nrf52840 and nrf52833 SAADC effective bit number question..

[colin_was_here]

Hi all,

We're starting to move to the nrf52840 platform and have a question about the adc..

We'd like about ~13 bits of effective resolution- right now we use an ads1115 externaly, but were wondering if it would be possible to use the internal adc...

The datasheets for both the 52840 and 33 are a bit sparse.. It does rate the enob at 3us aperture... this is only 9.1 bits, which really isn't adequate for what we're doing... What I'm wondering is if this improves a lot if you increase the aperture? what does this look like at 40 uS ? would we be able to oversample down to 13 bit? if so- how long would that take? would it be worth it power wise?? (we care much more about power than cost)

just to be clear- it's not the end of the world if we have to use an external one... We're still going to be using the nrf.. I'd rather an honest answer than an optimistic one ;)

Thanks, Colin

[haakonsh]

13ENOB is not possible with the nRF SAADC. 

Acquisition time is independent of the SAADC performance as long as it complies with Table 1 in the Acquisition time chapter of the PS.

Oversampling will certainly help with increasing the signal-to-noise ratio and there are steps that you can take to decrease the offset, but you can't affect the integral and differential non-linearity of the SAADC. 

The SAADC's gain error and offset are fairly temperature-dependent, the instrumentational precision like what the ads1115 provides is far outside our intended use-case, the pricing of the ads1115 reflects that :/ 

[colin_was_here]

Thanks, a bunch.. exactly what I wanted to know ;)