Why does a resistor and diode make this circuit temperature insensitive

tindel

Joined Sep 16, 2012
936
upload_2015-12-14_21-39-50.png

Here is a temperature model of a similar part. note the straightness of the curve that my crosshairs are on in the plot. I calculate that slope at ~8ppm. And we haven't even taken into account the tempco of the resistors! So double this value (swag) to account for resistors and the diode variance, making it somewhere on the order of 16ppm, worst case. So that makes your tempco across the entire range of the part equal to about 16ppm/degC + 10ppm /degC = 26ppm/degC *65degC = 1690ppm*5uV = +/-8.45mV The only thing you've gained is an initial trim accuracy. maybe you want that, but it doesn't buy you much over the total temperature range of the part.

(Note: the simulation had a tempco at 0ppm without the trim circuitry, so we know that the tempco is based on just the diode and the bias current of the diode.)
 
Top