2016-02-09 04:54 PM
Hi!
I have been using the F373 and its SDADC successfully for a while now but in my most recent project I'd like to do a precise gain calibration without the use of additional external components. I'm using a 3V external reference voltage connected to VREFSD+. In order to do a gain calibration, I now need to connect another known voltage < 3V to one of the analogue inputs of the same SDADC to measure the gain error. I don't want to buy another reference IC ony for that. So my thought was if it might be possible to connect the unused internal 1.2V bandgap reference signal to one of the SDADC inputs? Basically exactly like it's done for the SAR ADC (Vrefint connected to Analog MUX): In my case, it would look something like this: So by measuring the internal, calibrated 1.2V bandgap voltage with the precise 3V external reference, I can calculate the exact gain and compensate any future measurements. Unfortunately, it looks like the bandgap voltage is not accessible for the SDADC to measure, nor is it connected to any other pin. Have I overlooked something? There must be a way to do proper gain calibration without using additional external parts? If not, I'll use a precise resistor divider fed from 3V. But then I'm relying on their accuracy ... I'd rather like to use the calibrated internal reference. Thanks! Edit: Or is there a way to use V_refint with the SAR adc so that I get the 1.2V at the VREF+ pin to route it into the SDADC? I couldn't find any code in any examples showing that the SAR adc actually is able to use Vrefint as reference ...2016-02-17 05:00 PM
*bump*
No one any idea? Impossible? Would be so nice to have!2016-02-21 11:08 PM
Hmmkay, I take that as a ''No''. Not possible. Well then, resistor divider it is.
Hey ST, how about you add that feature to the SDADC MUX for a next hardware revision? Just like for the SAR ADC? That would allow a nice auto-calibration when using VCC or an external voltage as Vref. And this makes much more sense for the SDADC, given its higher resolution ... don't you think?2016-02-22 12:07 AM
The internal reference varies by + / - 3.3% over -40 .. +85 C. Would that have been good enough for you?
From the Data Sheet: –40 °C < TA < +85 °C 1.16 1.20 1.242016-02-25 06:53 PM
Hmm, does it really vary by so much when going from cold to hot?
Or is it only the initial setpoint that can be within +/- 3.3% but stays relatively constant afterwards? I mean, why should STM calibrate the reference and store the data in a register if the voltage varies so much afterwards?