Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
As it is, you are using 0 - 5v range which equals (5/0.01) a 500°C range (which the sensor isn't even capable of).
You need to narrow your range either by lowering your A/D Vref or amplifying the temperature sensor output by using an op-amp to get the resolution you require.
For example you can have your -Vref set to 2.2195V and your +Vref set to 3.2435V to give you a range of -51.2°C to +51.2°C and a resolution of (102.4/1024) 0.1°C.
Also, if you do it that way then each LSB equals 0.1°C making your programming a lot easier.
I would cange my reference voltages but the 16F88 controller
uses RA2 and RA3 for that and those ports are already
committed to another function on my board.
I would cange my reference voltages but the 16F88 controller
uses RA2 and RA3 for that and those ports are already
committed to another function on my board.
hi,
This in not correct, the accuracy is ~+/-0.25C.
The resolution can be whatever you chose, once set up the 'repeatability' is good, so if you are using the LM35 for measuring temperature 'change' , it works OK.
The response time for change in temperature is slow, check the datasheet.
If you use the LM35 and can determine the 'error' on the accuracy, it fairly easy in the program to adjust for this error.
These comments are based on my experience using the LM35 over many years.
I agree the 'overall' resolution of the LM35 and the PIC at 10bits is 0.488C.
But the way I read your post suggested that the 'best' LM35 resolution is only 0.5C.
This is always a problem with short text posts, they can be read in different ways.
He didn't say he was forced to use the 5v as Vref? Just that he didn't want to use opamps?
It's pretty easy to use something like a 1.8v blue LED and a resistor to be a 1.8v Vref and use that for the ADC unit. The datasheets usually say 2v is the minimum that Vref should be used, but this is to ensure full linearity at full range, especially the top of ADC range. For an LM32 that puts out 0.020v to (say?) 0.800v a 1.8v Vref will work fine and be linear enough.
...I would cange my reference voltages but the 16F88 controller
uses RA2 and RA3 for that and those ports are already
committed to another function on my board...
He didn't say he was forced to use the 5v as Vref? Just that he didn't want to use opamps?
It's pretty easy to use something like a 1.8v blue LED and a resistor to be a 1.8v Vref and use that for the ADC unit. The datasheets usually say 2v is the minimum that Vref should be used, but this is to ensure full linearity at full range, especially the top of ADC range. For an LM32 that puts out 0.020v to (say?) 0.800v a 1.8v Vref will work fine and be linear enough.
LEDs are not too bad as a Vref reference. The light sensitivity is mainly when they are reverse biased, but forward baised they are fairly voltage stable (better than most Vdd) and make less noise than a zener (which is reverse biased of course).
But I should apologise as I missed the point about the PIC pins being already allocated. I'll shut up now.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
