16F1705 internal Opamp woes

[richard]

Testing with 200mV injected into pin 10 (non-inv input of Opamp1, Gain of 10) should
: see 2V at pin 8 output. When set as buffer should see 200mV. Both tests show 0V.

are you seeing this voltage at opamp out pin with a meter ?
your description is not clear whether the opamp is not working or the adc measurement .
working with opamps with inputs near rail potentials is never easy unless chip is designed to have rail to rail capability

[wjsmarine]

Hi Richard,

I knew you'd be the first to respond

I've been monitoring the output (pin 8) throughout the testing and it doesn't go above ground at any stage. Ultimately I want to process the opamp output with AN6 (or another if this isn't possible) and is how I initially breadboarded it, but when nothing worked I tried disabling the ADC module and repeated all tests. Zip, diddly squat, nada.

The data sheet and some of the App Notes rave about the specs of the internal opamp and, yes, it is Rail to Rail IO.

I have the pcb space to put in an external opamp (probably use an instrument amp) but don't see why I should and, sooner or later, I will need to use this feature so it's worth pursuing.

Cheers,
Bill

[tumbleweed]

OPA1CON = %1100000          ' Opamp 1 Enabled, SP set.

Isn't that missing a bit?
Try

OPA1CON = %11000000          ' Opamp 1 Enabled, SP set.

For OPAMP1 you have

' RC2  (pin 8)    AN6, Opamp1Out.       ### output - see notes.
' RC1  (pin 9)    AN5, Opamp1in-.       Junction of RF and RG.
' RC0  (pin 10)   AN4, Opamp1in+.       + sense of 0R1 series load resistor.

So, to use these for the opamp:
- set the TRIS bits for RC0, RC1, and RC2
- set ANSEL bits for AN4, AN5, and AN6

[wjsmarine]

Hi Tumbleweed,

You are a star, thank you.

I don't know how I missed that bit, correcting it saw the Opamp spring to life... in fact I tried a bunch of combinations with TRIS and ANSEL and it was difficult to make the Opamp fail to do its job!

It sounds obvious but to use the Opamp as a buffer one needs to remove any feedback resistors (of course).

Thanks again, I hope this is of use to others.

Kind regards,
Bill