Zero and Span Calibration

Because they don't weigh accurately unless this is done.

(It may have already been done at the factory; doing it again corrects for any shift that might occur during shipmant.)

Cheating is ubiquitous. There are homework cheaters and sellers who cheat buyers by not providing a true weight. Surely one can recognize the other?

Sounds like homework to me!

thank you for your good answers gentlemen. Additional question. when the system has undergone Zero calibration and consequently, it reads out zero considering no process material. Why is there a need to perform span calibration when the vessel is filled with process material or a percentage of full scale?

Because of you don't do the span, maybe there would be 100 kg of material in the scale, but the scalewould read only 50 kg (or 200 kg, or whatever.)

It is assumed that the system is linear and several known (calibrated ) weights are "measured", in several steps as well the zero as the span are changed till the weights and the "display" are identical within the tolerance limits accepted for the weighing uncertainty.

Good Answer but you meant to say:-

Calibration of an analog instrument involves adjusting the M (Span) and B (Zero) of the instrument so that the resultant Y (Output) for any given X (Input) is correct.

I want to responnd to the first question.

In the field of engineering quantities are measured to know the values in other to ascertain if there are within the normal range or not, hence the need for instrument calibration.For all measuring instrument, there is a range

For any instrument to measure or weigh accurately, the must be no zero and span error.

So zero calibration helps to remove zero error and return the measuring instrument to a starting point of zero.

Span calibration helps to determine and assign a limit to which the instrument should weigh. That explains why most instruments have inscriptions like range 0-200gr.

In general calibrations are NOT done with only 2 points since errors can give a wrong "line". If the error is let say "+" at one end and "-" at the other the line does not have the right slope thus values are wrong.

The points are in general as far as possible from the middle of the range in order to reduce above mentioned effect.

Depending on the required uncertainty the number of points is at least 3 and can go as high as 10. The higher the number of points the lower the uncertainty at the end.

There are 2 possibilities to calibrate a sensor: either with physical known values as mentioned (melting ice and boiling temperatures ) or via a comparison with the output of a reference sensor/device. For instance for pressure calibration it can be done either with a pressure generator using weights (very high precision) or with an other sensor calibrated and used as a transfer reference. It is much more easy to calibrate parallel measures as pressure, temperature or velocity as serial measures as weight, force or torque where the reference sensor has to be placed in series with the controlled one in order to have both "feel" same "load".

wouldn't M be the span and B the zero

When you purchase some grocery suppose, when you pay for 1kg, and the weighing machine shows 1 kg, would you like to get only 1/2 kg of it only?

That simple

When you do a zero and span calibration you are fixing two points to a set gradient for eg. 0mV=0kg and 10mV=100 kg it follow 5mV will be 50kg and so on depending on the weight.Over time this gradient can drift,Depending on loadcells and weights a small drift can mean a large percentage error.

Doing Zero and Span calibration is not specific to weighing instruments. It is required for any measuring device. This is to offset the drifts that has taken place over a period of time. To negate the drifts the calibration is to be done periodically as advised by the manufacturer.

Why it is required before putting into service?

In a weighing instrument the 'g' factor plays an important role. Since the 'g' varies between places all over the world, it is recommended to do a calibration of the weighing instrument at the installed location before it is being put into service.

Discussion is great. Thank you gentlemen for your bright ideas..

Any measuring instrument, may it be weigh scale, has a particular capacity or range/scale for measuring.If it is factory set and calibrated it is supposed to give accurate measurement ( in weigh scale case the weigh in KG or Tonnes as the case may be), through out the range. If the weigh scale is a precision one for very sensitive measurement like a goldsmith scale, or even in larger scale like truck /wagon weigh bridge one needs to get checked its acuuracy on part of measurement and for that purpose, one needs to get it calibrated the same with any other standard cerified method/agencies. In Calibration the ZERO and SPAN calibrations are necessary to see that the measurements are correct throughout the range. Once ZERO is adjusted, SPAN CHECK/adjustment is necessary repeatedly until no more corrections are required. After ZERO and SPAN need no more adjustments, the scale or instrument is supposed yo behave normally and linerally throught the scale. Calibration of an instrument is necessary depending on its frequency of usage,purpose of usage, type and accuracy etc.

I have worked as a weighing instrument engineer and know that the balances were set right at the factory but almost always needed further fine tuning on site. There are many reasons for this being necessary.

In balances that use knife edge pivots they need protection from impact so in transport the beam is locked in the raised position. This puts different stress on it to when in use. Beams need to be as light as possible to aid accuracy and reduce wear. This means it is easy to introduce minute distortion which affects accuracy. They also expand and contract by noticeable amounts through temperature changes so need to be set at the temperature they will be used in. Indeed a good jolt in transit might have altered the alignment of the scale relative to the beam, and so on . . .

Think about the accuracy asked from a weighing machine. It might register one in a million units. Ideally you would want the knives to be parallel in two planes to within one millionth if the distance between them. If you achieve that, you probably can also lose it quite easily, in transport and in use. Similarly the scale adjustment.

This is more important for the more pivots that are involved in the machine. (Weigh bridges and some balances do not have freely swinging platforms like a typical chemist's balance and this is achieved by having more beams and therefore pivots and adjustments and potential errors.)

In practice you can probably get adequate performance from less tight limits but the accuracy will suffer a bit. In particular the linearity will be affected and this will not show up with just end of scale checks. Other checks, in particular, quarter, half and three-quarter span readings, should be obtained. For this you need calibrated weights of appropriate values. Another check is to make sure the span is the same with minimum load as well as with maximum load. This is sometimes eliminated by designing the balance to always work at full load (termed constant load in the trade) in which case the items you weigh are substituting for some of the inbuilt weights which have to be lifted free to regain "balance".

With scales used for measuring cooking ingredients these sort of errors aren't crucial so you can unpack and use it 'straight from the box' but in other spheres you need to have a proper check made to be sure you are getting what you have paid for.

Finally, in any instrument, it is essential to re-check the zero and full scale positions after every adjustment session since one will always affect the other and you could be ending up with no room for further user adjustment in one direction and even might have parts of the assembly losing their running clearance.

Those are the problems of mechanical systems and cannot be avoided since friction and deformation is totally related to mechanical structures. The problems you mention are less present if not totally absent in weighing systems using load cells. And as far as I know almost ALL modern systems are based on load cells even for very high loads. the introduction of load cells was accelerated by the cost related to the maintenance of the mechanical systems.

Now those systems have their problems too. One of the most frequent source of errors is the differences in sensitivity of the different load cells which generate an erroneous output when the load is not perfectly centred (same problem with 3 or 4 LC). There are ways to completely eliminate those errors but many are not aware of them and do not know how to do. I was once confronted with a system where the position gave about 5% error in the weight on a commercial balance.

I have a personal experience with the design and the manufacturing of a weighing system for loads up to 4000 metric tones on 8 loads cells the results were repetitive within less than 0.1%. the system was dedicated to the weighing of off-shore modules and to the computation of the COG position (important for the transport and lifting). The compensation was so good that even if the load was changed as distribution the result did not move out of above uncertainty range. the advantage of such a system was that it was able to assembly it and after the operation to disassembly and transport on an other spot without all the problems you mentioned. It was a building site weighing system which maintained its precision which ever the conditions were.

You will argue that the other type an measure with less errors, I agree but the specification was for 1% maximal error in all conditions and we were at about only 1/10 of it so that we did not try to do better but it could have been possible.

What is meant by zero and span calibration?

@Nickname

This discussion looks great.

I've a doubt regarding the load cells. Whenever an equipment is mounted on Load cells. How exactly the load sharing takes place. Of course this depends on how many load cells are installed & their respective capacities but there are some points like centre of gravity or centre of mass & angle(circularly installed LCs). do they play any role in the sharing of load?

thanks & regards,

Amey D

If there are more than 3 load cells then the system is hyper-static and the loads is distributed according to the relative stiffness at the support points. When we speak about stiffness we have to consider as well the structure as the load cell itself and of course the support of the last.

If calibrations have been made the right way it is possible to show that the weight is an invariant which ever is the position the load cells or they are changed as position under the load. I had such a demonstration with a difference less the uncertainty !

http://www.isa.org/Template.cfm?Section=books3&template=Ecommerce/FileDisplay.cfm&ProductID=7577&file=ACFBA59.pdf

this will help in understanding the principals and terms used in calibration.