 So, today's lecture is going to be on basic ideas behind measurement, so this is a departure from electronics which you have gotten some exposure to mainly pertaining to instrumentation electronics. So, we will just do a brief recap to understand where we are right now in the course and what we are trying to learn in the next couple of couple of weeks after which you will put everything together and actually conduct an experiment where you verify or validate a hypothesis. So, where are we what we have done so far is that we have learnt the structure of modern day instrumentation electronics, so that is what that is what we have learnt so far. So, structure of a modern day instrumentation system consists of a sensor which gives you an analog input typically need not always be analog, but oftentimes it is an analog input that we will call U of t which is then processed using a set of electronic components. So, you typically have an ADC to convert the analog information into digital information which is then sampled typically using a computational device like a microcontroller after which signal processing or data processing is done on the microcontroller possibly in real time that information is later on utilized in a variety of means. One way is to is if you want to see stuff on an oscilloscope you can reconstruct the signal and the process signal and verify if your processing makes sense etcetera or if you want to store the information that is coming out of microcontroller put it out through some communication device like a serial port or something and store it in your computer ok. So, the what we have learnt so far is this structure. So, you have an analog signal ADC processing through a microcontroller and then whatever you want to do with it ok. So, for you to so, this is this structure spans across the entire gamut of instrumentation electronics ok. So, that is the reason why we have we have focused on it. So, in abstract sense every measurement system where you are utilizing a computer to store information is going to ultimately look like this ADC and then a microcontroller. Therefore, you needed to have familiarity with voltages currents ADCs DACs microcontrollers some notion of digital discrete etcetera. So, all of that has hopefully been built to some extent during the lecture and then partly reinforced during the course of the experiments ok. So, that is that is what we have ended up doing so far. Now, the interesting questions for today I will come to them, but the nature of the questions will be different. We will be specifically interested in this measured signal U and more importantly the process of getting the measured signal U ok. So, I will come to the question sooner or later, but the focus is not on the electronics now. It is on how do you generate a measured signal U and what can you make you have to make sense out of the measured signal U ultimately you are trying to measure something so, that you can make sense out of it ok and the U is a representation of the measurement ok. So, we will we will discuss what I mean by representation, but the point is we are discussing the abstract ideas behind the process of generating U in a given setup. Is it clear? So, what we are discussing today is not electronics, it is the basic ideas in abstraction which go behind the process of getting to the U ok. Some of the basic ideas we will we will discuss and the point of this week's or this series of experiments is for you to explore some of these ideas and reinforce them you know. Hopefully, some more important questions will come through during the course of your experiments to those who are those who are interested, but the idea is to help you explore these ideas or the purpose is to explore these ideas and help you connect with these ideas a little better. So, notions like accuracy, precision, calibration etcetera that is the focus of the this week's experiments it is not the electronics. Of course, you will use electronics, but you have working level familiarity with the electronics or reasonable number of you have working level familiarity. So, we are going to take that as granted ok. So, before we get into the discussion for today a couple of points that I need to make this U of t which you saw is the which is the analog input we have assumed that the analog input comes from a measurement ok. So, sensor is providing you an analog input which you then later on process in a certain way ok. This U of t is not the measured value it is an it is a representation of the measured value that is that is an important point for you to understand ok. So, it is it is typically a voltage, but that voltage just by measuring some voltage as being 3.5 volt in your multimeter it does not tell you anything about what you are measuring ok. That voltage by itself means nothing, because suppose I put a black box or a screen in front of you and I say that this is the output of a sensor that is U of t you can do whatever processing with with it later on you know a lot of electronics you can do whatever you want with it later on, but I put the screen and I do not tell you where it comes from. All you can do is take multimeters and oscilloscopes and ammeters etcetera and find out what the value in the sense of the units you usually expose to volts, amperes etcetera of the variable that you are measuring is. So, you can measure U of t ok, but U of t itself is not the measured value it is the representation of the measured value is that is that clear to everybody ok. It is an this is an important point because most people confuse U of t itself as the measured value ok. The second point of course, I have already said that U of t is not the measured value and U of t is the is the representation ok. So, these two things you need to appreciate before we go on to the questions that we will we will address during the course of this lecture. So, any questions at this point what we mean by a representation of something U of t is like a proxy. So, if you are not attending some class you send some tell some fellow you will act as my proxy right that person is not the person who is did not come to class that person represents the person who who did not come to class ok. So, just by knowing who the proxy is I have no idea who the person he is representing is right just shout ah yes sir you do not know anything about who did not did not attend the same way U of t is a proxy U of t is a proxy for the measured value right ok. Any any questions about about this important idea ok. So, the questions for today these questions may seem simple, but they are not very simple and I can guarantee you most of you have to have to think before you answer some of these questions ok. The three questions we are going to focus these are the sort of broad questions there are a lot of sub sub questions that come up if you start discussing these in detail and the idea is to get you to think about these questions through the course of a discussion that we will have here take these ideas back and correlated with what you are going to end up doing during the course of this week ok. So, the first question is if U is only a representation right then given U how do we get to the measured value U m. So, you are you are interested in measuring something displacement velocity ah you are interested in temperature pressure lots of different things ok. Suppose I gave you only U how do you get to the measured value and I put a screen in front of you how do you get to the measured value you ultimately you are interested in the temperature being read as 83 degrees Fahrenheit ok or pressure being read as 0.7 bar or velocity being read as some centimeters per second or some displacement as being 30 degrees with respect to reference ok. So, those are the numbers you are interested in you are not interested in voltages because that is just a just a proxy ok. So, how do you get to the to these values that you are interested in ok that is that is one question that we will discuss in some some detail. The second question is even if you knew by some process what the measured value and I will call that measured value is U m is different from U U is a proxy for U m even if you knew the measured value U m how do you know that that measured value is equal to the true value because ultimately you are interested in the true value of whatever you are measuring ok. So, you do not know whether the measured value is the true value even if the. So, your representation to the measurement may be ok ok, but how do you know from the measurement to the true value you are fine that is another question. The third question is suppose you are interested in some variable and you have figured out a process of going from U to U m and you keep measuring the same quantity or lot of different people keep measuring the same thing and get the same or similar values what can you infer about the true value U of t or can you infer anything at all about the true value ok. So, these are questions that keep coming up in every measurement scenario that is why I have sort of picked them off sort of basic questions that keep coming up in every measurement process and they are very natural you do not need to you do not need to think too much about it the questions themselves are very natural the answer sometimes may be obvious sometimes may not be obvious, but the questions are very natural. So, we are going to discuss the questions clear what the questions are. So, let us take up the first question I am going to get some people to shoot their mouth off about the first question. So, we are discussing the question if U is only a representation how do you ever get to the value U m which is probably an intermediary between U and U t which is a true value. Now, let us say how do you get to the value U t ok if you are given only U what do you do? So, to the first question no not all questions only the first question. So, given U you can measure U m by comparing it by using some other standard value. So, like suppose you are having you are given some U some voltage for voltage representation for some measured value of U m which might be displacement. So, you might be having some other standard value of U m what do you mean you might be having the initial value like sort of thing. So, like for zero displacement you may assume U to be zero volts that is how you define it and go about this. No. So, so envision yourself sitting in front of a setup you do not know where the value is coming from and you measure some 3 volt what are you going to do with this 3 volt how will you tell me that this 3 volt corresponds to some temperature pressure or whatever whatever you are measuring how will you tell me that? I will measure at some other value say 5 volts or 2 volts. Ok, I measure 5 volt now what? I will take the change and I do not think you are understanding my question I am measuring 3 volt then I do another measurement I get 5 volt how do I say anything about what I am measuring some mathematical formula which would relate this to that is what I wanted to get to this has nothing to do with mathematics you can't just because you have measured something at 3 volt something at 5 volt you have no idea of what what you are measuring you understand my question has this happened to you that you measure something So, you need to have basically if we just keep on measuring these values like 3 volt 5 volt we the point I am trying to make is we need to have some measured value that is we just can't predict like you said you just So, what should you have done you are right of course you can't just predict from 3, 5 what what is it that you are trying to measure so what should you have done when I am taking down these 3, 5 and whatever values I must have some mechanism to you know get the values of what you are measuring you know some representative other values so you must have a mechanism so I give you a sheet I give you a sheet saying 3 volt means 30 degrees 4 volt means 40 degrees 5 volt means 50 degrees right so I gave you the sheet then you just look at the sheet and say okay 3 volt means this 4 volt means this 5 volt means this right so you you definitely need a sheet okay that's the first thing for you to realize the measurement itself has has no meaning 3 volt or 4 volt or 0.3 amp or whatever has no meaning you need a sheet that comes along with it who who generates that sheet it could be some standard reference somebody's already performed the experiment somebody already means as in like you have standards for who will generate the sheet if you are you are the person who's measuring sir we we would take a reference so reference can be something which we already know so we need to have some relation between the input and the output suppose we have you which is representing some variable suppose displacement or temperature so first of all we need to assume whether it is a linear relation or quadratic relation or something like that so that we can suppose I have a relation like u equals to ax plus b so when I get wait wait wait don't don't run here u equal to ax plus b what is x x is the real variable which we are trying to measure suppose it is a displacement so I'm trying to get a relation between that's what you asked I'm trying to get a relation between what I am observing at the oscilloscope and what is it in the real world right so how can you just construct something like why linear quadratic who told you you have to use these relationships we have to check which kind of relation is going to be most suitable for the given set of observations if we don't have that sheet which you are I think this is the reason I wanted to have the discussion I think the main question is not clear to you okay you can check whether a certain relationship whatever you may construct the relationship structure to be as being true or false or valid or invalid only if you know what is it that you're measuring if I gave you a sheet then you can try and fit ax plus b and ax square plus bx plus c or whatever it is to that and say that the relationship between what you're measuring and what you're observing has this bearing if I gave you the sheet I'm asking where did you get the sheet from where do you think that sheet comes from you need to have done something else you started off well but you went off somewhere else I think there should be another way of measuring it also I mean some physical way of measuring it so we know how I mean what by the way the answer is not complicated okay you just need to think like a layman and there has to be a scale relating you with um okay something else must have happened before I just want someone to just hit the nail on its head be done with it one more one more try there any thoughts on this we could have a standard reference or something we could take some some reference value like an assign say if u is equal to zero then um will be zero and then no who tells you all this you can you arbitrarily assume something like this you need to begin somewhere so that is how we could what do you mean you need to begin somewhere so you're measuring zero you need to get the relative idea about the measured values of you so okay that's some somebody here answer is quite simple I'm surprised nobody has come up if that is the first instrument which is measuring that physical quantity then there is no need to do anything further but if there was a previous instrument which measured that quantity then you just need to compare the values measured by both of them and calibrate the new device right so okay so that's that's probably the a nice answer what you need to do is that now don't everybody is going to be like what's the big deal about this the fact that you didn't know it okay or you could not articulate it properly the the point is that you need to know the value you expect for a certain measurement that you want to make okay so for example suppose I I want to get a value corresponding to I'm suppose I'm measuring mass okay it translates to measuring weight if I know that the value that I measure for me to know that the value that I measure the u corresponds to 100 gram weight I should know that a 100 gram weight a priori I should know before I perform the experiment that a 100 gram weight produced this 200 gram weight produce something else therefore if I have a value between u1 and u2 which is between the 100 and 200 grams okay and I assume the measurement process to be whatever you said some relationship linear quadratic blah blah blah those sort of relationships then I will be able to make something then I will be able to make a statement on what I'm measuring do you understand what I'm saying okay just by looking at some voltage value you are not going to be able to say anything about the stuff you're measuring unless you're given a sheet which we spoke about that this voltage corresponds to this this voltage corresponds to this where do you get that sheet from somebody should have used the same measurement process and should have known what the output of the measurement process needs to be for you to be able to make that sheet this where is the sheet going to come from the person who is measuring it should know a priori where the sheet came from so this this is an important point most people don't understand how many of you have not understood this point so far okay so many people have not understood it okay so suppose I'm measuring displacement okay I'm measuring some angle and I have a a measurement process which gives me u of t at the end of it okay and I measure one volt now I have to say how much I have rotated with respect to some reference that's what I have to say what I'm what I am saying is that just by looking at one volt you're going to get nowhere you have to accompany it with a sheet saying one volt means that you have rotated by 30 degrees two volt means you have rotated by 45 degrees okay who is going to prepare that sheet for you that is a question that we are trying to answer that sheet just by virtue of you not thinking about what is happening may just present itself as a data sheet and on your computer but we are not we are not talking about that we are talking about the process of getting to the sheet so the process of getting to the sheet is that you need to know that I am moving by 30 degrees first and then see what your output u of t is you need to know that I am moving by 45 degrees first then see your output what what your output is put all this in a sheet and give it to the person who's actually doing the measurement okay so you are basically you have you have measured a priori and the actual act of measurement is just correlating what you measure versus what comes on the sheet what is given to you in the sheet okay now without this process having a priori you can't make any sense of what you're measuring okay so now the question is how do you know that you rotated by 30 degrees in the first place when you're doing this process of arriving at the sheet you rotated by 30 degrees and measured one volt rotated something else by 45 degrees measured something else how do you know you actually rotated by 30 degrees okay so that's that's where something that everybody agrees on comes into place okay let's say I'm measuring displacements something that everybody's agreed on as what it means to write the symbol one meter has already been agreed on by the people by human beings have agreed upon what constitutes one meter and therefore subdivisions of it okay based on that agreement you're saying that I'm moving this by 2 mm and I'm going to measure what I'm moving by okay so if I get one volt then I'm going to measure by move by 4 mm and so I get something else so how much you have moved by or what exactly happened during the process during this process of generating of the sheet depends on an agreed consensus between you and other human beings okay there is no basis for that agreement you just agree that this means that I will put a symbol 1 mm that's the agreement is it clear okay so fairly insightful answer he said if this measurement of this variable or this quantity did not exist in the parlance of human beings then there is no point doing this exercise whatever you measure you that is it now you can agree that these sort of measurements measured in this way these sort of numbers mean something else so that agreement has to be reached reached first okay is this clear you have to give an input for which input as an input to the measurement process the process of measurement gives you a value okay you have to input to the measurement process which is the sensor the way of measurement how you are measuring it whether you are giving it to an ADC blah blah blah the entire measurement process an input for which you already know the output okay and then correlate the known output which is the measured value the voltage you are measuring okay so that process is called calibration so what you do in calibration is that you construct a table you give an input so known value of measured variable so this is then u of t you populate this table so this this table or the process of arriving at this table is called calibration it's a fairly straightforward idea if you think about it but not so straightforward to explain all the questions that go with it okay any questions about what calibration is without calibration there is no point to measurement that's that's what you need to understand right you know people go through entire engineering programs they look at some some number some some measurement is happening somewhere strain measure over there is a number on a screen they'll note down that number what does what does that number mean is never something that you pay attention to okay if you do not pay attention to it there is no point of the measurement okay how many of you have already done this you must have done this in your in your it days here gone to lab taken some numbers down how many of you have done this and not not ever questioned or not ever felt the need to question it one because of laziness two because of not in no interest third even if you are interested did not even strike you that somebody else must have done this before a b c how many of you choose a no interest no b is laziness okay c this is a good the honest people so you keep measuring lots of different things noting down lots of things so from now on if there is any prof who tells you okay you go do this experiment note down these these values these are the values for pressure this so you should what the question you should ask is show me the calibration sheet when was this calibrated how was it calibrated you will realize that it wasn't calibrated at all then you tell him I won't do the experiment unless you can give me the calibration sheet I won't do it okay so try this go and tell your prof okay the point is that you need to understand there is no point in doing a measurement unless there is calibration so what you will end up doing tomorrow is you will learn how to calibrate a load cell you will also know learn how to calibrate a pressure measurement device that we we have just concocted very simple to put together so I'll I'll show both of those things to you so you will calibrate a load cell you know load cell is go to this electronic weighing scales you must have seen in grocery shops the guy put something and some number comes there kilogram gram right how do you know it is right I can display anything I want right somebody must have said that this weighing scale if it shows these numbers it means that it is measuring this many grams somebody must have said this otherwise that guy is just cheating you don't need to even measure you can just keep displaying something so the person or the body that does it has sanctioned from set of human beings so government saying that you are allowed to tell whether this instrument with this measurement process if it yields these values then it you can certify that the number you're measuring is grams okay somebody has already done that otherwise it makes no sense so when we are calibrating these devices so in that case what shall we be using to calibrate them okay so the question so this is an important question and ultimately it boils boils down to what I briefly mentioned but I did not elaborate when you use something to calibrate an instrument that instrument or whatever you're using for calibration needs to be at least better than what what is it that you are trying to calibrate is that clear to you like I said if I need to know if I say that I move 30 degrees and the voltage value was one volt I need to know that I actually move 30 degrees and not 32 degrees is there otherwise my table will be all wrong so how do I know I moved by 30 degrees so there has to be a measurement there okay so the instrument that is used for calibrating another instrument needs to be at least as good otherwise it doesn't make any sense in fact it has to be one order of magnitude better typically as a ballpark so what about calibrating that instrument where how do you know that that is right so you can keep on asking this question and you know you have a infinite series going there what is the answer to that okay so the the short answer to that is at some point because you people you have to realize that there is no end to this at some point you will have to agree people have to agree on something arbitrarily with no basis okay and give a symbol saying that this measurement implies this symbol it's an arbitrary spec specification given by human beings you may not all be party to making the decision but in some sense people have agreed that these people will decide sort of thing okay so the answer to this question is an important question is that how do you know or what happens to the instrument that you're using to calibrate at at the end you have to agree on something if you do not agree then you can keep going going on and on it doesn't make any sense also because the numbers that you're coming up with 30 degrees 0.7 bar etc those things have to be agreed upon by people before you start using those is that is that clear it's an important question any other questions okay so we're going to move I'll come to what exactly you're going to do in your experiment in this time around okay but the idea of this discussion is to throw open these questions and get you to start thinking about them okay at the least whenever you see a number next time on some screen and somebody asks you to take it down and say okay this is the value of whatever you're measuring at the least you should know that this is bogus okay if you if you have not been told what the calibration sheet is it is bogus that whole measurement is bogus okay that's at the least is what you should go back with but if you know more then you can make use of whatever sure ask ask ask us for it and you'll know whether we are doing bogus bogus giri or not okay so that's that's about calibration so the next next question that I had thrown up is now suppose you have calibrated something okay so during the process of calibration what you're doing is let me just come to this calibration process again you're creating a table between u of t and u is it clear what what table you're creating it's a map between u of t and u so this u of t comes from some assumed value of u of t which you think is correct so called true value now whether that is correct or not depends on the calibration of that itself so that process itself okay so this is the process of calibration establishing this map so the question we're asking is this u unfortunately is not a representation of u of t it is a representation of u of m which is the measured value of whatever you're trying to measure okay is it clear u of t and u establishing that map is calibration but given a you you only know something about u of m which is the measured value which may be different from u of t okay so let me just highlight that so let's say let's say we are measuring temperature this is just an example so suppose I have some numbers like this okay so this is some volts this is some degrees Celsius suppose my calibration table came out to be this I stuck my sensor into a into a chamber which I knew a priory was at 30 degrees and that sensor put out a voltage value u which is one volt I do the same thing when it was at 40 and when it was at 50 I get 1.5 volt and 2 volt okay so now I have established a map between u of t and u which is the true value and u now this process is done so I don't have chambers anymore I'm just going to believe the measured value or the u and say something about the true value that is the actual process of measurement is it clear so this this table is given to you so the question is this u is a representation of the measured value not u of t so there is an abstract variable u of m so how do you know that what you're measuring that is the representation of the measured value u of m which is u and this table how do you know that this table is valid so for example when the calibration was happening if there was an error between u of t and u of m that error is going to carry all the way through you're going to be given a sheet everybody is going to keep making the same error all the time do you understand that do you understand the issue that that is that we are discussing so if you want to believe a sheet a calibration sheet that is given to you then it has to be or it better be for for you to make sense out of u better be that u of t is as close as possible to u m that is the measured value is as close as possible to the true value okay and the difference between this is usually called inaccuracy or also use the word accuracy is the difference between u of t and u of m okay this is a fundamental notion in measurement we'll discuss this in greater detail it's not not not an easy notion to to understand it's fine fine to define it like this but if i ask somebody to speak about it for two minutes you'll all be stuttering i know that okay because inherent to this is a notion of a true value what is the true value i mean that you can ask a philosophical question the point is that if your true value deviates from your measured value by a significant amount okay then and you're using that for calibration that process for calibration then the calibration table will have less and less meaning you will have less and less confidence in you saying that okay 1.25 means 35 degrees that statement you want to make ultimately okay is this clear the the issue is the issue clear okay so it's good if u of t is as close as possible to u m yeah so we will assume that the the calibrating device is true okay that's a basic assumption it may not be true but we are in the process of calibration we always assume the u of t that you're giving is the right value now if that is incorrect its calibration is error is also going to get added up is that clear to everybody so this this table u of t you will assume as being true only this is variable so the notion of accuracy okay this it's not always called inaccuracy but it's called accuracy is all also this this difference difference between the true value and the measured value okay you want that to be small now when can what is a typical sort of situation where where you have what is the typical situation where you have bad accuracy give me an example of a situation where you have bad accuracy okay so anyway we'll we'll come to that in a little while okay so i'm going to give you an example of a situation so you all seen a dart board dart board darts even if you have not used darts you must have crumpled pieces of paper try to hit some target at some point so the dart board in which you throw darts and this is your target okay so i'm going to draw draw three pictures this is your target i throw five darts each dot hitting i'm going to use a into mark okay and we are going to discuss the difference between situation a and situation b what is the difference what is the difference between situation a and situation b both are accurate both are precise one is accurate okay so the point is that this collection of darts okay seems to indicate that your process of throwing darts is fairly repeatable you keep doing the same thing but you do the wrong thing what why why will you do this suppose you had a squint or you can't see the target properly you are thinking that i am hitting the target all the time or sometimes you know that you're not hitting the target but you'll keep doing the same thing okay this is that situation this is a situation where you hit the target and you're pretty repeatable so this is a good shooter okay this is a situation where you're a bad shooter you're all over the place so c is bad of course a is good of course b it seems that you have some hope you just need to either wear specs or need to change the angle in which you're throwing and keep repeating that so i'm going to give some names to this so in this case ut is my desired location okay and the difference between i can talk about a euclidean distance between my target and what i actually achieve as the inaccuracy okay and i can i can come up with the lee i can come up with the sum of square sort of euclidean distance which represents the inaccuracy of all the five throws okay take the euclidean distance add all of them up for this you will have the smallest number for this you will have a larger number and for this maybe you have a larger number or maybe you have a smaller number the point is if the inaccuracy is small then you call the measurement accurate obviously so this is accurate not accurate again not accurate so we are going to introduce another notion called precision the precision captures this notion b that you're not talking about the true value at all you're talking about the difference between two measurements so um1 and um2 okay accuracy talks about difference between the true value and the measured value precision talks about um1 and um2 it has nothing to do with the true value okay so you call imprecision okay so this is precise this is also precise okay in some sense this is not precise so two things you need to remember precision has nothing to do with accuracy in the sense of a measurement process okay accuracy has nothing to do with precision either what about the second statement is that true how many of you say accuracy has nothing to do with precision if i'm precise i may or may not be accurate but if i'm accurate i'm precise okay so let for those of you who are confused i'll confuse you even more okay so i'm going to give you this this is one situation second situation is all of them are here which is more accurate which is more precise yeah so it depends on the relative distance of these dots away from the crosses of course but the point is you can be accurate and not precise that is if you keep repeating the same thing you'll approximately get to the right value but you'll you may be off in a different direction you would throw something and you kind of over correct and you throw something else kind of over correct again you throw something else so you're reasonably accurate but not precise you can keep throwing the same place then you're precise so these two are independent notions accuracy is not related to precision precision is not related to accuracy okay yeah so we'll we'll discuss about the infinite infinities soon okay so tomorrow you're going to do something and hopefully this accuracy precision etc will will come into play you have to you have to think about it so this is a load cell okay so this this guy is similar to what the grocer uses this is similar to what the grocer uses you put a weight there then you connect up your electronics in a certain way you will be told about the wheat stones bridge etc which you may have been exposed to already you must have definitely done problems 3j e on wheat stones bazaar 1 by r2 equal to r3 by r4 all that okay so this is this is what you will use for calibrating this load cell you will keep putting different weights weights will be given to you you can ask the question where do you get these weights from I will say go and blame the standards organization Indian standards organization okay so these weights will be given you need to put them here measure the value of you and create the calibration table you will assume these to be the ut either ut good question no answer okay Indian standards organization we will tell you that that's for this load cell okay so you will do a more interesting experiment so you will be exposed to this device okay so this is a very simple device on this device so it's just a piece of metal okay which is open to the atmosphere right now but if I close this whatever pressure is maintained inside is maintained okay so this this device is going to be used for measurement of pressure okay let's say this device is going to be used for measurement of pressure obviously I can make this guy standard different points in time at different different heights okay if I take this off it drops the physics of this should be fairly straightforward for you to for you to understand so I'm going to ask somebody who has done lots of physics let's see if you can apply it to this simple situation tell me the physics of this why does this why are you able to see something here I'm able to stop this at different different heights so I mean whenever you put a weight on it there'll be that rod right so that rod will stop the vertical thingy falling through why because there's a hole all the way through and you're putting it putting a pin through it no you're not asking answering my question I am able to make this guy stop at different different heights for example I've made him stop here I made a stop here oh the pin doesn't go all the way through yeah then it's just friction friction I mean you're just how many how many of you say it's friction or there might be holes inside that thing where you're putting the there is no there might be an all I don't know what's inside so I can speculate zero you're going to venture to talk about physics of this simple device by the way it's not friction anybody else willing to venture into explaining the physics of this okay tell me this so look at this condition in which this has rested somewhere okay so this this guy is it's just a rod on one side it sees some pressure on the other side it sees some other pressure what can you tell me the pressure inside vis-a-vis the pressure outside yeah so that itself is the physics the pressure inside has to be higher in the pressure outside because if you do do a simple force balance of this guy okay on this side you have pressure acting through this area and then the weight and that is contracted by the pressure here so that's the physics sorry I took a fake too okay so as simple as that so what you're going to do now is you are going to use this as a pressure measurement device and the way you're going to use that is you utilize weights to first calibrate this device okay so what do you mean by that so this is connected to this piece of electronics tiny piece of electronics which is going to generate some voltages here okay and these voltages will be different for different weights put in here okay so again you establish a calibration table and then say if I measure this voltage I can correlate with with the pressure that must be inside this because if I believe my force balance the pressure should be that yeah so you will only get p into a you're right you're right okay so good good question but if I gave you the area of the piston then you'll be able to get to the pressure okay so this is going to be a pressure measurement device for you this will give you load this is some weight weight measurement measurement device we'll come to that in the next next time around but you will be doing calibration of two instruments for sure one is this guy the other is your load cell okay so you have to create the table okay so this each each group will be doing this each of you will create this table okay and we are going to ask you to measure some unknown quantity so-called unknown quantity from here which is not used in the calibration table so each group is going to measure something okay we are going to use that data so the data of each of the groups to make some statements about precision okay you have lots of um1 um2 um3 etc we are going to utilize that information to find out if our measurement process which hopefully everybody will replicate but the point is that people will not replicate it people will keep doing their own thing you will all get different um1 um2 um3 um4 values if you do not get different um1 um3 um3 um4 values you're all lying together okay so they will make a statement about precision of the entire measurement process how you keep this how that thing is plugged in is there a hole here is there a hole there whether whether this is repeatable etc the entire process leads to your precision or imprecision so we will we will make a statement about precision okay so each I think each lab has 40-50 people right or 20-30 people so we will make a statement about precision from the data of each each person all collected together okay now the purpose of that exercise is not to say that this is precise or not that is not the purpose the purpose is to indicate to you that this question should come at every point in time when you're using using a measuring equipment because the what is its accuracy what is its precision whether it has been calibrated or not etc these are basic questions that you need to ask before utilizing a measurement for the purposes of verifying some hypothesis okay so the the process of asking those questions or why those questions make sense is what we are trying to highlight it's not the specific answer of whether this is precise or not that's immaterial because you're probably never going to use this instrument to measure pressure is that clear what you're going to end up doing okay any questions about about what we are overall trying the overall intent of the of the discussion is to let you explore notions of precision accuracy okay and the get exposed to the idea of calibration and do calibration a couple of times okay there are lots of questions that one can ask related to this simple experiment okay we will ask some of them to you during your during an experiment itself and some in your exam of course so if you don't understand what you're doing from a from a conceptual understanding point of view you will you will be under a handicap okay so pay attention to what is it that you're doing because the questions are not very easy about this they're confusing questions they're not if you're not clear about what you're doing you can get easily confused okay any any overall questions on what the learning objective is learning objective is to get exposed to ideas of accuracy precision calibration this is these are excuses for you to get exposed to that idea yeah if accuracy is zero or inaccuracy is zero then you can say precision is zero or imprecision is zero that is correct nowhere else you can say anything accuracy is zero okay so you keep measuring and you keep meeting the right target all the time then obviously precision is imprecision is also zero so but you said that they don't have any relation sure in this particular so you can in the as in or in the extreme cases there are relationships okay but in general if for example if inaccuracy is some number i'm just giving you a number for some measurement device is 0.1 mm right right no the question you are asking is is is there a condition where you can say something about precision by knowing only accuracy and my answer is yes you can if the accuracy is zero or inaccuracy is zero so the definition for inaccuracy is this ut minus um yeah this is inaccuracy but it is also used somewhat incorrectly as the term accuracy itself okay that's why i put it in brackets so when you see accuracy is so much it actually means inaccuracy okay so if the inaccuracy is zero which means that every time you measure you hit the true value then and the true value remains constant and the precision is also zero or imprecision is zero okay see if you're you understand the spirit of what i'm saying the spirit of what i'm saying is that just because you knew inaccuracy you cannot say anything about imprecision but yes in one particular case if inaccuracy is zero then you can also say imprecision is zero that doesn't mean that the spirit of the previous statement is wrong okay so if you're trying to do a one-upmanship saying i found something where you are wrong sure great but if you want to understand the spirit of the argument any other questions okay so happy labbing