Halp with computer architecture

[Fallso]

RIGHT.

Basically I've got this assignment to do on computer architecture, but our teacher is so incompetant that nobody in the class has any idea wtf to do, would anybody be nice enough to add me on msn or skype and lend a hand for 20-30 mins? ;_;

[Blubbartron]

I'm in a computer architecture class right now, no access to either of those atm as I'm on campus. Do feel free to ask questions in the thread or PM though.

[The Stig]

Extremely vague, any particular area?

[Fallso]

Sorry ;;

It's easier to explain if I just copy paste the whole fucking thing in lol

Undowtyble Electronics are developing a new temperature sensor for use in Turbo-Prop aircraft. The sensor will have a variety of uses including:

• outside air temperature
• cabin air temperature
• pitot tube temperature
• carburettor air input temperature.

The sensor must measure temperatures from minus 9.9 to plus 49.9 degrees Celsius to an accuracy of 0.1 degrees (one tenth of a degree).

A Betatherm Ni1000SOT thermistor will be used as the actual sensor. This is relatively linear over the required range at 5.42ohms per degree. An operational amplifier will change this to a voltage of 14.19647927mV per degree. In other words, -10˚C will produce zero volts and +50˚C will produce five volts.

This voltage is fed into the analogue input of a Microchip PIC17F873 as this computer chip includes a 10 bit analogue to digital convertor, which will be configured to read from zero to five volts. The PIC16F873 will convert the analogue voltage to a digital signal and send it to the main aircraft computer over a USB interface. USB is achieved by taking the serial data (UART) output from the PIC16F873 and feeding this through an FTDI UM232 convertor.

The calculations we have to do are deriving values from the analogue to digital converter when the temperature is at X degrees.

idk, it doesn't make sense to me.

[The Stig]

Holy shit I knew my degree was useless.

I do feel slightly uneasy reading that.

[Fallso]

What kind of level is that stuff? We've never done anything like this before and it baffles me.

[The Stig]

Well there's no real question from what I can see. Are you supposed to ... design a device that does what it says? Or work out some complex equation/code that does what it says?

Difficulty? Not sure. What's your course name? And god damn which uni are you placed in. If it's Cambridge then hats off to you . . .

[Fallso]

This is college lol, going to uni next year which is why I need to get this ><

The idea is that we do some calculations, one of them is

• Calculate the value that will be derived by the analogue to digital convertor when the temperature is at the freezing point of water (zero degrees Celsius) and express this:
  
  • as a ten bit binary number
  • as a four digit hexadecimal number

c. as a four digit decimal number

[Blubbartron]

From what it looks like, you're expressing temp * 10 (to account for accuracy in tenths of a degree) as a binary number, with a shift of +100, because the temperate is measured from -10 to 50. Thus -10 would be 0000000000. 0 degrees C should be +100 to that previous binary number, or +64 + 32 + 4. This becomes 0001100100.

edit: I assume you're capable of the basic conversions from that point.

Further Edit: To clarify, this converter chip is attempting to express each temperature as a unique series of 10 bits. It is 10 bits because it is attempting to uniquely identify 600 items. It is 600 items because we have a range of 60 degrees, 10 sub-units per degree. If we had 9 bits, we could only identify 512 items. This is of course assuming that the output isn't meant to be in Two's Complement binary expression. So far your directions have been fairly vague, and if I were in that class, I would punch the professor in the face immediately after reading the question.

[Fallso]

From what it looks like, you're expressing temp * 10 (to account for accuracy in tenths of a degree) as a binary number, with a shift of +100, because the temperate is measured from -10 to 50. Thus -10 would be 0000000000. 0 degrees C should be +100 to that previous binary number, or +64 + 32 + 4. This becomes 0001100100.

edit: I assume you're capable of the basic conversions from that point.

Further Edit: To clarify, this converter chip is attempting to express each temperature as a unique series of 10 bits. It is 10 bits because it is attempting to uniquely identify 600 items. It is 600 items because we have a range of 60 degrees, 10 sub-units per degree. If we had 9 bits, we could only identify 512 items. This is of course assuming that the output isn't meant to be in Two's Complement binary expression. So far your directions have been fairly vague, and if I were in that class, I would punch the professor in the face immediately after reading the question.

Thanks a ton lol, that looks good atm (and thanks for the explaination, I actually understood that!).

Did a few calculations and it looks as if your method is coming out with results that look right, so I'm gonna go with that for figuring out the rest.