables even exist, can vary from reality, and need to be checked says that you will learn pretty quickly. It's this kind of analysis that will eventually make you more self-aware of both the big-picture and the micro-details. The vast majority of "help me" threads are a result of the beginning user not knowing what can even go wrong in the first place; the self-analysis is not there just yet. They don't know what to check and how to check it. I have a feeling that won't be you all that much.

BTW, I'm the resident mad scientist AFAICT.

I have a feeling we're gonna get along just fine! I have a background in research and testing so I tend to look at everything in that regard.

Unfortunately I may be on the other side of the spectrum. While most beginners might be posting "help me" threads, I'm here posting annoying pedantic over-analysis threads.....

[Munkul]

your experiment is interesting, but the data is completely flawed unless we know the regulators upstream of the flowmeters were both set at identical pressure.

Since the US style is to integrate a pressure regulator into a flowmeter with a single pressure gauge indicating tank pressure, we have no way of knowing.

You can have a higher flowrate at lower pressure and it may inflate a balloon less than a low flowrate at high pressure.Just saying, there's a variable which you have no control of, and also no way to measure it.

Over here, the flowmeters are nearly always sold separately to the regulators - and you set them independently. It's better for tinkering, but worse for getting it wrong - it's too easy to crank the pressure up and end up with a whole load of start-up surge.

If you keep on practicing TIG you will eventually recognize inadequate gas coverage and meter/gauge be damned you'll turn it up or learn to turn it down to what is sufficient.

^^ this, completely

You don't *need" x flowrate of argon. you *need* a consistent flowrate at what's correct for your setup, your style, and your job, and to know what numbers worked for you last time. That only comes with experience and some of us are still learning...

[MinnesotaDave]

Interesting experiment.

This is the flow checker I use.


I use it to get repeatable results no matter what meter I'm using.

If you get one, it'll also help you verify your testing methods for you - since you are looking at differences in flow.

Nice thing is no wasted time wondering if the rate you wanted is what you are getting.

[Mach_Zero]

your experiment is interesting, but the data is completely flawed unless we know the regulators upstream of the flowmeters were both set at identical pressure.

Since the US style is to integrate a pressure regulator into a flowmeter with a single pressure gauge indicating tank pressure, we have no way of knowing.

You can have a higher flowrate at lower pressure and it may inflate a balloon less than a low flowrate at high pressure.Just saying, there's a variable which you have no control of, and also no way to measure it.

Over here, the flowmeters are nearly always sold separately to the regulators - and you set them independently. It's better for tinkering, but worse for getting it wrong - it's too easy to crank the pressure up and end up with a whole load of start-up surge.

I'm having trouble understanding this. What does the upstream pressure have to do with the flowrate downstream? The purpose of these devices are to offer a specified flowrate at atmospheric pressure. They should all be calibrated to produce the desired flowrate no matter what their upstream pressure or internal design looks like.

If I dial in 25scfh on a flow meter, it should provide 25scfh. The balloon provides very little back pressure. It should be considerably less than 1psi.

As far as the start-up surge, the only way this can alter the results is to provide extra total flow. If the device is providing a perfect 25scfh at all times other than the start-up surge, then it would have more gas in the balloon than it's supposed to. Clearly this is not the case.

I personally don't think any of these arguments are disproving the results I have showed. Balloon pressures are almost negligible, and start-up surge would only cause an overflow in the balloon rather than underflow.

Interesting experiment.

This is the flow checker I use.


I use it to get repeatable results no matter what meter I'm using.

If you get one, it'll also help you verify your testing methods for you - since you are looking at differences in flow.

Nice thing is no wasted time wondering if the rate you wanted is what you are getting.

This is a good idea to have the same flow checker as a reference no matter what. Because even if it's wrong, you still have a reference and follow the same device each time.

But this experiment was also to troubleshoot some issues that I was having. If you bought a regulator and you had to turn it up to 50scfh to get a decent weld, wouldn't you wonder if there was something wrong?

At least as a beginner that doesn't have confidence, it's easy to wonder if I'm doing something wrong or if the meter is wrong. And that's where I believe the value of these tests lies. I was able to verify that this new flowmeter is completely off so I don't have to wonder if it's a technique issue. It's easy for an experienced tig welder to hear or feel the gas flow at the cup and know that it's too low or too high. But a beginner like me has no experience to reference whether it's too low or high. And all I have is the number on the regulator until I gain experience.

And do you think it would be best for a beginner to learn on a damaged flowmeter that I have to turn it up to 50scfh to get 15scfh? I think that would be a terrible idea so that when I get a new machine I'll turn it up to 50scfh and wonder why the torch sounds like an air compressor blower lol.

[G-ManBart]

I'm having trouble understanding this. What does the upstream pressure have to do with the flowrate downstream? The purpose of these devices are to offer a specified flowrate at atmospheric pressure. They should all be calibrated to produce the desired flowrate no matter what their upstream pressure or internal design looks like.

The Harris 355 flow meters that I put on my TIG welders are all set for 20PSI delivery pressure. The Victor on my MIG is set for 25PSI delivery pressure. I don't think they would work well at atmospheric pressure.

[Mach_Zero]

The Harris 355 flow meters that I put on my TIG welders are all set for 20PSI delivery pressure. The Victor on my MIG is set for 25PSI delivery pressure. I don't think they would work well at atmospheric pressure.

Right, this one works at 50psig. But the torches are all at atmospheric. So their purpose is to ultimately deliver a flow rate at atmospheric, right? Unless I'm fundamentally misunderstanding something (which is completely likely lol).

I was just trying to say that the balloon shouldn't create a significant enough backpressure to change the results.