What is the purpose of this? What measurements are the different colors belonging to?
There are 5 ADuC7061. In industrial environments with ambient temperature ranges from −40°C to +40°C is necessary to use another ADC.
I think this is not a good method to predict the quality of the ADC.
Also it is essential how you have connected the circuitry to the part!
1. the temperature drift is mainly influenced by the reference used - internal Vref, external Vref - external voltage source or reference resistor
2. on-chip reference is worst case 20 ppm/°C = 24µV/°C - to improve this you can use a external better reference
3. unless you use specific ultra low TC, accurate, but very expensive resistors as reference you see also the drift from that
For normal application you need to use calibration in software to compensate for drift.
So in the first step I suggest you use a external voltage from an calibrator as a source for the ADC input and repeat the measurements to see the accuracy of the part.
If this is not accurate enough, you can use a better external reference.
Then you can do tests with a resistor to predict the performance of your resistor and circuitry.
But please also read this http://www.analog.com/static/imported-files/rarely_asked_questions/moreInfo_raq_resistors.html
I agree with everything. ADuC7061 ADC is invalid without additional calibration in the temperature range -40°C - +40°C.
Can you provide data you have captured as I explained as a first step?
I used external reference resistor 1 kOm 5 ppm/°C at REF0-REF1
Doesn't mean anything, can you provide the schematics and the software you used to do this test?
There are multiple possible sources which can cause this.
What board did you use?
What type of filter cap's did you use?
How did you connect the sources to the part/board....?
Can you provide the theoretical calculations showing with all possible factors the worst case scenario?
Thanks. I don't have any questions. I just shared information.
Others in the community would appreciate to know the details how this data is generated - so can you give some more detail?
OK. The diagram on page 55 Figure 20. Example of an RTD Interface Circuit without PGA, internal and external filters and op-amp. Line connect RTD=100 Om (10 ppm/°C) is a 2 twisted pair (50sm). RTD temperature equal 24°C. ADC0 is calibrated at 24°C too. Range 80 Om - 120 Om. Iexc=1ma Rref=1kOm (5 ppm/°C). Internal temperature sensor at ADC1. I did test ADuC from -40°C to +50°C.Thats all.
How did you configure the ADC - in ratiometric mode?
Ratiometric mode is normally considered to be when the device output is a ratio of the power supply. Is it right?
I used external reference resistor as Vref, i.e. I meassured RTD/Rref. Is it ratiometric mode? RTD/Rref does not depend on the power supply.
So if this is really done correctly - using Rref as reference and used a normal 100 ohm, than the only conclusion here can be that one of the used resistors TC is worse than specified, because the ADC itself shouldn't influence the result in a ratiometric setup as you confirmed. So it would be of interest if you can provide 2 sets of date where you keep each of both resistors at a constant temp to see the influence from the other one.
Also a 3rd data set would be of interest where you keep both resistors at constant temperature and just cycle the part through the temperature range.
And I'm also interested about the SW, how you have configured the ADC in detail - MMR settings.
Each of both resistors at a constant temp 23°C.
POWKEY1 = 0x1;
POWCON0 = 0x7B; // Set core to CPU speed of 1.28Mhz
POWKEY2 = 0xF4;
POWKEY3 = 0x76;
POWCON1 = 0x20; // Uart enabled, PWM & I2C disabled
POWKEY4 = 0xB1;
IEXCON = BIT1 + BIT2 + BIT6 + BIT7; // 600uA[1:3], Enable
ADCMSKI = BIT0 + BIT1;
ADCFLT = 0;
ADC0CON = BIT4 + BIT10 + BIT15; // Gain = 32[0:3], Ext.ref[4:5], Unipolar, Ext.ref , enable 
// [10:7=1011] = internal temperature sensor at ADC1
ADC1CON = BIT7 + BIT8 + BIT10 + BIT11 + BIT15; // Gain[0:1], reference[4:6], ADC2-3[7:10], Unipolar, Int Reference, enable ADC0
ADCMDE = BIT0; // Enable Continuous conversion mode & 133 khz
ADC temperature drift calibration
Retrieving data ...