ADC measuring between VREF- and VREF+

Thread Starter

Potato Pudding

Joined Jun 11, 2010
688
Before I put a lot of work into this can somebody give me a quick check answer regarding some PIC features.

Confirm that I can constrain the ADC measurements between VREF- and VREF+. I would like to use that to limit measurements to within a 1 volt range and need to know if doing that will give a full 10 bit accuracy across that 1 Volt of measurement so that in that range I will get measurements to the Millivolt.

Have I misunderstood the documentation? (PIC16f887 checking 16f88# family datasheet F version dated 2009).

I was expecting to set up a complicated external window amplifier to fill the normal 5volt input range from that 1 volt window to get a similar effect, until I looked at the PIC data again.
 

Markd77

Joined Sep 7, 2009
2,806
There is this table which suggests the error may be above 1 Lsb with 1 V.
There is no mention of Vref- in the tables, which isn't very helpful.

 

Attachments

Thread Starter

Potato Pudding

Joined Jun 11, 2010
688
Interesting.

So it doesn't say that it won't work as far as I can tell.

Minimum error is +-1least significant bit at 2.7 Volts so +-2.7 mV.

I think that it could still be worth locating that 5.4mV error bar with 1 mV accuracy if the min and max Vrefs work the way that I am thinking they do.

I am just worried that I am missing another meaning for them, and assuming they do something different than intended.

I searched the Vrefs in the PDF and none of the references for them seemed to show anything else.

I would not bother with an external amplifier. I can get a limit to a 2.5 Volts range located in the middle or even at the top end of the 5 Volts supply range and that will be plenty. If I really wanted more accuracy I would use a device with higher resolution ADC, and there are plenty of those to choose from.

Being able to locate my sample range where the input naturally falls is actually more important.
 
Top