# How to specify analog gages for pressure and temperature?

### Specify your gages for the middle of the range

As a process and test guy and all-around controls technician, I started my career in calibration working on pressure and temperature gauges. I spent a lot of time becoming intimate with the inner workings of these analog gages.

You know the kind.

They have a circular face with a needle anchored the dead center of the face.  I became very familiar with curved bourdon tubes and circular bi-metallic strips with return springs.

### Percent of Full scale

I can remember one of my first calibration jobs for a customer having a large RO water system and spending the whole day calibrating pressure gages with my dad, who first introduced me to calibration in 1997. The customer wanted all their pressure gages calibrated to 5%. So, when I came across a gage that was reading well off its zero, I asked my Dad if it failed. He said we needed to check the other calibration points before we know. Sensing my confusion, he explained to me that pressure gages like these were specified based on the percentage of the full scale, not the reading on the gage, and that meant that this 100 PSI gage had to be within 5 PSI at any point in its measuring scale. That made more sense to me, once I thought about it.

### How well can you read a gage

Since then I’ve calibrated a lot of pressure gages and their close cousin’s temperature gages, and I’m amazed how confident everyone is, in their performance. All of these devices suffer from a compromised design. Higher accuracy units have more and more mechanical compensation to account for the issues, but these come with a higher cost as well. The basic issue is that these devices are designed to function best near the center of their operating scales. The most common process pressure gages are of the 3-2-3 variety. They have an accuracy of 3% in the bottom 1/3 of the scale 2% in the middle 1/3 of the scale, and then back to 3% at the top third of the scale, and these are all in percent of full scale.

So, if you try to read 10 PSI on a 100 PSI gage, then you can have an error of 30% of the reading, and be within the calibration tolerance. Psst, auditors learned about this years ago, and love to go down that rabbit hole.

Another thing to keep in mind is the graduations on the gage itself. The number one driver of measurement uncertainty for analog gages like these is their resolution. So, feel free to read between the graduations, there is nothing wrong with that, just know your minimum uncertainty will be roughly half of the smallest graduations on the device.

### How to specify a gage

When purchasing gages always consider your process first. All too often we consider the max pressure a system might receive and then do our ordering based on that. What we need to also consider is what the typical measurement will be, how much it will fluctuate, and how accurately we need to be able to measure it. If you need to measure a wide range of values accurately consider digital devices.  If you are worried about leaving a more expensive device online all the time use an analog age, but add a port next to it that you can tap into with a digital device when more precise readings are required. If your process pressures fluctuate rapidly, consider liquid-filled gages. These use a liquid in the gage to act as a shock absorber for the internal working of the device to prevent damage. Think real-world with your specifications.

Does the pressure really need to be read to 1% or is 5% sufficient?

Research how much the difference will impact the product.

Don’t assume anything. There is very little standardization across manufacturers, and when it comes to cheap gages, you get what you pay for…if you’re lucky!

Still, have questions on how to select the correct gage?