Question: Why can't you calibrate SPG?

[Robert Gillcash]

SPG's can read very much incorrectly. Had a student show me his gauge before his last pool session, it showed his tank only having 2200 psi. I wasn't too concerned as it was a short dive and I knew he watched his gauge like a hawk anyways. Halfway thru the dive, when I asked him how much air he had, he signed 2200. Huh? Yup, gauge still showed 2200. At the surface, to do another bouyancy check, (dump air down to 500 psi etc.) his gauge did not move. Aborted the air dump for him of course.
When we got out, his gauge still read 2200. Removed reg unit from tank, pressure dropped to 1100, not even connected.
Obviously a faulty gauge, what was sad is that it worked perfectly fine the dive right before that, same day.
Unfortunately, they stop working. The real trouble, is figuring out when. If that gauge had read 3000 at the start and then just stopped at 2200, would you notice or just keep diving?

 

[g1138]

I don't buy that. If the pressure is zero, then the SPG will read zero. Very easy to verify. If it doesn't read zero with the tank turned off, then you know it's bad. Ok. So with the SPG reading zero at zero pressure, why would it jump up to 250 with an only 50 psi increase in pressure? While still reading 3000 at 3000 psi? That would be seriously off. I'm not saying that it's technically impossible, but the realistically plausible chance of this happening gotta be close to zero. It's much more likely that the inaccuracy goes the other way, i.e. the SPG reading zero when there's still some pressure left.

Most gauges work by a metal wire spiralling inside the gauge, your reading needle is connected to this wire. As pressure is exerted it contracts (the spiral gets smaller), the needle is turned, and your gauge reads a number greater than zero. Let say your wire is loosened over time due to metal fatigue or what not. It's exerting a force on your needle that wants to take the reading below zero, but because of the pin stop at your zero reading, it doesn't drop down. So your pressure reading is lower than what your tank pressure actually is. Not a problem per say.

Now lets talk about corrosion or metal fatigue. Your wire doesn't contract properly as it did brand new, so your reading can be higher than what you actually have. Big problem. It doesn't jump, it just increments at a very different rate from before.

Another issue can be corrosion of your wire, so it won't expand consistently as your tank drains. There was a post a month or two about an SPG failure, where upon banging his SPG, his pressure then dropped 200psi. Turns out his SPG would always stick at the half way mark.

So no, it's not that simple.

 

[dfx]

Doesn't work that way. Zero is a mechanical stop on cheap gauges. See second gauge in post #4.

Most gauges work by a metal wire spiralling inside the gauge, your reading needle is connected to this wire. As pressure is exerted it contracts (the spiral gets smaller), the needle is turned, and your gauge reads a number greater than zero. Let say your wire is loosened over time due to metal fatigue or what not. It's exerting a force on your needle that wants to take the reading below zero, but because of the pin stop at your zero reading, it doesn't drop down. So your pressure reading is lower than what your tank pressure actually is. Not a problem per say.

Yes exactly, I can see that happen.

Now lets talk about corrosion or metal fatigue. Your wire doesn't contract properly as it did brand new, so your reading can be higher than what you actually have. Big problem. It doesn't jump, it just increments at a very different rate from before.

I can also see that happen. In fact I did have a (rental) gauge once that was always reading 250 bar on regular 200-bar filled tanks. But during the dive, the needle just dropped faster than it normally would, and by the time it was getting below half, the inaccuracy was resolved. I know that because my buddy had an almost identical SAC rate.

But what I don't see how it could happen is the needle being on zero (or wanting to be below zero) at zero pressure, the needle also being on 3000 at 3000 pressure, but on 250 with 50 psi pressure. Or something like that. I'm thinking that if the needle shows a higher reading at low pressure, then it should also show either a higher reading at zero pressure, or a higher reading at high pressure. Unless there's something else seriously wrong with the SPG of course, like the needle getting stuck. But you would/should notice that, and that's something else than a simple inaccuracy.

Again, not saying that it's impossible, but I just can't envision any mechanical circumstances that would result in an otherwise normally functioning SPG that's giving you high readings only in a selective pressure range.

 

[oly5050user]

I don't buy that. If the pressure is zero, then the SPG will read zero. Very easy to verify. If it doesn't read zero with the tank turned off, then you know it's bad. Ok. So with the SPG reading zero at zero pressure, why would it jump up to 250 with an only 50 psi increase in pressure? While still reading 3000 at 3000 psi? That would be seriously off. I'm not saying that it's technically impossible, but the realistically plausible chance of this happening gotta be close to zero. It's much more likely that the inaccuracy goes the other way, i.e. the SPG reading zero when there's still some pressure left.

Buy it or not..I have seen plenty of gauges that stick and read wrong all the time.. You may read that you have 800psi but the needle is stuck there.Keep on breathing from it and it will still read 800psi when the tank has been sucked dry.

 

[robertarak]

I don't buy that. If the pressure is zero, then the SPG will read zero. Very easy to verify. If it doesn't read zero with the tank turned off, then you know it's bad. Ok. So with the SPG reading zero at zero pressure, why would it jump up to 250 with an only 50 psi increase in pressure? While still reading 3000 at 3000 psi? That would be seriously off. I'm not saying that it's technically impossible, but the realistically plausible chance of this happening gotta be close to zero. It's much more likely that the inaccuracy goes the other way, i.e. the SPG reading zero when there's still some pressure left.

How happy are you when you get a "good fill"? YAY! I got 3200! ....or you only got 3000 and your spg is off by 200. See where this is going?

 

[nimoh]

Here is a scenario. The guage was working fine on your last dive, so when you took it off of the tank it read 0. However, there is some corrosion inside the guage that is worsening over the course of the next couple of weeks in between dives. Now, on your next dive, you pressurize your guage and the needle is forced all the way up, and works just fine for most of the dive, but then you get down to about 200 and it sticks due to the corrosion.

Of course it could stick at higher pressures, but there is a greater likelihood of you noticing it since while checking your guage during your dive you would notice that two times in a row it was reading 1000psi.

 

[dfx]

Here is a scenario. The guage was working fine on your last dive, so when you took it off of the tank it read 0. However, there is some corrosion inside the guage that is worsening over the course of the next couple of weeks in between dives. Now, on your next dive, you pressurize your guage and the needle is forced all the way up, and works just fine for most of the dive, but then you get down to about 200 and it sticks due to the corrosion.

Of course it could stick at higher pressures, but there is a greater likelihood of you noticing it since while checking your guage during your dive you would notice that two times in a row it was reading 1000psi.

A sticking needle would definitely a problem. But that's not what I had in mind when I said that I don't buy that. I was thinking of a simple inaccuracy issue, since the thread is talking about calibration. Those are two different kinds of problems in my thinking. Kinda hard to explain. Anyway, it's not important, so carry on.

 

[lowviz]

...//... But what I don't see how it could happen is the needle being on zero (or wanting to be below zero) at zero pressure, the needle also being on 3000 at 3000 pressure, but on 250 with 50 psi pressure. ...//...

I'll try to be more clear. This type of gauge is most inaccurate at either end of the scale. Lets assume it passed final test and shipped with a plus or minus 2% error. 0.02 x 4000 psi = 80 psi. So they put zero somewhere behind the pin to keep you from seeing the error. Back on post #4, second gague, you will see that the pin is at the 100 psi mark.

So let's say that the true zero wandered somewhere so that the indicator is just ready to lift off the pin. You can't see this by looking at the gauge. In addition, there is the gauge error to consider. So if my gauge is reading 200 psi, I can be pretty sure that I have at least 80 psi (10x the error, sorry about the two 80's *confusing*) in the cylinder if it was reading zero before the dive...

 

[nimoh]

A sticking needle would definitely a problem. But that's not what I had in mind when I said that I don't buy that. I was thinking of a simple inaccuracy issue, since the thread is talking about calibration. Those are two different kinds of problems in my thinking. Kinda hard to explain. Anyway, it's not important, so carry on.

yes, I have always been skeptical about the inaccuracy discussions as well. If the guage is off by 200psi, at the very least you should be at least noticing that you seem to get consistently get better fills than your buddies

However, it is my understanding that guages are consistently more inaccurate in the lower ranges, and this isn't limited to pressure guages -- I am also under the impression that torque wrenches are off by a bit in their lower ranges.

although, I have never put any effort into verifying any of this.

 

[cocoajoe]

I appreciate all the comments and insight. In the pass I've read some 'horror stories'....one sticks out in my mind. This guy has like 4 or 5 regulators. He connected one then the other and so on. With a couple reading way diff than the rest Forget the exact details but it was a scary story.