Post by thread:[OT]: thermistors in op-amp feedback

Hi Group, I have an op-amp circuit operating in the non-inverting mode with 36x amplification. The Rf is 100k and the Rv is appr 3K. I need to circuit to back down the gain at cold temperatures. At the low end of 0 deg, the gain would be 18 and at the high end 120 deg, the gain would be 36x, as a function of temperature. So we would start with Rv near 3k and end with Rv near 6k. Is there a better method than using a thermistor for Rv? Could a transistor or diode, FET or other be used to control the gain? I could use a digital POT but that may be more expensive than desired. Thanks, Paul

PDRUNEN@AOL.COM wrote: > > Hi Group, > > I have an op-amp circuit operating in the non-inverting mode with 36x > amplification. The Rf is 100k and the Rv is appr 3K. I need to circuit to > back down the gain at cold temperatures. > > At the low end of 0 deg, the gain would be 18 and at the high end 120 deg, > the gain would be 36x, as a function of temperature. So we would start with > Rv near 3k and end with Rv near 6k. > > Is there a better method than using a thermistor for Rv? Could a transistor > or diode, FET or other be used to control the gain? I could use a digital > POT but that may be more expensive than desired. > > Thanks, > > Paul have you worked out the conversions yet to get a readout in F or C.. PITA.. depending on your temperature range you might check into Dallas's digital themp sensors.. of course they are like $6-8 in single quantity but need no extra parts. Damon Hopkins

I too am curious as to how thermistors and the like are used to perform temperature compensation. I believe you have ventured into another one of those black arts. Why not use the PIC to do the work, assuming there is one? If 8 bit resolution is not enough, then use a 12 bit A/D. If you want this gain to adjust linearly over temp, you will not likely get that with a thermistor. Your challenge is not just at two temperatures, but at all temperatures. Anyone out there know of a good book reference on the subject?? Chris Eddy spam_OUTPDRUNENTakeThisOuTAOL.COM wrote: {Quote hidden}> Hi Group, > > I have an op-amp circuit operating in the non-inverting mode with 36x > amplification. The Rf is 100k and the Rv is appr 3K. I need to circuit to > back down the gain at cold temperatures. > > At the low end of 0 deg, the gain would be 18 and at the high end 120 deg, > the gain would be 36x, as a function of temperature. So we would start with > Rv near 3k and end with Rv near 6k. > > Is there a better method than using a thermistor for Rv? Could a transistor > or diode, FET or other be used to control the gain? I could use a digital > POT but that may be more expensive than desired. > > Thanks, > > Paul

Hi, thermistor tempco for opamps usually only works in some sort of closed loop or after shunting the thermistor copiously so its non-linearity becomes less important. A simpler and more linear method relies on a silicon junction (diode) designed into the feedback or reference side of a DC coupled opamp such that the desired temperature compensation is achieved. A typical example places the D between +in and GND (with some more R's etc) in an inverting opamp configuration. If the input produces positive output then this scheme produces gain decrease with temperature. The zero point moves too however. A cheap RTD can be used in the feedback loop to compensate linearily for gain over temperature. The thermistor is actually only used as the RTD's poor brother. And this is not a black art. The black art is to avoid needing a tempco on gain by using tempco circuits everywhere imho. Matching the required curves (of the required tempco and of the thermistor(s)) requires tens or hundreds of lab hours and a heat box. Reproductibility is not guaranteed. Peter 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