5 Replies Latest reply on Feb 20, 2015 8:33 AM by user_14586677

    IDAC : Code 0xFF=31.875 or 32 uA ?

    user_347933176

      I setup the IDAC to code FF in the lowest range. I dont need to change the current but I need to know as accurate as possible how much the current is.

         

      The documentation says this is 31.875 but when I measure it with my precision voltage meter over a fixed resistor I get :

         

      >> Comparing to 31.875 uA >> deviation = 0.4%

         

      >> Comparing to 32uA >> deviation <0.1%

         

      Tested it for as well as Sink and Source mode. Fixed resistor 100 and  220 ohm.

         

       

         

      So my question : Is it correct that the IDAC is calibrated to 32 uA and NOT to 31.875 ?

         

      Patrick

        • 1. Re: IDAC : Code 0xFF=31.875 or 32 uA ?
          user_1377889

          In the datasheet the data are specified for a 600 Ohms resistor, can it be that the low resistance is giving you bad results?

             

           

             

          Bob

          • 2. Re: IDAC : Code 0xFF=31.875 or 32 uA ?
            user_347933176

             Thanks for the answer,

               

            No I just mentioned the 2 resistance values. I measured with E12 from 100 to 100k, knowing that on the high end the voltage over the resistance will become too high.

               

            See attached picture -

               

            X-Axis=resistance 100-120-150...82k-100k (logaritmic scale)

               

            D1a=Source; deviation in % from 31.875uA

               

            D1b= Source; deviation in % from 32uA

               

            D2a=Sink; deviation in % from 31.875uA

               

            D2b= Sink; deviation in % from 32uA

               

            It looks like D1a and D2a goes to 0.4% at origin (255/256=0.4% !!!)

               

            It looks like D1b and D2b goed to 0% at origin

               

            For calculating these values, I measured the resistance of the resistors used also with the HP precision meter.

               
                    
            • 3. Re: IDAC : Code 0xFF=31.875 or 32 uA ?
              user_14586677

              If you do an error budget analysis the results you are seeing well

                 

              within spec -

                 

               

                 

               

                 

               

                 

              The voltage ref specs are -

                 

               

                 

               

                 

               

                 

              So you could use a precision R, the DelSig A/D, and possibly have a built in self cal

                 

              correction to improve the overall specs for the IDAC. However you are intrinsically

                 

              limited to 8 bits, ~ +/- .4%, so not sure even this approach would be useful.

                 

               

                 

              You could use PWM, OpAmp, Bipolar PNP and develop a more accurate I source,

                 

              but this would require a precision ref driving Vdd.

                 

               

                 

              Regards, Dana.

              1 of 1 people found this helpful
              • 4. Re: IDAC : Code 0xFF=31.875 or 32 uA ?
                user_347933176

                Thanks for answering. 

                   

                I know it is within the range of the datasheet. I am just trying to get a current source which is as accurate as possible. The 32uA / 0.1% which I have now suits me very well. When using the IDAC on less then Full Scale, the output gets more inaccurate (but still within datasheet mentoined limits)

                   

                While working on my first PSOC project, I am positively surprised every day about the possibilities of these devices (PSOC3/5LP) and the flexibility and accuracy I can get. :-)

                • 5. Re: IDAC : Code 0xFF=31.875 or 32 uA ?
                  user_14586677

                  This is off topic a little but take a look at this technique for removing

                     

                  offsets and noise in many applications. Correlated Double Sampling.

                     

                   

                     

                      

                     

                            http://www.cypress.com/?rID=49159

                     

                   

                     

                  Regards, Dana.