In the datasheet of the AD9235 there is a specification for gain drift when using an external reference.
is there any data available regarding typical gain drift when using the internal reference?
I'm reading figure 41 as 1.1% to 0.8% error from 1V. So I calculate about 80 ppm/C?? Still not great.
I too would recommend using much newer parts on newer processes, better testing, more chance of
being around in ten years, etc.
Figure 41 shows an example of a typical devices internal voltage reference drift which shows it drifting from approximately 1.1 to 0.8 V (i.e. delta of 300 mV or 300,000 ppm)............NOT GOOD. Assuming linear drift, this would amount to 2400 ppm/C..................which is 200x greater than the typ Gain Drift of +/-12 ppm on page 3 of datasheet. For this reason, it would appear that the datasheet encourages an external VREF if absolute accuracy is requierd.
Note, this device is over 10 years old and has been replaced by newer devices that may have tighter gain drift specs using internal VREF. Take a look at the recently released AD9237 ADC family which shows that its internal VREF shows far less temp drift (figure 41) than the older AD9235.
Thanks for catching the error. Anyways, looks like both ADC are about equal on the internal VREF drift (i.e. 0.3%) but AD9237 would still be preferred part for reasons stated by Harry.
Thanks for the information,
Obviously I overlooked this figure.
Upgrading to AD9237 is however no option to me due to nontechnical reasons.
However an external Reference seems useful.
Retrieving data ...