$(8)$ An engineering student, Mr I. B$.$ Nofool was employed over the summer vacation to set up an electronic sampling and processing system. He was told to set up a system that will monitor the force on a metal bar that supports a heavy sensor$.$ For various reasons, the force on this metal bar can vary, and a sensing system is required to log the force variations. He went about this task as follows: $($i$)$ Assuming that the operators can not observe a change that occurs faster than $20$ times a second, he created the system to sample at $100$ Hz$,$ far more than what he considers to be twice the required sampling rate. $($ii$)$ Knowing that one of the easiest ways to measure force on a bar is with a strain gauge, he set two of them up$,$ both along the longitudinal axis, so that via a push$-$pull setup he could compensate for the non$-$linearities. $($iii$)$ Knowing that there could be temperature fluctuations in the environment, he chooses operational amplifiers that have small temperature drift. He also places a fan in the vicinity of the sensor to try to keep the temperature constant. $($iv$)$ In the control software he monitors the results from the strain gauges, and has the computer notify the operator that an error has occurred if the output from the sensor is outside some bracket of "reasonable" results. $($v$)$ In order to minimise costs he designs the system to work with a Successive Approximation Analogue to Digital Converter, uses standard ribbon cabling $($for parallel data transfer$),$ and rather than designing a totally separate power supply for his data presentation, he implements a standard VDU screen, powered by mains supply $(240$ V$,50$ Hz$),$ which he positions close to the sensor to minimise cabling length, and hence reduce the amount of noise on the cabling. The ADC provides for a resolution of $0\mathrm{.}01$V$,$ which is within the specifications of required system accuracy. $($vi$)$ Rather than worry about logic level interfacing, he arranges the op$-$amps from the strain gauge to send out a signal between $0$ and $5$ volts, and designs all subsequent logic to operate at this TTL logic level. $($vii$)$ Being aware of the fact that fatal errors can occur, he sets up a monitor routine that periodically polls the status of the various elements of the system, warns the operator of a partial failure of the system $($not just that the sensor is producing strange results$),$ and will shut the system down if the fault is serious enough. $($viii$)$ Mr Nofool does not worry about documentation because he thinks he can remember everything he has done, and thinks that he can be assured of a permanent job if he keeps the design reasonably secret. To Mr Nofool's way of thinking if there is no documentation, they must employ him to do any maintenance or repairs. Now Mr Nofool has not created an optimal solution. Can you suggest any improvements to his system, before he implements it $($and subsequently gets fired!$)?$ Note that not everything he has done is wrong! State the reasons behind any suggested modifications. Your answer does NOT need to be in essay form. A list of the appropriate points, with an explanation is satisfactory, $($and preferred as it is considerably easier to mark!$).[10$ marks$]$ Marking note: In the past, the answers to this question have indicated a reasonable amount of paranoia by the author. To combat this, I will SUBTRACT marks for suggestions that are overly pedantic, or do very little $($if anything$)$ to improve the system. There are some very obvious errors in the above approach, though some things are correct $($at least to the extent that they will work satisfactorily. Only correct those things that are decidedly wrong.