Compare the voltage against a precision voltage ref, available at precision of 1%. I use loads of ZRB500F01s, cost < #0.40, Temp coeff 50ppm/degC. Stuart ----- Original Message ----- From: "Donovan Parks" To: Sent: 24 March 2002 05:45 Subject: [EE]: AD592 Temperature Sensor (coverting relative voltage to abs. voltage) Hello, I'm hoping to use an AD592 temperature sensor to *accurately* measure temperature (+/- 1 degree C). The device outputs 1uA/K and has a rated accuracy of (+/-0.5 degree C). Now, to measure the current produced by the AD592 I am measuring the voltage over a known resistance value (~10K) using a PIC. Unfortunately, I didn't account for the fact that the PIC AtoD really gives you a relative voltage (i.e. what percentage is the measured voltage compared to Vref+). I'm using a LM2940 to achieve +5V from a 9V source, but the LM2940 actually outputs 5.4V (beyond the maximum value specified in the datasheet) and will decrease as the 9V battery voltage decreases. Thus, I am left with no accurate way to measure the absolute value of the voltage across the resistor connected to the AD592. Note, this is for a one-off design so I am more than happy to do any tricks. For example, the resistor I am using is just a 10K resistor I grabbed out of my junk box and used a DMM to get a precise value that I use for calculations on the PIC. If there a similar type 'trick' I can use to determine absolute voltage. I'm thinking something along the lines of using a diode to get an absolute voltage value (0.7V) so I can determine the voltage across the resistor (of course, a diode is not precisely 0.7V and changes with temperature so this probably isn't a good approach). Any suggestions? Regards, Donovan Parks -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics