Analog Pressure Gauge Calibration - smallest resolution practice

I've been a Metrologist for many, many years..back when Lowry was teaching it..so am having a hard time teaching these whipper snappers some techniques...

Anyways, got a question, maybe some of you Old Schoolers can help...

Set up:

Analog pressure gauge, 0 - 100 PSI, minor increment 1 psi

I recall I was trained to "interpret" the space between the divisions. I.E. 1/2 way would be 0.5 PSI, "not quite half" 0.25 PSI.

Does anyone have any documentation they can direct me to that teaches that technique? or in fact teaches how to perform that type of analog measurement?

I'm in a debate with some peers and the frame of thouhgt is to aaply pressure until the UUT reads a nominal set point, then read the digital pressure source, and record it.

The debate then becomes which is correct? You get more resolution reading the digital standard, however the UUT is the devce under test. So, philosophically that would be incorrect.

TIA

Re: Analog Pressure Gauge Calibration

Welcome to the Cove!

I think I recall reading a specification on calibrating analog pressure gauges some time back, but for the life of me, I cannot recall what it is.

Basically I have always seen and utilized the smallest resolution practice, which is 1/2 between two marked increments. So in your case, you have marked resolutions at 1 psi. So if I apply a pressure and the gauge displays "something" greater than 50.0 psi and less than 51 psi, I can reasonably infer it to be 50.5. Anything more or less would be mere speculation and in my opinion, not really necessary anyway. Five people could agree its greater than 50.0 and less than 51, so settle on half. But anything finer than that, and you are going to have arguments. Then... try to explain how you read 50.6 to a customer.

So I have always noted the best resolution possible is 1/2 the smallest resolution. That applies on chart recorders also.

With analog pressure gauges, I typically perform a reverse calibration. I apply a pressure to hit a nominal reading on the pressure gauge, then record the reading from the pressure standard. So I would apply pressure to achieve 50.0 on the analog gauge, then observe what the standard displays. I'm still able to determine what error exist in the gauge, and can accurately record the error to .1 psi. In your case, if you have a 0-100 psi gauge with 1 psi resolution, I'm guessing the tolerance is 1psi.

Just if you decide to do a reverse calibration, I would mark it as such on the calibration certificate. That way people know how you were able to accurately record the error that you do.

Anyway, I know I did not provide the documentation you asked for. But hopefully, something in my posts gives you something helpful.

Re: Analog Pressure Gauge Calibration

You can interpret the space betqween the numbers but for calibration purposes it is in 1psi increments, you cannot guess a reading more than .5 +\-

Bazzle

Re: Analog Pressure Gauge Calibration

Hi,

I know it is not from your neck of the woods but maybe BS EN 837-1:1998 (Pressure gauges: Part 1. Bourbon tube gauges – Dimensions, metrology, requirements and testing) may help – can get it from BSI (www.bsiamerica.com). This covers analogue scales, different accuracy classes for gauges, permissible errors, scale intervals etc. Sorry I cannot supply a copy (copyright etc) and I know I may be teaching grandma to suck eggs here etc.

Steve

Re: Analog Pressure Gauge Calibration

That's precisley what's being done. However now were into picking Nits.
So, because of this method, the argument is now were not really performing a calibration of the guage (or something like that.) The argument is that the nominal amount is supposed to be applied t the UUT, so therefore it's never going to fail because it's always set to the nominal amount. Damn Nits sure get in the way.

My experience, via the US NAVY and US Air Force & US Civil service, is that your applying a nominal amount, and taking a measurement. the standard is just that "The Standard" and your DUT, UUT or TI, is the device being tested, so that is the device you measure.

Re: Analog Pressure Gauge Calibration

SteveK said:

Hi,

I know it is not from your neck of the woods but maybe BS EN 837-1:1998 (Pressure gauges: Part 1. Bourbon tube gauges – Dimensions, metrology, requirements and testing) may help – can get it from BSI . This covers analogue scales, different accuracy classes for gauges, permissible errors, scale intervals etc. Sorry I cannot supply a copy (copyright etc) and I know I may be teaching grandma to suck eggs here etc.

Steve

Click to expand...

Thanks, I'll hunt it down, I appreciate your response. No worryss on the copy, I understand.

Cheers

Re: Analog Pressure Gauge Calibration

bazzle said:

You can interpret the space betqween the numbers but for calibration purposes it is in 1psi increments, you cannot guess a reading more than .5 +\-

Bazzle

Click to expand...

I agree, thanks. That's what I've been taught,

Re: Analog Pressure Gauge Calibration

VP93Navy said:

That's precisley what's being done. However now were into picking Nits.
So, because of this method, the argument is now were not really performing a calibration of the gauge (or something like that.) The argument is that the nominal amount is supposed to be applied to the UUT, so therefore it's never going to fail because it's always set to the nominal amount. Damn Nits sure get in the way.

My experience, via the US NAVY and US Air Force & US Civil service, is that your applying a nominal amount, and taking a measurement. the standard is just that "The Standard" and your DUT, UUT or TI, is the device being tested, so that is the device you measure.

Click to expand...

I think you know what you're doing, and have this figured out. I think the whipper snappers are making you second guess yourself.

There are many different definitions of calibration (another thread; another time), but essentially, you are verifying accuracy. You are using a traceable standard with substantially greater accuracy than your UUT. That is calibration.

It matters not where the nomimal amount is being applied. It matters that the error present in the UUT can be measured. If the analog gauge is displaying 50.0 PSIG, and my standard says 49.3, who cannot accurately state that the UUT is reading high?

You've messed with gauges enough to know you'll see a predictable pattern of error up to the full span of the gauge from there (assuming the zero was not initially off).

Once you get it adjusted, both instruments should read 50.0psi, and your mission is accomplished.


I got initially trained on gauge work by a wonderful old gentlemen from the Navy. He was on a minesweeper sub in WW2, and with instruments you made them work. Period.

Re: Analog Pressure Gauge Calibration

The 'interprited' resolution on an analog scale is not always 1/2 the marked resolution. ASTM E4 discusses determining the resolution of analog scales, dials and recorders. A paraphrased version of this is its a ratio between the width of the pointer/indicator to the space between ticks/marks. If 4 pointer widths could fit between each mark than you can interpret to a quarter of the resolution or a ratio of 1:4. In some cases pressure gauges are a type of force gauge so it wouldn't be much of a stretch to apply this method to all analog pressure gauges. Another example is ASME B89.1.10, which requires that drop/deflection/dial gauges be read to the nearest 1/5 resolution.

Hope this is useful information on this point at least,
Tessa

Re: Analog Pressure Gauge Calibration

NTAinc said:

The 'interprited' resolution on an analog scale is not always 1/2 the marked resolution. ASTM E4 discusses determining the resolution of analog scales, dials and recorders. A paraphrased version of this is its a ratio between the width of the pointer/indicator to the space between ticks/marks. If 4 pointer widths could fit between each mark than you can interpret to a quarter of the resolution or a ratio of 1:4. In some cases pressure gauges are a type of force gauge so it wouldn't be much of a stretch to apply this method to all analog pressure gauges. Another example is ASME B89.1.10, which requires that drop/deflection/dial gauges be read to the nearest 1/5 resolution.

Hope this is useful information on this point at least,
Tessa

Click to expand...

That's precisely what I was looking for, can you give me a little more description? I went to ASTM site and can't find it.