Input resolution ?

When you digitize any analog signal it will have distinct equal spaced values. As an example a +/- 10.000V dc input or output will be represented in the controller by a range of binary integer number values. The number range depends on how many bits of resolution are used, which maybe low resolution of typically 8 bits = 256 values or steps to cover the range or increasing resolution for 10, 12, 16 or higher bits in the number of steps for resolution for the range.

Thomas,

A digital signal is not a smooth line but a series of small steps. Resolution is the amount of one step compared to the total scale. An on/off switch is a digital device with a resolution of 1:2. A temperature measurement system with a scale of -40 to +40 C that gives readings to the nearest 1 C has a resolution of 1:80. You could also say it has a resolution of 1 C. My bathroom scale reads to the nearest 0.5 pound and has a full scale of 300 lbs. has a resolution of 0.5 lb or 1:600.

When analog signals are received by digital devices they are converted by an ADC (analog to digital converter). Let's say that an air pressure measurement of 0 to 10"WC is converted to a single byte of computer memory. Then the ADC can be said to have a resolution of one byte, or eight bits, or

1:256 ( 2^8 =256) or 10"WC / 256 = 0.04"WC.

Keep in mind that any resolution exceeding the precision of the source measurement is of no value. If your instruments are good to 1%, (1:100) any ADC resolution greater than 8 bits (1:256) isn't worth much.

Higher resolution costs more and it is easy to pay for something you don't need.

Walter.

Note that in Walter's 1% ( 100 parts input FULL range resolution ) example, that the 8 bits = 256 steps. This is the correct digital range that allows for 2 steps minimum per the 100 parts of a 1% resolution input signal range. There should be at least 2 steps / bits in the digital range per resolution step in the analog input or you will cut the resolution in half for some readings that are on the border between two steps. You must also consider if the analog range FS and Digital FS range are the same, 0-5Vdc signals are often feed into +/- 10V ADC = 1/4 the digital range covers the input range.

A simple rule is that a 3.5 digit multimeter has 12 bits of resolution and a

4.5 digit multimeter has 16 bits.

Caution, instrumentation resolution is often based on percent FULL SCALE as this is the low value. A better and truer resolution number is percent of READING. Thus scaling and the linearity will also make a difference. Even higher resolution ADC's are often used in conjunction with a calibration table to allow for non-linear inputs in dedicated measurements.

I apologize for the delayed response but I would like to thank both of you for taking the time to answer my questions.

Thomas

There are 4 parameters you need to worry about.

1 Resolution. A 13 bit plus signe digital to analog convertor divides a 0-10v singnl into 8192 steps. (ie about 1.25mV). If measuring a load cell mass in the range 0-100kg you could resolve down to +/-12.5 gramms.

2 Analog Linearity.

All that accuracy will do you not necessarily be of use you if the analog linerity of your front end is not up to it.

An ordinary Siemens 14 bit A to D card for an S7 has a linearity of

1.0% with 0.3% losses due to temperature coefficient, linearity and hysteriesis etc. (Its generaly much much better than that but that is what they quote)

So if the card is used to measure a weight in the 0-100Kg range it could be of by +/-1kg. Due to the high resolution it would however detect a 12.5 gram change in weight!

A high accuracy siemens analog card has a linearity of 0.05% to 0.1% instead of 1% and it has 15 or 16 bit resolution.

The high accuracy cards cost about 60% more.

If you are trying to control a process: eg dosing or saving of material you may need the higher accuracy. In an oild refinery saving

0.5% material can be a lot.